<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.1 20151215//EN" "http://jats.nlm.nih.gov/publishing/1.1/JATS-journalpublishing1.dtd">
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" article-type="research-article" dtd-version="1.1" specific-use="sps-1.9" xml:lang="pt">
  <front>
    <journal-meta>
      <journal-id journal-id-type="nlm-ta">Arq Bras Cardiol</journal-id>
      <journal-id journal-id-type="publisher-id">abc</journal-id>
      <journal-title-group>
        <journal-title>Arquivos Brasileiros de Cardiologia</journal-title>
        <abbrev-journal-title abbrev-type="publisher">Arq. Bras. Cardiol.</abbrev-journal-title>
      </journal-title-group>
      <issn pub-type="ppub">0066-782X</issn>
      <issn pub-type="epub">1678-4170</issn>
      <publisher>
        <publisher-name>Sociedade Brasileira de Cardiologia - SBC</publisher-name>
      </publisher>
    </journal-meta>
    <article-meta>
      <article-id specific-use="scielo-v3" pub-id-type="publisher-id">5XS3sr4LTQqRqNWT54VzSSt</article-id>
      <article-id specific-use="scielo-v2" pub-id-type="publisher-id">S0066-782X2021000701091</article-id>
      <article-id pub-id-type="publisher-id">abc.20190329</article-id>
      <article-id pub-id-type="doi">10.36660/abc.20190329</article-id>
      <article-categories>
        <subj-group subj-group-type="heading">
          <subject>Artigo Original</subject>
        </subj-group>
      </article-categories>
      <title-group>
        <article-title>Performance Diagn&#243;stica da FFR por Angiotomografia de Coron&#225;rias atrav&#233;s de Software Baseado em Intelig&#234;ncia Artificial</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0001-9652-2927</contrib-id>
          <name>
            <surname>Morais</surname>
            <given-names>Thamara Carvalho</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Assun&#231;&#227;o-Jr</surname>
            <given-names>Antonildes Nascimento</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-3053-3583</contrib-id>
          <name>
            <surname>Dantas</surname>
            <given-names>Roberto Nery</given-names>
            <suffix>J&#250;nior</suffix>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Silva</surname>
            <given-names>Carla Franco Grego da</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Paula</surname>
            <given-names>Caroline Bastida de</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0001-5547-6821</contrib-id>
          <name>
            <surname>Torres</surname>
            <given-names>Roberto Almeida</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-7232-4552</contrib-id>
          <name>
            <surname>Magalh&#227;es</surname>
            <given-names>Tiago Augusto</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
          <xref ref-type="aff" rid="aff3">
            <sup>3</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Nomura</surname>
            <given-names>C&#233;sar Higa</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>&#193;vila</surname>
            <given-names>Luiz Francisco Rodrigues de</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Parga</surname>
            <given-names>Jos&#233; Rodrigues</given-names>
            <suffix>Filho</suffix>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
      </contrib-group>
      <aff id="aff1">
        <label> 1 </label>
        <institution content-type="orgname">Hospital S&#237;rio-liban&#234;s</institution>
        <addr-line>
          <named-content content-type="city">S&#227;o Paulo</named-content>
          <named-content content-type="state">SP</named-content>
        </addr-line>
        <country country="BR">Brasil</country>
        <institution content-type="original"> Hospital S&#237;rio-liban&#234;s , S&#227;o Paulo , SP - Brasil </institution>
      </aff>
      <aff id="aff2">
        <label> 2 </label>
        <institution content-type="orgname">Universidade de S&#227;o Paulo</institution>
        <institution content-type="orgdiv1">Faculdade de Medicina</institution>
        <institution content-type="orgdiv2">Departamento de Imagem Cardiovascular</institution>
        <addr-line>
          <named-content content-type="city">S&#227;o Paulo</named-content>
          <named-content content-type="state">SP</named-content>
        </addr-line>
        <country country="BR">Brasil</country>
        <institution content-type="original"> Universidade de S&#227;o Paulo Faculdade de Medicina - CDI - InCor/HCFMUSP - Departamento de Imagem Cardiovascular , S&#227;o Paulo , SP - Brasil </institution>
      </aff>
      <aff id="aff3">
        <label> 3 </label>
        <institution content-type="orgdiv1">Complexo Hospital de Cl&#237;nicas</institution>
        <institution content-type="orgname">Universidade Federal do Paran&#225;</institution>
        <addr-line>
          <named-content content-type="city">Curitiba</named-content>
          <named-content content-type="state">PR</named-content>
        </addr-line>
        <country country="BR">Brasil</country>
        <institution content-type="original"> Complexo Hospital de Cl&#237;nicas da Universidade Federal do Paran&#225; (CHC-UFPR) - Cardiovascular CT/MR, Curitiba , PR - Brasil </institution>
      </aff>
      <author-notes>
        <corresp id="c01">Correspond&#234;ncia: Thamara Carvalho Morais &#8226; Hospital S&#237;rio-liban&#234;s - Rua Adma Jafet, 91. CEP 01308-050, S&#227;o Paulo, SP - Brasil E-mail: <email>thamaramorais@hotmail.com</email> </corresp>
        <fn fn-type="con">
          <p>Contribui&#231;&#227;o dos autores</p>
          <p>Concep&#231;&#227;o e desenho da pesquisa: Magalh&#227;es TA, Nomura CH, &#193;vila LFR, Parga Filho JR; Obten&#231;&#227;o de dados: Morais TC, Silva CFG, Paula CB, Torres RA, Magalh&#227;es TA; An&#225;lise e interpreta&#231;&#227;o dos dados: Assun&#231;&#227;o-Jr AN, Dantas J&#250;nior RN, Parga Filho JR; An&#225;lise estat&#237;stica: Assun&#231;&#227;o-Jr AN, Dantas J&#250;nior RN; Reda&#231;&#227;o do manuscrito: Morais TC, Assun&#231;&#227;o-Jr AN, Dantas J&#250;nior RN, Magalh&#227;es TA, Parga Filho JR; Revis&#227;o cr&#237;tica do manuscrito quanto ao conte&#250;do intelectual importante: Morais TC, Assun&#231;&#227;o-Jr AN, Dantas J&#250;nior RN, Nomura CH, &#193;vila LFR, Parga Filho JR.</p>
        </fn>
        <fn fn-type="conflict">
          <p>Potencial conflito de interesses</p>
          <p>Declaro n&#227;o haver conflito de interesses pertinentes.</p>
        </fn>
      </author-notes>
      <pub-date date-type="pub" publication-format="electronic">
        <day>14</day>
        <month>06</month>
        <year>2021</year>
      </pub-date>
      <pub-date date-type="collection" publication-format="electronic">
        <month>06</month>
        <year>2021</year>
      </pub-date>
      <volume>116</volume>
      <issue>6</issue>
      <fpage>1091</fpage>
      <lpage>1098</lpage>
      <history>
        <date date-type="received">
          <day>26</day>
          <month>05</month>
          <year>2019</year>
        </date>
        <date date-type="rev-recd">
          <day>11</day>
          <month>02</month>
          <year>2020</year>
        </date>
        <date date-type="accepted">
          <day>16</day>
          <month>03</month>
          <year>2020</year>
        </date>
      </history>
      <permissions>
        <license license-type="open-access" xlink:href="https://creativecommons.org/licenses/by/4.0/" xml:lang="en">
          <license-p>This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.</license-p>
        </license>
      </permissions>
      <abstract>
        <title>Resumo</title>
        <sec>
          <title>Fundamento</title>
          <p>A quantifica&#231;&#227;o n&#227;o invasiva da reserva fracionada de fluxo mioc&#225;rdico (FFR <sub>TC</sub> ) atrav&#233;s de <italic>software</italic> baseado em intelig&#234;ncia artificial em vers&#227;o mais atualizada e tom&#243;grafo de &#250;ltima gera&#231;&#227;o (384 cortes) apresenta elevada <italic>performance</italic> na detec&#231;&#227;o de isquemia coronariana.</p>
        </sec>
        <sec>
          <title>Objetivos</title>
          <p>Avaliar o desempenho diagn&#243;stico da FFR <sub>TC</sub> na detec&#231;&#227;o de doen&#231;a arterial coronariana (DAC) significativa em rela&#231;&#227;o ao FFRi, em tom&#243;grafos de gera&#231;&#245;es anteriores (128 e 256 cortes).</p>
        </sec>
        <sec>
          <title>M&#233;todos</title>
          <p>Estudo retrospectivo com pacientes encaminhados &#224; angiotomografia de art&#233;rias coron&#225;rias (TCC) e cateterismo (FFRi). Foram utilizados os tom&#243;grafos Siemens Somatom Definition Flash (256 cortes) e AS+ (128 cortes). A FFR <sub>TC</sub> e a &#225;rea luminal m&#237;nima (ALM) foram avaliadas em <italic>software</italic> (cFFR vers&#227;o 3.0.0, Siemens Healthineers, Forchheim, Alemanha). DAC obstrutiva foi definida como TCC com redu&#231;&#227;o luminal &#8805;50% e DAC funcionalmente obstrutiva como FFRi &#8804;0,8. Todos os valores de <italic>p</italic> reportados s&#227;o bicaudais; e quando &lt;0,05, foram considerados estatisticamente significativos.</p>
        </sec>
        <sec>
          <title>Resultados</title>
          <p>Noventa e tr&#234;s pacientes consecutivos (152 vasos) foram inclu&#237;dos. Houve boa concord&#226;ncia entre FFR <sub>TC</sub> e FFRi, com m&#237;nima superestima&#231;&#227;o da FFR <sub>TC</sub> (vi&#233;s: &#8211;0,02; limites de concord&#226;ncia: 0,14 a 0,09). Diferentes tom&#243;grafos n&#227;o modificaram a rela&#231;&#227;o entre FFR <sub>TC</sub> e FFRi (p para intera&#231;&#227;o = 0,73). A FFR <sub>TC</sub> demonstrou <italic>performance</italic> significativamente superior &#224; classifica&#231;&#227;o visual de estenose coronariana (AUC 0,93 <italic>vs.</italic> 0,61, p &lt;0,001) e &#224; ALM (AUC 0,93 <italic>vs.</italic> 0,75, p &lt;0,001) reduzindo o n&#250;mero de casos falso-positivos. O melhor ponto de corte para a FFR <sub>TC</sub> utilizando um &#237;ndice de Youden foi de 0,85 (sensiblidade, 87%; especificidade, 86%; VPP, 73%; NPV, 94%), com redu&#231;&#227;o de falso-positivos.</p>
        </sec>
        <sec>
          <title>Conclus&#227;o</title>
          <p>FFR <sub>TC</sub> baseada em intelig&#234;ncia artificial, em tom&#243;grafos de gera&#231;&#245;es anteriores (128 e 256 cortes), apresenta boa <italic>performance</italic> diagn&#243;stica na detec&#231;&#227;o de DAC, podendo ser utilizada para reduzir procedimentos invasivos.</p>
        </sec>
      </abstract>
      <kwd-group xml:lang="pt">
        <kwd>Reserva Fracionada de Fluxo Mioc&#225;rdico</kwd>
        <kwd>Doen&#231;a Arterial Coronariana</kwd>
        <kwd>Tomografia Computadorizada</kwd>
        <kwd>Isquemia Mioc&#225;rdica</kwd>
        <kwd>Aprendizado de M&#225;quina</kwd>
      </kwd-group>
      <counts>
        <fig-count count="8"/>
        <table-count count="4"/>
        <equation-count count="0"/>
        <ref-count count="20"/>
        <page-count count="8"/>
      </counts>
    </article-meta>
  </front>
  <body>
    <sec sec-type="intro">
      <title>Introdu&#231;&#227;o</title>
      <p>De acordo com as diretrizes cl&#237;nicas mais recentes, <sup><xref ref-type="bibr" rid="B1">1</xref> - <xref ref-type="bibr" rid="B3">3</xref></sup> o manejo da doen&#231;a arterial coronariana (DAC) cr&#244;nica e sintom&#225;tica pode ser guiado por testes adicionais para avalia&#231;&#227;o anat&#244;mica (extens&#227;o, severidade, morfologia) ou funcional (fun&#231;&#227;o ventricular, presen&#231;a/extens&#227;o de isquemia), com algumas evid&#234;ncias apontando superioridade da avalia&#231;&#227;o funcional sobre anat&#244;mica. <sup><xref ref-type="bibr" rid="B4">4</xref> - <xref ref-type="bibr" rid="B6">6</xref></sup></p>
      <p>Para este prop&#243;sito, sobretudo em pacientes com probabilidade pr&#233;-teste intermedi&#225;ria para DAC obstrutiva, a angiotomografia computadorizada de art&#233;rias coron&#225;rias (TCC) vem se destacando dentre os v&#225;rios testes n&#227;o invasivos como m&#233;todo robusto para descartar DAC obstrutiva, por seu elevado valor preditivo negativo. <sup><xref ref-type="bibr" rid="B7">7</xref></sup> Particularmente em estenoses moderadas (50% a 69%), a quantifica&#231;&#227;o n&#227;o invasiva da reserva fracionada de fluxo mioc&#225;rdico (FFR <sub>TC</sub> ) pode ajudar na correta discrimina&#231;&#227;o de quais destas est&#227;o associadas &#224; isquemia. <sup><xref ref-type="bibr" rid="B8">8</xref></sup> Estudos recentes demonstraram que a TCC tem elevada acur&#225;cia para identificar isquemia mioc&#225;rdica por meio da quantifica&#231;&#227;o n&#227;o invasiva da FFRTC quando comparada ao padr&#227;o-ouro, a FFR invasiva pelo cateterismo card&#237;aco (FFRi). <sup><xref ref-type="bibr" rid="B8">8</xref> - <xref ref-type="bibr" rid="B10">10</xref></sup></p>
      <p>A grande restri&#231;&#227;o do uso da FFR <sub>TC</sub> na pr&#225;tica cl&#237;nica t&#234;m sido sua baixa disponibilidade, especialmente devido &#224; necessidade de <italic>software</italic> espec&#237;fico que requeria supercomputadores em grandes centros internacionais, encarecendo e prolongando substancialmente o processo. <sup><xref ref-type="bibr" rid="B8">8</xref></sup> Recentemente, um prot&#243;tipo n&#227;o comercial de <italic>software</italic> (dispon&#237;vel para computadores pessoais de configura&#231;&#227;o-padr&#227;o) que utiliza ferramentas de intelig&#234;ncia artificial &#8211; rede neural convolucional ( <italic>deep learning</italic> ) &#8211; para avalia&#231;&#227;o da FFR <sub>TC</sub> foi testado por Rother et al., <sup><xref ref-type="bibr" rid="B11">11</xref></sup> Quando comparado com FFRi, a FFR <sub>TC</sub> calculada por esse <italic>software</italic> demonstrou elevada acur&#225;cia para detec&#231;&#227;o de isquemia, com significativa redu&#231;&#227;o no tempo de execu&#231;&#227;o do seu c&#225;lculo quando comparado com modelos j&#225; existentes que utilizam supercomputadores. <sup><xref ref-type="bibr" rid="B8">8</xref></sup> Deve-se ressaltar, contudo, que este trabalho utilizou apenas imagens de um tom&#243;grafo de &#250;ltima gera&#231;&#227;o (Siemens Somatom Force &#8211; 384 cortes). Uma vez que esse <italic>software</italic> se prop&#245;e a calcular a FFR <sub>TC</sub> em imagens adquiridas em tom&#243;grafos com diferentes tecnologias, objetivamos utiliz&#225;-lo para investigar a acur&#225;cia diagn&#243;stica da FFR <sub>TC</sub> em tom&#243;grafos de gera&#231;&#245;es anteriores, em compara&#231;&#227;o com FFRi, com qualidade de imagens que podem potencialmente afetar os resultados do algoritmo utilizado no <italic>software</italic> . Esse estudo tamb&#233;m comparou a acur&#225;cia diagn&#243;stica da FFR <sub>TC</sub> com a avalia&#231;&#227;o anat&#244;mica isolada pela TCC.</p>
    </sec>
    <sec sec-type="methods">
      <title>M&#233;todos</title>
      <sec>
        <title>Popula&#231;&#227;o do estudo</title>
        <p>Retrospectivamente, este estudo incluiu pacientes com sintomas sugestivos de DAC significativa encaminhados para TCC. Ap&#243;s os achados desse exame, foram encaminhados por decis&#227;o cl&#237;nica para o cateterismo card&#237;aco (com intervalo menor de 30 dias) e que realizaram an&#225;lise de FFRi no Hospital S&#237;rio-Liban&#234;s (S&#227;o Paulo-SP), entre janeiro de 2014 e fevereiro de 2018. No total, houve 17 exclus&#245;es: 14 por fatores limitantes para o c&#225;lculo da FFR <sub>TC</sub> descriminados pelo fabricante da ferramenta (8 por les&#227;o de tronco de coron&#225;ria esquerda [TCE], &#243;stios ou bifurca&#231;&#245;es; 6 pela presen&#231;a de <italic>stent</italic> ); e 3 por imagem com qualidade insuficiente devido a calcifica&#231;&#227;o excessiva e significativos artefatos de movimento. Durante o p&#243;s-processamento das imagens, n&#227;o houve exclus&#227;o de pacientes por impossibilidade t&#233;cnica de o <italic>software</italic> realizar a quantifica&#231;&#227;o da FFR <sub>TC.</sub> Ressalta-se que as recomenda&#231;&#245;es do fabricante com rela&#231;&#227;o a les&#245;es de TCE, &#243;stios ou bifurca&#231;&#245;es t&#234;m sido seguidas tamb&#233;m por outros autores, <sup><xref ref-type="bibr" rid="B11">11</xref></sup> e parecem estar relacionadas ao reconhecimento limitado das bordas anat&#244;micas nesses cen&#225;rios. Esse estudo foi aprovado pelo Comit&#234; de &#201;tica em Pesquisa do Hospital S&#237;rio-Liban&#234;s.</p>
      </sec>
      <sec>
        <title>Aquisi&#231;&#227;o de imagens da TCC</title>
        <p>As imagens foram obtidas utilizando-se os tom&#243;grafos Siemens Somatom Definition Flash de 256 cortes (resolu&#231;&#227;o temporal &#8211; TR &#8211; de 75ms; resolu&#231;&#227;o espacial &#8211; SR &#8211; de 0,30mm) e Somaton Definition AS+ de 128 cortes (TR de150ms; SR de 0,30mm) (Siemens Healthineers, Forchheim, Alemanha). O preparo do paciente seguiu recomenda&#231;&#245;es das diretrizes atuais, incluindo jejum de 4h, pun&#231;&#227;o venosa com jelco calibre 18G, preferencialmente &#224; direita, e monitoramento eletrocardiogr&#225;fico cont&#237;nuo. <sup><xref ref-type="bibr" rid="B12">12</xref></sup> Sempre que necess&#225;rio, para controle da frequ&#234;ncia card&#237;aca (FC), foi administrado betabloqueador (tartarato de metoprolol 50 a 100mg oral 1h antes do exame e/ou 5 a 20mg endovenoso poucos minutos antes da aquisi&#231;&#227;o), objetivando-se manter a FC em torno de 55 a 60bpm. Todos os pacientes receberam ainda nitrato sublingual (isordil 2,5mg) poucos minutos antes da aquisi&#231;&#227;o, exceto nos casos de hipotens&#227;o sintom&#225;tica ou uso de inibidores de fosfodiesterase tipo 5 (de acordo com o tempo de a&#231;&#227;o de cada f&#225;rmaco).</p>
        <p>A aquisi&#231;&#227;o das imagens foi planejada ap&#243;s a realiza&#231;&#227;o de teste <italic>bolus</italic> para c&#225;lculo do tempo de pico do contraste na aorta, com volume de 10 a 15mL, seguido de 30 a 50mL de soro fisiol&#243;gico a 4,5 a 5,5mL/s. As imagens foram adquiridas utilizando o modo <italic>Flash</italic> (no tom&#243;grafo <italic>Definition Flash</italic> ) ou de forma retrospectiva (em ambos os tom&#243;grafos) com acoplamento eletrocardiogr&#225;fico na di&#225;stole (55% a 75% do RR), corrente do tubo de 100 a 120kVp (ajustado pelo &#237;ndice de massa corporal [IMC] individual), tempo de rota&#231;&#227;o de 0,28 ( <italic>Flash</italic> )/0,33 (AS+) segundos, 160 a 320mAs e espessura de corte de 0,6/0,3mm. A infus&#227;o de contraste iodado Optiray 350 (Ioversol 350mg/mL, Mallinckrodt-EUA) utilizou os mesmos par&#226;metros do teste <italic>bolus</italic> (60 a 90mL).</p>
      </sec>
      <sec>
        <title>An&#225;lise de imagens da TCC e FFRct</title>
        <p>A an&#225;lise das imagens da TCC foi realizada por meio do <italic>software</italic> Syngo.via ( <italic>Siemens Healthineers, Forchheim, Germany</italic> ). Ap&#243;s a escolha de imagens de melhor qualidade t&#233;cnica, foram feitas avalia&#231;&#245;es da &#225;rvore coronariana em reformata&#231;&#227;o tridimensional, multiplanar e curva ( <italic>vessel probe</italic> ), com quantifica&#231;&#227;o do grau de estenose e composi&#231;&#227;o predominante da placa, quando presente (n&#227;o calcificada, calcificada e mista). A quantifica&#231;&#227;o da redu&#231;&#227;o luminal foi realizada conforme recomenda&#231;&#227;o da Sociedade de Tomografia Computadorizada Cardiovascular: <sup><xref ref-type="bibr" rid="B13">13</xref></sup> ausente, m&#237;nima (&lt;25%), discreta (25% a 49%), moderada (50% a 69%), importante (70% a 99%) e oclus&#227;o (100%). As estenoses foram tamb&#233;m classificadas como obstrutivas (&#8805;50%) e n&#227;o obstrutivas (&lt;50%) pela TCC.</p>
        <p>O p&#243;s-processamento da FFR <sub>TC</sub> foi feito na mesma s&#233;rie em que foi realizada a an&#225;lise visual anat&#244;mica (descrita anteriormente), utilizando a plataforma <italic>Frontier</italic> e o prot&#243;tipo n&#227;o comercial do <italic>software</italic> cFFR, vers&#227;o 3.0 (Siemens Healthineers, Forchheim, Alemanha) por um m&#233;dico experiente em imagem cardiovascular (&gt;4 anos).</p>
        <p>No c&#225;lculo da FFR <sub>TC</sub> , inicialmente, foi realizada a detec&#231;&#227;o autom&#225;tica da linha central e dos contornos do l&#250;men das art&#233;rias coron&#225;rias, que foram revisados e corrigidos pelo m&#233;dico especialista, quando necess&#225;rio. Na sequ&#234;ncia, o m&#233;dico definiu o limite superior e inferior de todas as placas dos vasos que tiveram os valores de FFRi calculados no cateterismo.</p>
        <p>Para a identifica&#231;&#227;o correta do local em que se calculou FFRi, o m&#233;dico hemodinamicista localizou o ponto de interesse nas imagens da fluoroscopia e documentou usando refer&#234;ncias anat&#244;micas (ramos das coron&#225;rias) e tamb&#233;m um modelo de segmenta&#231;&#227;o coronariana sugerido pela <italic>Society of Cardiovascular Computed Tomography</italic> (SCCT). O m&#233;dico especialista em TCC utilizou essa documenta&#231;&#227;o para selecionar a mesma localiza&#231;&#227;o anat&#244;mica das placas que tiverem a FFR <sub>TC</sub> calculada.</p>
        <p>O <italic>software</italic> cFFR calcula valores de FFR para toda a &#225;rvore coronariana com di&#226;metro &#8805;1,5mm e calcula automaticamente o valor ALM de cada placa delimitada. As etapas para c&#225;lculo da FFR <sub>TC</sub> est&#227;o demonstradas na <xref ref-type="fig" rid="f01">Figura 1</xref> . Esse <italic>software</italic> foi desenvolvido com novas ferramentas de intelig&#234;ncia artificial, utilizando-se t&#233;cnicas de aprendizado de m&#225;quina ( <italic>deep learning</italic> ). Todos eles foram instalados em computador padr&#227;o para laudos de radiologia. O tempo total de p&#243;s-processamento de todas as etapas foi em torno de 10 minutos.</p>
        <p>
          <fig id="f01">
            <label>Figura 1</label>
            <caption>
              <title>&#8211; Etapas para c&#225;lculo da FFR TC utilizando o software cFFR. A) Detec&#231;&#227;o visual da les&#227;o coronariana. Defini&#231;&#227;o do centerline B) e contornos luminais C) pelo software cFFR D) defini&#231;&#227;o dos limites da les&#227;o e do ponto de maior redu&#231;&#227;o luminal pelo operador. E) Resultado do FFR no ponto de maior redu&#231;&#227;o luminal demonstrado na &#225;rvore coronariana (ap&#243;s determinar centerline e contornos luminais nas tr&#234;s principais art&#233;rias coron&#225;rias: DA, Cx e CD).</title>
            </caption>
            <graphic xlink:href="1678-4170-abc-116-06-1091-90329-gf01.tif"/></fig>
        </p>
      </sec>
      <sec>
        <title>Cateterismo card&#237;aco e an&#225;lise de FFRi</title>
        <p>O cateterismo card&#237;aco foi realizado via radial ou femoral, utilizando cateteres diagn&#243;sticos de 6 ou 7 <italic>french</italic> (F). Para o c&#225;lculo de FFRi, o grau de estenose foi avaliado visualmente pelo hemodinamicista, utilizando-se pelo menos duas proje&#231;&#245;es ortogonais. As medidas de FFR foram realizadas com cateteres-guia de 6 ou 7F. Nitroglicerina intracoron&#225;ria (0,2mg) foi injetada em todos os pacientes antes dos angiogramas. Um fio-guia de monitoramento de press&#227;o foi posicionado distal &#224; les&#227;o-&#237;ndice, e as press&#245;es m&#233;dias foram registradas no momento em que ficam est&#225;veis. Adenosina intracoronariana foi injetada manualmente atrav&#233;s do cateter-guia, em inje&#231;&#227;o de <italic>bolus</italic> de 80&#956;g (art&#233;ria coron&#225;ria esquerda) ou 40&#956;g (coron&#225;ria direita) em 10mL de soro fisiol&#243;gico. Ap&#243;s sua administra&#231;&#227;o, o menor valor de FFR est&#225;vel durante o estado estacion&#225;rio hiper&#234;mico foi registrado. Esse valor corresponde &#224; raz&#227;o entre a press&#227;o coronariana m&#233;dia distal &#224; estenose e a press&#227;o a&#243;rtica m&#233;dia no momento da hiperemia induzida farmacologicamente.</p>
        <p>A exata posi&#231;&#227;o do sensor de medida de FFRi foi documentada pelo hemodinamicista no laudo, e esta documenta&#231;&#227;o foi utilizada pelo m&#233;dico especialista em TCC para a medida da FFR <sub>TC</sub> na mesma localiza&#231;&#227;o anat&#244;mica.</p>
      </sec>
      <sec>
        <title>An&#225;lise estat&#237;stica</title>
        <p>An&#225;lises descritivas foram expressas como frequ&#234;ncia (porcentagem) para vari&#225;veis categ&#243;ricas e como m&#233;dia &#177; desvio padr&#227;o para vari&#225;veis cont&#237;nuas. A distribui&#231;&#227;o das vari&#225;veis cont&#237;nuas foi avaliada por meio visual atrav&#233;s de gr&#225;ficos QQ e checada pelo teste de Shapiro-Wilk. Compara&#231;&#245;es entre as vari&#225;veis cont&#237;nuas encontradas na TCC e no cateterismo (grau de estenose coronariana, FFR) dos pacientes foram realizadas pelo teste t de Student para amostras pareadas. Da mesma forma, correla&#231;&#227;o entre essas vari&#225;veis foi realizada por meio de correla&#231;&#227;o de Pearson.</p>
        <p>Concord&#226;ncia entre FFR <sub>TC</sub> e FFRi foi determinada por an&#225;lise de Bland-Altman. Assumindo FFRi &#8804;0,8 como padr&#227;o-ouro para presen&#231;a de isquemia, a <italic>performance</italic> diagn&#243;stica da FFR <sub>TC</sub> e de outros par&#226;metros anat&#244;micos da TCC foi avaliada por meio do c&#225;lculo de sensibilidade, especificidade, valor preditivo positivo (VPP) e valor preditivo negativo (VPN). Al&#233;m disso, a &#225;rea sob a curva ROC (AUC) para detec&#231;&#227;o de les&#245;es coronarianas associadas com isquemia foi calculada, e compara&#231;&#245;es entre as AUC foram realizadas de acordo com o m&#233;todo descrito por DeLong et al., <sup><xref ref-type="bibr" rid="B14">14</xref></sup> O melhor ponto de corte da FFR <sub>TC</sub> para detec&#231;&#227;o de isquemia (FFRi &#8804;0,8) foi calculado utilizando-se o &#237;ndice de Youden, que corresponde &#224;quele com maior valor na equa&#231;&#227;o (sensibilidade [Sens] + especificidade [Espe] &#8211; 1). <sup><xref ref-type="bibr" rid="B15">15</xref></sup> Ressalta-se que o grau de estenose foi explorado como vari&#225;vel cont&#237;nua e categ&#243;rica (DAC obstrutiva, &gt;50%, ou n&#227;o). A escolha pela forma categ&#243;rica para inclus&#227;o no modelo se baseou em se tratar de um limiar clinicamente acion&#225;vel para tomadas de decis&#245;es cl&#237;nicas subsequentes.</p>
        <p>Considerando a potencial correla&#231;&#227;o entre m&#250;ltiplos vasos no mesmo indiv&#237;duo, o m&#233;todo de equa&#231;&#245;es de estima&#231;&#227;o generalizadas com estrutura de correla&#231;&#227;o permut&#225;vel foi usado para comparar amostras pareadas em um n&#237;vel por-vaso. As an&#225;lises estat&#237;sticas foram realizadas no <italic>software</italic> R ( <italic>R Foundation for Statistical Computing</italic> , Viena, &#193;ustria). Todos os valores de p reportados s&#227;o bicaudais e, quando &lt;0,05, foram considerados estatisticamente significativos.</p>
      </sec>
    </sec>
    <sec sec-type="results">
      <title>Resultados</title>
      <sec>
        <title>Caracter&#237;sticas dos pacientes e das placas</title>
        <p>Noventa e tr&#234;s pacientes foram inclu&#237;dos no estudo, com um total de 152 vasos. Cinquenta pacientes (54%) realizaram a TCC em tom&#243;grafo Flash (256 colunas de detectores) e 43 (46%) em tom&#243;grafo AS+ (128 colunas de detectores). A FC m&#233;dia nas aquisi&#231;&#245;es foi de 58 &#177; 8bpm.</p>
        <p>Setenta e quatro pacientes (80%) apresentavam DAC obstrutiva (estenose &gt;50%) na TCC, sendo 48 com estenose moderada (50% a 69%) e 26 importante (&gt;70%). Na an&#225;lise por-vaso, as placas foram mais frequentemente mistas (70%), com localiza&#231;&#227;o mais comum (73%) na coron&#225;ria descendente anterior (DA) e com ALM m&#233;dia de 3,2 &#177; 1,6mm <sup>2</sup> . Caracter&#237;sticas cl&#237;nicas e tomogr&#225;ficas dos pacientes est&#227;o expostas nas <xref ref-type="table" rid="t1">Tabelas 1</xref> e <xref ref-type="table" rid="t2">2</xref> , respectivamente.</p>
        <p>
          <table-wrap id="t1">
            <label>Tabela 1</label>
            <caption>
              <title>&#8211; Dados demogr&#225;ficos</title>
            </caption>
            <table frame="hsides" rules="groups">
              <colgroup width="50%">
                <col/>
                <col/>
              </colgroup>
              <thead>
                <tr>
                  <th align="left">Vari&#225;veis</th>
                  <th>n = 93</th>
                </tr>
              </thead>
              <tbody>
                <tr>
                  <td>Idade, anos*</td>
                  <td align="center">64 &#177; 11</td>
                </tr>
                <tr>
                  <td>Sexo masculino, n (%)</td>
                  <td align="center">70 (75)</td>
                </tr>
                <tr>
                  <td>Hipertens&#227;o, n (%)</td>
                  <td align="center">54 (58)</td>
                </tr>
                <tr>
                  <td>Dislipidemia, n (%)</td>
                  <td align="center">45 (48)</td>
                </tr>
                <tr>
                  <td>Diabetes, n (%)</td>
                  <td align="center">24 (26)</td>
                </tr>
                <tr>
                  <td>Tabagismo, n (%)</td>
                  <td align="center">7 (8)</td>
                </tr>
                <tr>
                  <td>IMC, kg/m <sup>2</sup> *</td>
                  <td align="center">28 &#177; 4</td>
                </tr>
                <tr>
                  <td>FC, bpm*</td>
                  <td align="center">58 &#177; 8</td>
                </tr>
              </tbody>
            </table>
            <table-wrap-foot>
              <fn id="TFN1">
                <p>
                  <italic>IMC: &#237;ndice de massa corporal; FC: frequ&#234;ncia card&#237;aca. *m&#233;dia &#177; desvio padr&#227;o.</italic>
                </p>
              </fn>
            </table-wrap-foot>
          </table-wrap>
        </p>
        <p>
          <table-wrap id="t2">
            <label>Tabela 2</label>
            <caption>
              <title>&#8211; an&#225;lise do FFRTC e FFRi</title>
            </caption>
            <table frame="hsides" rules="groups">
              <colgroup width="50%">
                <col/>
                <col/>
              </colgroup>
              <thead>
                <tr>
                  <th align="left">Por paciente</th>
                  <th>n = 93</th>
                </tr>
              </thead>
              <tbody>
                <tr>
                  <td>Estenose &#8805; 50%, n (%)</td>
                  <td align="center">74 (80)</td>
                </tr>
                <tr>
                  <td>Estenose 50-69%, n (%)</td>
                  <td align="center">48 (52)</td>
                </tr>
                <tr>
                  <td>Estenose &#8805; 70%, n (%)</td>
                  <td align="center">26 (28)</td>
                </tr>
                <tr>
                  <td>FFR <sub>TC</sub> &#8804; 0,8, n (%)</td>
                  <td align="center">32 (34)</td>
                </tr>
                <tr>
                  <td>FFRi &#8804; 0,8, n (%)</td>
                  <td align="center">39 (42)</td>
                </tr>
                <tr>
                  <td>
                    <bold>Por vaso</bold>
                  </td>
                  <td align="center">
                    <bold>n = 152</bold>
                  </td>
                </tr>
                <tr>
                  <td>
                    <bold>Localiza&#231;&#227;o</bold>
                  </td>
                  <td> </td>
                </tr>
                <tr>
                  <td>DA, n (%)</td>
                  <td align="center">111 (73)</td>
                </tr>
                <tr>
                  <td>Cx, n (%)</td>
                  <td align="center">26 (17)</td>
                </tr>
                <tr>
                  <td>CD, n (%)</td>
                  <td align="center">16 (10)</td>
                </tr>
                <tr>
                  <td>Estenose &#8805; 50%, n (%)</td>
                  <td align="center">124 (82)</td>
                </tr>
                <tr>
                  <td>Estenose 50-69%, n (%)</td>
                  <td align="center">95 (63)</td>
                </tr>
                <tr>
                  <td>Estenose &#8805; 70%, n (%)</td>
                  <td align="center">29 (19)</td>
                </tr>
                <tr>
                  <td>ALM, mm <sup>2</sup> *</td>
                  <td align="center">3,2 &#177; 1,6</td>
                </tr>
                <tr>
                  <td>
                    <bold>Morfologia</bold>
                  </td>
                  <td> </td>
                </tr>
                <tr>
                  <td>Calcificada, n (%)</td>
                  <td align="center">16 (10)</td>
                </tr>
                <tr>
                  <td>Mista, n (%)</td>
                  <td align="center">106 (70)</td>
                </tr>
                <tr>
                  <td>N&#227;o calcificada, n (%)</td>
                  <td align="center">30 (20)</td>
                </tr>
                <tr>
                  <td>
                    <bold>FFR <sub>TC</sub> *</bold>
                  </td>
                  <td align="center">
                    <bold>0,88 &#177; 0,08</bold>
                  </td>
                </tr>
                <tr>
                  <td>FFR <sub>TC</sub> &#8804; 0,8, n (%)</td>
                  <td align="center">32 (21)</td>
                </tr>
                <tr>
                  <td>
                    <bold>FFRi*</bold>
                  </td>
                  <td align="center">
                    <bold>0,86 &#177; 0,08</bold>
                  </td>
                </tr>
                <tr>
                  <td>FFRi &#8804; 0,8, n (%)</td>
                  <td align="center">47 (31)</td>
                </tr>
              </tbody>
            </table>
            <table-wrap-foot>
              <fn id="TFN2">
                <p>
                  <italic>DAC: doen&#231;a arterial coronariana; FFR: reserva fracionada de fluxo invasiva; FFR <sub>TC</sub>: reserva fracionada de fluxo por tomografia; ALM: &#225;rea luminal m&#237;nima. <sup>*</sup> m&#233;dia &#177; desvio padr&#227;o.</italic>
                </p>
              </fn>
            </table-wrap-foot>
          </table-wrap>
        </p>
      </sec>
      <sec>
        <title>Compara&#231;&#227;o de FFR TC com FFRi</title>
        <p>Houve forte correla&#231;&#227;o entre os valores da FFR <sub>TC</sub> e FFRi (r = 0,73, p&lt;0,001) ( <xref ref-type="fig" rid="f02">Figura 2</xref> ). Na m&#233;dia, os valores da FFR <sub>TC</sub> foram discretamente superiores aos encontrados em FFRi (0,88 &#177; 0,08 <italic>vs.</italic> 0,86 &#177; 0,08, p = 0,02), erro sistem&#225;tico que se confirma na an&#225;lise de Bland-Altman (vi&#233;s de &#8211;0,02 com intervalo de confian&#231;a de &#8211;0,14 a 0,09) (ver <xref ref-type="fig" rid="f02">Figura 2</xref> ). O tipo de tom&#243;grafo utilizado n&#227;o mudou a rela&#231;&#227;o entre FFRi e FFR <sub>TC</sub> (p-valor para intera&#231;&#227;o de 0,73).</p>
        <p>
          <fig id="f02">
            <label>Figura 2</label>
            <caption>
              <title>&#8211; Correla&#231;&#227;o (A) e concord&#226;ncia por an&#225;lise de Bland-Altman (B) entre FFR TC e FFRi (an&#225;lise por vaso): APARELHO: AS refere-se a tom&#243;grafo de 128 cortes e FLASH, a tom&#243;grafo de 256 cortes.</title>
            </caption>
            <graphic xlink:href="1678-4170-abc-116-06-1091-90329-gf02.tif"/></fig>
        </p>
      </sec>
      <sec>
        <title>Detec&#231;&#227;o de isquemia</title>
        <p>Para identifica&#231;&#227;o de les&#245;es coronarianas obstrutivas com limita&#231;&#227;o de fluxo (FFRi &#8804;0,8 como padr&#227;o-ouro), a FFR <sub>TC</sub> demonstrou <italic>performance</italic> significativamente superior &#224; classifica&#231;&#227;o visual isolada de obstru&#231;&#227;o coronariana (AUC 0,93 <italic>vs.</italic> 0,61, p &lt;0,001) e &#224; ALM pela TCC (AUC 0,93 <italic>vs.</italic> 0,75, p &lt;0,001) ( <xref ref-type="fig" rid="f03">Figura 3</xref> ). O melhor ponto de corte (com menor n&#250;mero de resultados falsos) para a FFR <sub>TC</sub> definido por meio do &#237;ndice de Youden, para discrimina&#231;&#227;o de les&#227;o com ou sem isquemia, foi 0,85, que apresentou valores de sensibilidade de 87%, especificidade de 86%, VPP de 73% e VPN de 94% nesse ponto de corte ( <xref ref-type="fig" rid="f04">Figura 4</xref> ). Tais m&#233;tricas de <italic>performance</italic> utilizando-se esse ponto de corte (0,85) foram discretamente superiores analisando-se apenas as placas com redu&#231;&#227;o luminal moderada (50% a 69%, n = 95), com sensibilidade de 89%, especificidade de 91%, VPP de 74%, VPN de 97%. Das 152 les&#245;es avaliadas, 3 (2%) foram falso-positivas e 18 (12%), falso-negativas, utilizando-se o ponto de corte tradicional (FFR <sub>TC</sub> &#8804;0,80). Por meio do ponto de corte mais elevado (FFR <sub>TC</sub> &lt; 0,85), 12 (7%) foram falso-positivas e 9 (6%), falso-negativas.</p>
        <p>
          <fig id="f03">
            <label>Figura 3</label>
            <caption>
              <title>&#8211; Performance para o diagn&#243;stico de les&#227;o obstrutiva com limita&#231;&#227;o de fluxo (FFRi &lt;0,8).</title>
            </caption>
            <graphic xlink:href="1678-4170-abc-116-06-1091-90329-gf03.tif"/></fig>
        </p>
        <p>
          <fig id="f04">
            <label>Figura 4</label>
            <caption>
              <title>&#8211; Performance diagn&#243;stica do valor de FFR TC &lt;0,85.APARELHO: AS refere-se ao tom&#243;grafo de 128 cortes e FLASH, ao tom&#243;grafo de 256 cortes.</title>
            </caption>
            <graphic xlink:href="1678-4170-abc-116-06-1091-90329-gf04.tif"/></fig>
        </p>
        <p>Ao avaliarmos o grau de redu&#231;&#227;o luminal, as placas visualmente consideradas moderadas (50% a 69%) tiveram um n&#250;mero de falso-positivos de 86 (56%) em rela&#231;&#227;o ao padr&#227;o-ouro (FFRi &#8804;0,8), enquanto as placas com redu&#231;&#227;o luminal visual importante (&#8805;70%) tiveram um n&#250;mero de falso-positivos de 23 (15%), o que, para esta &#250;ltima, representa uma magnitude 50% maior em rela&#231;&#227;o aos resultados da FFR <sub>TC</sub> &lt;0,85 (15% <italic>vs.</italic> 7%).</p>
      </sec>
    </sec>
    <sec sec-type="discussion">
      <title>Discuss&#227;o</title>
      <p>A an&#225;lise da FFR <sub>TC</sub> atrav&#233;s de <italic>software</italic> baseado em aprendizado de m&#225;quina demonstrou boa concord&#226;ncia em rela&#231;&#227;o &#224; medida de FFRi, destacando-se que o p&#243;s-processamento das imagens de TCC foi feito em computadores-padr&#227;o do pr&#243;prio hospital e com tempo de p&#243;s-processamento em torno de 10min. Com rela&#231;&#227;o &#224; <italic>performance</italic> diagn&#243;stica, mesmo em tom&#243;grafos de gera&#231;&#245;es anteriores, a FFR <sub>TC</sub> foi superior &#224; avalia&#231;&#227;o anat&#244;mica isolada tanto na observa&#231;&#227;o de estenoses coronarianas como no c&#225;lculo da ALM, reduzindo significativamente o n&#250;mero de falso-positivos.</p>
      <p>Em linha com o demonstrado por Rother et al., <sup><xref ref-type="bibr" rid="B11">11</xref></sup> retrospectivamente em coorte com 71 pacientes e utilizando a mesma vers&#227;o de <italic>software</italic> desse estudo (cFFR vers&#227;o 3.0), <sup><xref ref-type="bibr" rid="B11">11</xref></sup> a FFR <sub>TC</sub> apresentou consider&#225;vel concord&#226;ncia com a medida de FFRi, com m&#237;nima superestima&#231;&#227;o. Esses resultados est&#227;o em desencontro com vers&#245;es anteriores desse mesmo <italic>software</italic> (cFFR vers&#227;o 1.4), <sup><xref ref-type="bibr" rid="B16">16</xref> - <xref ref-type="bibr" rid="B19">19</xref></sup> em que uma subestima&#231;&#227;o foi descrita, e provavelmente reflete mudan&#231;as do algoritmo com a atualiza&#231;&#227;o do <italic>software</italic> .</p>
      <p>Apesar de compar&#225;vel com os tr&#234;s principais estudos multic&#234;ntricos at&#233; hoje publicados (DISCOVER-FLOW, DeFacto and NXT), <sup><xref ref-type="bibr" rid="B8">8</xref> - <xref ref-type="bibr" rid="B10">10</xref></sup> h&#225; que se real&#231;ar que o limite de concord&#226;ncia do nosso estudo foi mais largo na an&#225;lise de Bland-Altman (~0,20), que significa menor repetibilidade do m&#233;todo, em compara&#231;&#227;o com o observado por Rother et al., <sup><xref ref-type="bibr" rid="B11">11</xref></sup> Como, na m&#233;dia, os pacientes do nosso estudo apresentaram uma FC m&#233;dia &lt;60 bpm, o que reflete uma boa qualidade das imagens em geral, acreditamos que o desempenho superior daquele estudo pode ser explicado, em parte, pela utiliza&#231;&#227;o de um tom&#243;grafo com resolu&#231;&#227;o espacial 20% superior (0,3 <italic>vs.</italic> 0,24mm), al&#233;m do uso de um moderno algoritmo de reconstru&#231;&#227;o (ADMIRE). Esses fatores podem ter levado a uma melhor detec&#231;&#227;o dos contornos coronarianos (linha central e l&#250;men) por aquele estudo, com consequente melhora dos resultados. Outra justificativa que n&#227;o pode ser descartada seria a maior experi&#234;ncia do observador daquele centro com a nova vers&#227;o de cFFR.</p>
      <p>Com rela&#231;&#227;o ao poder de discrimina&#231;&#227;o de estenoses coronarianas com/sem limita&#231;&#227;o de fluxo, a FFR <sub>TC</sub> foi superior em compara&#231;&#227;o com a avalia&#231;&#227;o isolada anat&#244;mica da TCC, tanto qualitativamente (classifica&#231;&#227;o visual de DAC obstrutiva) quanto quantitativamente (ALM). Utilizando-se um ponto de corte de 0,85 para a FFR <sub>TC</sub> , os VPN e VPP foram compar&#225;veis com os de outras coortes que utilizaram esse <italic>software.</italic> <sup><xref ref-type="bibr" rid="B16">16</xref> - <xref ref-type="bibr" rid="B19">19</xref></sup> Al&#233;m disso, ressaltamos os seguintes aspectos: 1) a <italic>performance</italic> da FFR <sub>TC</sub> foi melhor nos casos com les&#245;es moderadas (50% a 69%); 2) a FFR <sub>TC</sub> levou a uma redu&#231;&#227;o de mais de 50% dos casos falso-positivos quando usada somente a avalia&#231;&#227;o anat&#244;mica de DAC grave (&#8805;70%). Esses achados apresentam grande relev&#226;ncia na pr&#225;tica cl&#237;nica, uma vez que les&#245;es moderadas na TCC s&#227;o relativamente frequentes e, muitas vezes, esses pacientes s&#227;o encaminhados para exames adicionais. <sup><xref ref-type="bibr" rid="B20">20</xref></sup> De fato, a oportunidade de redu&#231;&#227;o global de encaminhamento desnecess&#225;rio para cateterismo pode ser ainda maior utilizando-se essa nova ferramenta de FFR <sub>TC</sub> , uma vez que apenas 42% dos nossos pacientes apresentaram FFRi &lt;0,8.</p>
      <p>Por fim, destaca-se o r&#225;pido p&#243;s-processamento de imagens desse novo <italic>software</italic> com base em tecnologia de aprendizado de m&#225;quina ( <italic>deep learning</italic> ). Nos <italic>softwares</italic> pioneiros <sup><xref ref-type="bibr" rid="B8">8</xref> - <xref ref-type="bibr" rid="B10">10</xref></sup> que utilizam algoritmos de din&#226;mica de fluidos, o c&#225;lculo da FFR <sub>TC</sub> leva de 1 a 4h de processamento, e &#233; realizado em supercomputadores localizados apenas em centros espec&#237;ficos dos EUA (sede na Calif&#243;rnia), Londres e T&#243;quio. Al&#233;m do elevado custo, no geral, isso demanda cerca de 24h para a obten&#231;&#227;o dos resultados e necessidade de enviar imagens DICOM para fora de ambientes da institui&#231;&#227;o. Portanto, esse novo <italic>software</italic> poder&#225; promover impacto real na pr&#225;tica cl&#237;nica no cuidado do paciente com DAC.</p>
      <sec>
        <title>Limita&#231;&#245;es</title>
        <p>Trata-se de um estudo retrospectivo, unic&#234;ntrico, com uma popula&#231;&#227;o de estudo relativamente pequena e predominantemente apresentando DAC obstrutiva. Ao seguirmos as recomenda&#231;&#245;es de aplicabilidade da ferramenta pelo fabricante, foram exclu&#237;dos os pacientes com estenose significativa em tronco de coron&#225;ria esquerda, &#243;stios de coron&#225;rias principais ou em bifurca&#231;&#245;es; oclus&#245;es arteriais cr&#244;nicas; hist&#243;ria pr&#233;via de cirurgia de revasculariza&#231;&#227;o ou implante de <italic>stent</italic> . Da mesma forma, pacientes com sintomas t&#237;picos eventualmente n&#227;o foram submetidos a exame funcional invasivo (FFRi) por decis&#227;o cl&#237;nica. Portanto, esse estudo deve ser interpretado com a devida aten&#231;&#227;o ao contexto cl&#237;nico da popula&#231;&#227;o participante (DAC menos grave/complexa e/ou cen&#225;rios cl&#237;nicos de maior d&#250;vida diagn&#243;stica).</p>
      </sec>
    </sec>
    <sec sec-type="conclusions">
      <title>Conclus&#227;o</title>
      <p>Essa nova vers&#227;o da FFR <sub>TC</sub> , mesmo em tom&#243;grafos de gera&#231;&#245;es anteriores, apresentou boa <italic>performance</italic> diagn&#243;stica na detec&#231;&#227;o de les&#245;es coronarianas obstrutivas com limita&#231;&#227;o de fluxo, com redu&#231;&#227;o expressiva do n&#250;mero de falso-positivos, o que pode reduzir significativamente a quantidade de pacientes que realizam testes adicionais. A import&#226;ncia cl&#237;nica desses achados precisa ser validada por estudos especificamente desenhados para avalia&#231;&#245;es de desfechos cl&#237;nicos. Esse <italic>software</italic> apresenta tecnologia inovadora, utilizando aprendizado de m&#225;quina, o que possibilita maior acessibilidade, r&#225;pida execu&#231;&#227;o e, potencialmente, redu&#231;&#227;o de custos.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <title>Refer&#234;ncias</title>
      <ref id="B1">
        <label>1</label>
        <mixed-citation>1. Fihn SD, Blankenship JC, Alexander KP,Bittl JA, Byrne JG, Fletcher BJ, et al. et al. 2014 ACC/AHA/AATS/PCNA/SCAI/STS Focused Update of the Guideline for the Diagnosis and Management of Patients With Stable Ischemic Heart Disease. <italic>J Am Coll Cardiol</italic> . 2014; 64(18):19291949. doi:10.1016/j.jacc.2014.07.017.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Fihn</surname>
              <given-names>SD</given-names>
            </name>
            <name>
              <surname>Blankenship</surname>
              <given-names>JC</given-names>
            </name>
            <name>
              <surname>Alexander</surname>
              <given-names>KP</given-names>
            </name>
            <name>
              <surname>Bittl</surname>
              <given-names>JA</given-names>
            </name>
            <name>
              <surname>Byrne</surname>
              <given-names>JG</given-names>
            </name>
            <name>
              <surname>Fletcher</surname>
              <given-names>BJ</given-names>
            </name>
            <etal>et al. et al</etal>
          </person-group>
          <article-title>2014 ACC/AHA/AATS/PCNA/SCAI/STS Focused Update of the Guideline for the Diagnosis and Management of Patients With Stable Ischemic Heart Disease</article-title>
          <source>J Am Coll Cardiol</source>
          <year>2014</year>
          <volume>64</volume>
          <issue>18</issue>
          <size units="pages">19291949</size>
          <pub-id pub-id-type="doi">10.1016/j.jacc.2014.07.017</pub-id>
        </element-citation>
      </ref>
      <ref id="B2">
        <label>2</label>
        <mixed-citation>2. Task Force Members, Montalescot G, Sechtem U, Achenbach S, Andreotti F, Arden C, et al. 2013 ESC guidelines on the management of stable coronary artery disease. <italic>Eur Heart J</italic> . 2013; 34(38):2949-3003. doi:10.1093/eurheartj/eht296.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Members</surname>
              <given-names>Task Force</given-names>
            </name>
            <name>
              <surname>Montalescot</surname>
              <given-names>G</given-names>
            </name>
            <name>
              <surname>Sechtem</surname>
              <given-names>U</given-names>
            </name>
            <name>
              <surname>Achenbach</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Andreotti</surname>
              <given-names>F</given-names>
            </name>
            <name>
              <surname>Arden</surname>
              <given-names>C</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>2013 ESC guidelines on the management of stable coronary artery disease</article-title>
          <source>Eur Heart J</source>
          <year>2013</year>
          <volume>34</volume>
          <issue>38</issue>
          <fpage>2949</fpage>
          <lpage>3003</lpage>
          <pub-id pub-id-type="doi">10.1093/eurheartj/eht296</pub-id>
        </element-citation>
      </ref>
      <ref id="B3">
        <label>3</label>
        <mixed-citation>3. Cesar LA, Ferreira JF, Armaganyan D, Gowdak LH, Mansur AP, Bodanese LC, et al. Sociedade Brasileira de Cardiologia Diretriz De Doen&#231;a Coron&#225;ria Est&#225;vel. <italic>Arq Bras Cardiol</italic> . 2014; 103(2 Suppl 2):1-59. doi:10.5935/abc.2014S004.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Cesar</surname>
              <given-names>LA</given-names>
            </name>
            <name>
              <surname>Ferreira</surname>
              <given-names>JF</given-names>
            </name>
            <name>
              <surname>Armaganyan</surname>
              <given-names>D</given-names>
            </name>
            <name>
              <surname>Gowdak</surname>
              <given-names>LH</given-names>
            </name>
            <name>
              <surname>Mansur</surname>
              <given-names>AP</given-names>
            </name>
            <name>
              <surname>Bodanese</surname>
              <given-names>LC</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Sociedade Brasileira de Cardiologia Diretriz De Doen&#231;a Coron&#225;ria Est&#225;vel</article-title>
          <source>Arq Bras Cardiol</source>
          <year>2014</year>
          <volume>103</volume>
          <issue>2</issue>
          <supplement>Suppl 2</supplement>
          <fpage>1</fpage>
          <lpage>59</lpage>
          <pub-id pub-id-type="doi">10.5935/abc.2014S004</pub-id>
        </element-citation>
      </ref>
      <ref id="B4">
        <label>4</label>
        <mixed-citation>4. Little WC, Constantinescu M, Applegate RJ, Kutcher MA, Burrows MT, Kahl FT, et al. Can coronary angiography predict the site of a subsequent myocardial infarction in patients with mild-to-moderate coronary artery disease? <italic>Circulation</italic> . 1988; 78(5 Pt 1):1157-66.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Little</surname>
              <given-names>WC</given-names>
            </name>
            <name>
              <surname>Constantinescu</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Applegate</surname>
              <given-names>RJ</given-names>
            </name>
            <name>
              <surname>Kutcher</surname>
              <given-names>MA</given-names>
            </name>
            <name>
              <surname>Burrows</surname>
              <given-names>MT</given-names>
            </name>
            <name>
              <surname>Kahl</surname>
              <given-names>FT</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Can coronary angiography predict the site of a subsequent myocardial infarction in patients with mild-to-moderate coronary artery disease?</article-title>
          <source>Circulation</source>
          <year>1988</year>
          <volume>78</volume>
          <issue>5</issue>
          <comment>Pt 1</comment>
          <fpage>1157</fpage>
          <lpage>1166</lpage>
        </element-citation>
      </ref>
      <ref id="B5">
        <label>5</label>
        <mixed-citation>5. Shaw LJ, Berman DS, Maron DJ, Mancini GB, Hayes SW, Hartigan PM, et al. Optimal medical therapy with or without percutaneous coronary intervention to reduce ischemic burden. <italic>Circulation</italic> . 2008; 117(10):1283-91. doi:10.1161/CIRCULATIONAHA.107.743963.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Shaw</surname>
              <given-names>LJ</given-names>
            </name>
            <name>
              <surname>Berman</surname>
              <given-names>DS</given-names>
            </name>
            <name>
              <surname>Maron</surname>
              <given-names>DJ</given-names>
            </name>
            <name>
              <surname>Mancini</surname>
              <given-names>GB</given-names>
            </name>
            <name>
              <surname>Hayes</surname>
              <given-names>SW</given-names>
            </name>
            <name>
              <surname>Hartigan</surname>
              <given-names>PM</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Optimal medical therapy with or without percutaneous coronary intervention to reduce ischemic burden</article-title>
          <source>Circulation</source>
          <year>2008</year>
          <volume>117</volume>
          <issue>10</issue>
          <fpage>1283</fpage>
          <lpage>1291</lpage>
          <pub-id pub-id-type="doi">10.1161/CIRCULATIONAHA.107.743963</pub-id>
        </element-citation>
      </ref>
      <ref id="B6">
        <label>6</label>
        <mixed-citation>6. Tonino PAL, De Bruyne B, Pijls NHJ, Siebert U, Ikeno F, van Veer M, Klauss Y, et al. Fractional flow reserve versus angiography for guiding percutaneous coronary intervention. <italic>N Engl J Med</italic> . 2009; 360(3):213-24. doi:10.1056/NEJMoa0807611.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Tonino</surname>
              <given-names>PAL</given-names>
            </name>
            <name>
              <surname>De Bruyne</surname>
              <given-names>B</given-names>
            </name>
            <name>
              <surname>Pijls</surname>
              <given-names>NHJ</given-names>
            </name>
            <name>
              <surname>Siebert</surname>
              <given-names>U</given-names>
            </name>
            <name>
              <surname>Ikeno</surname>
              <given-names>F</given-names>
            </name>
            <name>
              <surname>van Veer</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Klauss</surname>
              <given-names>Y</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Fractional flow reserve versus angiography for guiding percutaneous coronary intervention</article-title>
          <source>N Engl J Med</source>
          <year>2009</year>
          <volume>360</volume>
          <issue>3</issue>
          <fpage>213</fpage>
          <lpage>224</lpage>
          <pub-id pub-id-type="doi">10.1056/NEJMoa0807611</pub-id>
        </element-citation>
      </ref>
      <ref id="B7">
        <label>7</label>
        <mixed-citation>7. Nielsen LH, Ortner N, Norgaard BL, Achenbach S, Leipsic J, Abdulla J. et al. The diagnostic accuracy and outcomes after coronary computed tomography angiography vs. conventional functional testing in patients with stable angina pectoris: a systematic review and meta-analysis. <italic>Eur Hear J &#8211; Cardiovasc Imaging</italic> . 2014; 15(9):961-71. doi:10.1093/ehjci/jeu027.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Nielsen</surname>
              <given-names>LH</given-names>
            </name>
            <name>
              <surname>Ortner</surname>
              <given-names>N</given-names>
            </name>
            <name>
              <surname>Norgaard</surname>
              <given-names>BL</given-names>
            </name>
            <name>
              <surname>Achenbach</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Leipsic</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Abdulla</surname>
              <given-names>J.</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>The diagnostic accuracy and outcomes after coronary computed tomography angiography vs. conventional functional testing in patients with stable angina pectoris: a systematic review and meta-analysis</article-title>
          <source>Eur Hear J &#8211; Cardiovasc Imaging</source>
          <year>2014</year>
          <volume>15</volume>
          <issue>9</issue>
          <fpage>961</fpage>
          <lpage>971</lpage>
          <pub-id pub-id-type="doi">10.1093/ehjci/jeu027</pub-id>
        </element-citation>
      </ref>
      <ref id="B8">
        <label>8</label>
        <mixed-citation>8. N&#248;rgaard BL, Leipsic J, Gaur S, Seneviratne S, Ko BS, Ito H, et al. Diagnostic performance of noninvasive fractional flow reserve derived from coronary computed tomography angiography in suspected coronary artery disease: The NXT trial (Analysis of Coronary Blood Flow Using CT Angiography: Next Steps). <italic>J Am Coll Cardiol</italic> . 2014; 63(12):1145-55. doi:10.1016/j.jacc.2013.11.043.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>N&#248;rgaard</surname>
              <given-names>BL</given-names>
            </name>
            <name>
              <surname>Leipsic</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Gaur</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Seneviratne</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Ko</surname>
              <given-names>BS</given-names>
            </name>
            <name>
              <surname>Ito</surname>
              <given-names>H</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Diagnostic performance of noninvasive fractional flow reserve derived from coronary computed tomography angiography in suspected coronary artery disease: The NXT trial (Analysis of Coronary Blood Flow Using CT Angiography: Next Steps)</article-title>
          <source>J Am Coll Cardiol</source>
          <year>2014</year>
          <volume>63</volume>
          <issue>12</issue>
          <fpage>1145</fpage>
          <lpage>1155</lpage>
          <pub-id pub-id-type="doi">10.1016/j.jacc.2013.11.043</pub-id>
        </element-citation>
      </ref>
      <ref id="B9">
        <label>9</label>
        <mixed-citation>9. Min JK, Leipsic J, Pencina MJ, Berman DS, Kocc BZ, van Mieghem C, et al. Diagnostic accuracy of fractional flow reserve from anatomic CT angiography. <italic>JAMA &#8211; J Am Med Assoc</italic> . 2012; 308(12):1237-45. doi:10.1001/2012.jama.11274.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Min</surname>
              <given-names>JK</given-names>
            </name>
            <name>
              <surname>Leipsic</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Pencina</surname>
              <given-names>MJ</given-names>
            </name>
            <name>
              <surname>Berman</surname>
              <given-names>DS</given-names>
            </name>
            <name>
              <surname>Kocc</surname>
              <given-names>BZ</given-names>
            </name>
            <name>
              <surname>van Mieghem</surname>
              <given-names>C</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Diagnostic accuracy of fractional flow reserve from anatomic CT angiography</article-title>
          <source>JAMA &#8211; J Am Med Assoc</source>
          <year>2012</year>
          <volume>308</volume>
          <issue>12</issue>
          <fpage>1237</fpage>
          <lpage>1245</lpage>
          <pub-id pub-id-type="doi">10.1001/2012.jama.11274</pub-id>
        </element-citation>
      </ref>
      <ref id="B10">
        <label>10</label>
        <mixed-citation>10. Koo BK, Erglis A, Doh JH, Daniels DV, Jegere S, Kim HS, et al. Diagnosis of ischemia-causing coronary stenoses by noninvasive fractional flow reserve computed from coronary computed tomographic angiograms: Results from the prospective multicenter Discover-Flow (Diagnosis of Ischemia-Causing Stenoses Obtained Via Noni). <italic>J Am Coll Cardiol</italic> . 2011; 58(19):1989-97. doi:10.1016/j.jacc.2011.06.066.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Koo</surname>
              <given-names>BK</given-names>
            </name>
            <name>
              <surname>Erglis</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Doh</surname>
              <given-names>JH</given-names>
            </name>
            <name>
              <surname>Daniels</surname>
              <given-names>DV</given-names>
            </name>
            <name>
              <surname>Jegere</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Kim</surname>
              <given-names>HS</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Diagnosis of ischemia-causing coronary stenoses by noninvasive fractional flow reserve computed from coronary computed tomographic angiograms: Results from the prospective multicenter Discover-Flow (Diagnosis of Ischemia-Causing Stenoses Obtained Via Noni)</article-title>
          <source>J Am Coll Cardiol</source>
          <year>2011</year>
          <volume>58</volume>
          <issue>19</issue>
          <fpage>1989</fpage>
          <lpage>1997</lpage>
          <pub-id pub-id-type="doi">10.1016/j.jacc.2011.06.066</pub-id>
        </element-citation>
      </ref>
      <ref id="B11">
        <label>11</label>
        <mixed-citation>11. R&#246;ther J, Moshage M, Dey D, Schwemmer C, Trobs M, Blachutzik F, et al. Comparison of invasively measured FFR with FFR derived from coronary CT angiography for detection of lesion-specific ischemia: Results from a PC-based prototype algorithm. <italic>J Cardiovasc Comput Tomogr</italic> . 2018; 12(2):101-7. doi:10.1016/j.jcct.2018.01.012.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>R&#246;ther</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Moshage</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Dey</surname>
              <given-names>D</given-names>
            </name>
            <name>
              <surname>Schwemmer</surname>
              <given-names>C</given-names>
            </name>
            <name>
              <surname>Trobs</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Blachutzik</surname>
              <given-names>F</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Comparison of invasively measured FFR with FFR derived from coronary CT angiography for detection of lesion-specific ischemia: Results from a PC-based prototype algorithm</article-title>
          <source>J Cardiovasc Comput Tomogr</source>
          <year>2018</year>
          <volume>12</volume>
          <issue>2</issue>
          <fpage>101</fpage>
          <lpage>107</lpage>
          <pub-id pub-id-type="doi">10.1016/j.jcct.2018.01.012</pub-id>
        </element-citation>
      </ref>
      <ref id="B12">
        <label>12</label>
        <mixed-citation>12. Abbara S, Blanke P, Maroules CD,Cheezum M, Choi AD Han BK, et al. SCCT guidelines for the performance and acquisition of coronary computed tomographic angiography : A report of the Society of Cardiovascular Computed Tomography Guidelines Committee Endorsed by the North America. Cardiovasc Comput Tomogr.2016;10(6): 435-49.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Abbara</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Blanke</surname>
              <given-names>P</given-names>
            </name>
            <name>
              <surname>Maroules</surname>
              <given-names>CD</given-names>
            </name>
            <name>
              <surname>Cheezum</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Choi</surname>
              <given-names>AD</given-names>
            </name>
            <name>
              <surname>Han</surname>
              <given-names>BK</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>SCCT guidelines for the performance and acquisition of coronary computed tomographic angiography : A report of the Society of Cardiovascular Computed Tomography Guidelines Committee Endorsed by the North America</article-title>
          <source>Cardiovasc Comput Tomogr</source>
          <year>2016</year>
          <volume>10</volume>
          <issue>6</issue>
          <fpage>435</fpage>
          <lpage>449</lpage>
        </element-citation>
      </ref>
      <ref id="B13">
        <label>13</label>
        <mixed-citation>13. Leipsic J, Co-Chair F, Abbara S, Achenbach S, Cury R, Earls JP, et al. SCCT Guidelines SCCT guidelines for the interpretation and reporting of coronary CT angiography: A report of the Society of Cardiovascular Computed Tomography Guidelines Committee.J Cardiovasc Comput Tomogr.2014;8(5):342-58. doi:10.1016/j.jcct.2014.07.003.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Leipsic</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Co-Chair</surname>
              <given-names>F</given-names>
            </name>
            <name>
              <surname>Abbara</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Achenbach</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Cury</surname>
              <given-names>R</given-names>
            </name>
            <name>
              <surname>Earls</surname>
              <given-names>JP</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>SCCT Guidelines SCCT guidelines for the interpretation and reporting of coronary CT angiography: A report of the Society of Cardiovascular Computed Tomography Guidelines Committee</article-title>
          <source>J Cardiovasc Comput Tomogr</source>
          <year>2014</year>
          <volume>8</volume>
          <issue>5</issue>
          <fpage>342</fpage>
          <lpage>358</lpage>
          <pub-id pub-id-type="doi">10.1016/j.jcct.2014.07.003</pub-id>
        </element-citation>
      </ref>
      <ref id="B14">
        <label>14</label>
        <mixed-citation>14. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. <italic>Biometrics</italic> . 1988; 44(3):837-45.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>DeLong</surname>
              <given-names>ER</given-names>
            </name>
            <name>
              <surname>DeLong</surname>
              <given-names>DM</given-names>
            </name>
            <name>
              <surname>Clarke-Pearson</surname>
              <given-names>DL</given-names>
            </name>
          </person-group>
          <article-title>Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach</article-title>
          <source>Biometrics</source>
          <year>1988</year>
          <volume>44</volume>
          <issue>3</issue>
          <fpage>837</fpage>
          <lpage>845</lpage>
        </element-citation>
      </ref>
      <ref id="B15">
        <label>15</label>
        <mixed-citation>15. WJ Youden. Index for rating diagnostic tests. <italic>Cancer</italic> . 1950: 32-5.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Youden</surname>
              <given-names>WJ</given-names>
            </name>
          </person-group>
          <article-title>Index for rating diagnostic tests</article-title>
          <source>Cancer</source>
          <volume>1950</volume>
          <fpage>32</fpage>
          <lpage>35</lpage>
        </element-citation>
      </ref>
      <ref id="B16">
        <label>16</label>
        <mixed-citation>16. Renker M, Schoepf UJ, Wang R, Meinel FG, Rier JD, Bayer RRnd, et al. Comparison of diagnostic value of a novel noninvasive coronary computed tomography angiography method versus standard coronary angiography for assessing fractional flow reserve. <italic>Am J Cardiol</italic> . 2014; 114(9):1303-8. doi:10.1016/j.amjcard.2014.07.064.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Renker</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Schoepf</surname>
              <given-names>UJ</given-names>
            </name>
            <name>
              <surname>Wang</surname>
              <given-names>R</given-names>
            </name>
            <name>
              <surname>Meinel</surname>
              <given-names>FG</given-names>
            </name>
            <name>
              <surname>Rier</surname>
              <given-names>JD</given-names>
            </name>
            <name>
              <surname>RRnd</surname>
              <given-names>Bayer</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Comparison of diagnostic value of a novel noninvasive coronary computed tomography angiography method versus standard coronary angiography for assessing fractional flow reserve</article-title>
          <source>Am J Cardiol</source>
          <year>2014</year>
          <volume>114</volume>
          <issue>9</issue>
          <fpage>1303</fpage>
          <lpage>1308</lpage>
          <pub-id pub-id-type="doi">10.1016/j.amjcard.2014.07.064</pub-id>
        </element-citation>
      </ref>
      <ref id="B17">
        <label>17</label>
        <mixed-citation>17. Coenen A, Lubbers MM, Kurata A, Kono A, Dedic A, Chelu RG, et al. Fractional flow reserve computed from noninvasive CT angiography data: diagnostic performance of an on-site clinician-operated computational fluid dynamics algorithm. <italic>Radiology</italic> . 2015; 274(3):674-83. doi:10.1148/radiol.14140992.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Coenen</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Lubbers</surname>
              <given-names>MM</given-names>
            </name>
            <name>
              <surname>Kurata</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Kono</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Dedic</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Chelu</surname>
              <given-names>RG</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Fractional flow reserve computed from noninvasive CT angiography data: diagnostic performance of an on-site clinician-operated computational fluid dynamics algorithm</article-title>
          <source>Radiology</source>
          <year>2015</year>
          <volume>274</volume>
          <issue>3</issue>
          <fpage>674</fpage>
          <lpage>683</lpage>
          <pub-id pub-id-type="doi">10.1148/radiol.14140992</pub-id>
        </element-citation>
      </ref>
      <ref id="B18">
        <label>18</label>
        <mixed-citation>18. Baumann S, Wang R, Schoepf UJ, Steinberg DH, Spearman JV, Bayer RR, et al. et al. Coronary CT angiography-derived fractional flow reserve correlated with invasive fractional flow reserve measurements &#8211; initial experience with a novel physician-driven algorithm. <italic>Eur Radiol</italic> . 2015;25(4):1201-7. doi:10.1007/s00330-014-3482-5.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Baumann</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Wang</surname>
              <given-names>R</given-names>
            </name>
            <name>
              <surname>Schoepf</surname>
              <given-names>UJ</given-names>
            </name>
            <name>
              <surname>Steinberg</surname>
              <given-names>DH</given-names>
            </name>
            <name>
              <surname>Spearman</surname>
              <given-names>JV</given-names>
            </name>
            <name>
              <surname>Bayer</surname>
              <given-names>RR</given-names>
            </name>
            <etal>et al. et al</etal>
          </person-group>
          <article-title>Coronary CT angiography-derived fractional flow reserve correlated with invasive fractional flow reserve measurements &#8211; initial experience with a novel physician-driven algorithm</article-title>
          <source>Eur Radiol</source>
          <year>2015</year>
          <volume>25</volume>
          <issue>4</issue>
          <fpage>1201</fpage>
          <lpage>1207</lpage>
          <pub-id pub-id-type="doi">10.1007/s00330-014-3482-5</pub-id>
        </element-citation>
      </ref>
      <ref id="B19">
        <label>19</label>
        <mixed-citation>19. Wang R, Renker M, Schoepf UJ, Wichmann JL, Fuller S, Rier J, et al. Diagnostic value of quantitative stenosis predictors with coronary CT angiography compared to invasive fractional flow reserve. <italic>Eur J Radiol</italic> . 2015; 84(8):1509-15. doi:10.1016/j.ejrad.2015.05.010.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Wang</surname>
              <given-names>R</given-names>
            </name>
            <name>
              <surname>Renker</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Schoepf</surname>
              <given-names>UJ</given-names>
            </name>
            <name>
              <surname>Wichmann</surname>
              <given-names>JL</given-names>
            </name>
            <name>
              <surname>Fuller</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Rier</surname>
              <given-names>J</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Diagnostic value of quantitative stenosis predictors with coronary CT angiography compared to invasive fractional flow reserve</article-title>
          <source>Eur J Radiol</source>
          <year>2015</year>
          <volume>84</volume>
          <issue>8</issue>
          <fpage>1509</fpage>
          <lpage>1515</lpage>
          <pub-id pub-id-type="doi">10.1016/j.ejrad.2015.05.010</pub-id>
        </element-citation>
      </ref>
      <ref id="B20">
        <label>20</label>
        <mixed-citation>20. Hadamitzky M, Distler R, Meyer T, Hein F, Kastrati A, Martinoff S, et al. Prognostic value of coronary computed tomographic angiography in comparison with calcium scoring and clinical risk scores. <italic>Circ Cardiovasc Imaging</italic> . 2011; 4(1):16-23. doi:10.1161/CIRCIMAGING.110.955351.</mixed-citation>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Hadamitzky</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Distler</surname>
              <given-names>R</given-names>
            </name>
            <name>
              <surname>Meyer</surname>
              <given-names>T</given-names>
            </name>
            <name>
              <surname>Hein</surname>
              <given-names>F</given-names>
            </name>
            <name>
              <surname>Kastrati</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Martinoff</surname>
              <given-names>S</given-names>
            </name>
            <etal>et al</etal>
          </person-group>
          <article-title>Prognostic value of coronary computed tomographic angiography in comparison with calcium scoring and clinical risk scores</article-title>
          <source>Circ Cardiovasc Imaging</source>
          <year>2011</year>
          <volume>4</volume>
          <issue>1</issue>
          <fpage>16</fpage>
          <lpage>23</lpage>
          <pub-id pub-id-type="doi">10.1161/CIRCIMAGING.110.955351</pub-id>
        </element-citation>
      </ref>
    </ref-list>
    <fn-group>
      <fn fn-type="other">
        <p>Vincula&#231;&#227;o acad&#234;mica</p>
        <p>N&#227;o h&#225; vincula&#231;&#227;o deste estudo a programas de p&#243;s-gradua&#231;&#227;o.</p>
      </fn>
    </fn-group>
    <fn-group>
      <fn fn-type="supported-by">
        <p><bold>Fontes de financiamento</bold>:  O presente estudo n&#227;o teve fontes de financiamento externas.</p>
      </fn>
    </fn-group>
  </back>
  <sub-article article-type="translation" id="TRen" xml:lang="en">
    <front-stub>
      <article-categories>
        <subj-group subj-group-type="heading">
          <subject>Original Article</subject>
        </subj-group>
      </article-categories>
      <title-group>
        <article-title>Diagnostic Performance of a Machine Learning-Based CT-Derived FFR in Detecting Flow-Limiting Stenosis</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0001-9652-2927</contrib-id>
          <name>
            <surname>Morais</surname>
            <given-names>Thamara Carvalho</given-names>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Assun&#231;&#227;o-Jr</surname>
            <given-names>Antonildes Nascimento</given-names>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-3053-3583</contrib-id>
          <name>
            <surname>Dantas</surname>
            <given-names>Roberto Nery</given-names>
            <suffix>J&#250;nior</suffix>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Silva</surname>
            <given-names>Carla Franco Grego da</given-names>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Paula</surname>
            <given-names>Caroline Bastida de</given-names>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0001-5547-6821</contrib-id>
          <name>
            <surname>Torres</surname>
            <given-names>Roberto Almeida</given-names>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-7232-4552</contrib-id>
          <name>
            <surname>Magalh&#227;es</surname>
            <given-names>Tiago Augusto</given-names>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
          <xref ref-type="aff" rid="aff3001">
            <sup>3</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Nomura</surname>
            <given-names>C&#233;sar Higa</given-names>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>&#193;vila</surname>
            <given-names>Luiz Francisco Rodrigues de</given-names>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Parga</surname>
            <given-names>Jos&#233; Rodrigues</given-names>
            <suffix>Filho</suffix>
          </name>
          <xref ref-type="aff" rid="aff1001">
            <sup>1</sup>
          </xref>
          <xref ref-type="aff" rid="aff2001">
            <sup>2</sup>
          </xref>
        </contrib>
      </contrib-group>
      <aff id="aff1001">
        <label>1</label>
        <country country="BR">Brazil</country>
        <institution content-type="original">Hospital S&#237;rio-liban&#234;s, S&#227;o Paulo, SP - Brazil</institution>
      </aff>
      <aff id="aff2001">
        <label>2</label>
        <country country="BR">Brazil</country>
        <institution content-type="original">Universidade de S&#227;o Paulo Faculdade de Medicina - CDI - InCor/HCFMUSP - Departamento de Imagem Cardiovascular, S&#227;o Paulo, SP - Brazil</institution>
      </aff>
      <aff id="aff3001">
        <label>3</label>
        <country country="BR">Brazil</country>
        <institution content-type="original">Complexo Hospital de Cl&#237;nicas da Universidade Federal do Paran&#225; (CHC-UFPR) - Cardiovascular CT/MR, Curitiba, PR - Brazil</institution>
      </aff>
      <author-notes>
        <corresp id="c01001">Mailing Address: Thamara Carvalho Morais &#8226; Hospital S&#237;rio-liban&#234;s - Rua Adma Jafet, 91. Postal Code 01308-050, S&#227;o Paulo, SP - Brazil E-mail: thamaramorais@hotmail.com</corresp>
        <fn fn-type="con">
          <p>Author contributions</p>
          <p>Conception and design of the research: Magalh&#227;es TA, Nomura CH, &#193;vila LFR, Parga Filho JR; Acquisition of data: Morais TC, Silva CFG, Paula CB, Torres RA, Magalh&#227;es TA; Analysis and interpretation of the data: Assun&#231;&#227;o-Jr AN, Dantas J&#250;nior RN, Parga Filho JR; Statistical analysis: Assun&#231;&#227;o-Jr AN, Dantas J&#250;nior RN; Writing of the manuscript: Morais TC, Assun&#231;&#227;o-Jr AN, Dantas J&#250;nior RN, Magalh&#227;es TA, Parga Filho JR; Critical revision of the manuscript for intellectual content: Morais TC, Assun&#231;&#227;o-Jr AN, Dantas J&#250;nior RN, Nomura CH, &#193;vila LFR, Parga Filho JR.</p>
        </fn>
        <fn fn-type="conflict">
          <p>Potential Conflict of Interest</p>
          <p>No potential conflict of interest relevant to this article was reported.</p>
        </fn>
      </author-notes>
      <abstract>
        <title>Abstract</title>
        <sec>
          <title>Background</title>
          <p>The non-invasive quantification of the fractional flow reserve (FFRCT) using a more recent version of an artificial intelligence-based software and latest generation CT scanner (384 slices) may show high performance to detect coronary ischemia.</p>
        </sec>
        <sec>
          <title>Objectives</title>
          <p>To evaluate the diagnostic performance of FFRCT for the detection of significant coronary artery disease (CAD) in contrast to invasive FFR (iFFR) using previous generation CT scanners (128 and 256- detector rows).</p>
        </sec>
        <sec>
          <title>Methods</title>
          <p>Retrospective study with patients referred to coronary artery CT angiography (CTA) and catheterization (iFFR) procedures. Siemens Somatom Definition Flash (256-detector rows) and AS+ (128-detector rows) CT scanners were used to acquire the images. The FFRCT and the minimal lumen area (MLA) were evaluated using a dedicated software (cFFR version 3.0.0, Siemens Healthineers, Forchheim, Germany). Obstructive CAD was defined as CTA lumen reduction &#8805; 50%, and flow-limiting stenosis as iFFR &#8804;0.8. All reported P values are two-tailed, and when &lt;0.05, they were considered statistically significant.</p>
        </sec>
        <sec>
          <title>Results</title>
          <p>Ninety-three consecutive patients (152 vessels) were included. There was good agreement between FFRCT and iFFR, with minimal FFRCT overestimation (bias: -0.02; limits of agreement:0.14-0.09). Different CT scanners did not modify the association between FFRCT and FFRi (p for interaction=0.73). The performance of FFRCT was significantly superior compared to the visual classification of coronary stenosis (AUC 0.93vs.0.61, p&lt;0.001) and to MLA (AUC 0.93vs.0.75, p&lt;0.001), reducing the number of false-positive cases. The optimal cut-off point for FFRCT using a Youden index was 0.85 (87% Sensitivity, 86% Specificity, 73% PPV, 94% NPV), with a reduction of false-positives.</p>
        </sec>
        <sec>
          <title>Conclusion</title>
          <p>Machine learning-based FFRCT using previous generation CT scanners (128 and 256-detector rows) shows good diagnostic performance for the detection of CAD, and can be used to reduce the number of invasive procedures.</p>
        </sec>
      </abstract>
      <kwd-group xml:lang="en">
        <kwd>Myocardial Fractional Flow Reserve</kwd>
        <kwd>Coronary Artery Disease</kwd>
        <kwd>Computed Tomography</kwd>
        <kwd>Myocardial Ischemic</kwd>
        <kwd>Machine Learning</kwd>
      </kwd-group>
    </front-stub>
    <body>
      <sec sec-type="intro">
        <title>Introduction</title>
        <p>According to the most recent clinical guidelines,<sup><xref ref-type="bibr" rid="B1">1</xref> - <xref ref-type="bibr" rid="B3">3</xref></sup> management of chronic and symptomatic coronary artery disease (CAD) may be guided by additional tests for either anatomical (extent, severity, morphology) or functional (ventricular function, presence/extent of ischemia) assessment, with evidence suggesting the superiority of the functional over the anatomical approach in some clinical scenarios.<sup><xref ref-type="bibr" rid="B4">4</xref> - <xref ref-type="bibr" rid="B6">6</xref></sup></p>
        <p>For this purpose, especially in patients with intermediate pretest probability of obstructive CAD, the coronary computed tomography angiography (CTA) stands out among the various existing non-invasive tests as a robust method to rule out obstructive CAD, given its high negative predictive value.<sup><xref ref-type="bibr" rid="B7">7</xref></sup> Particularly in moderate stenosis (50-69%), the non-invasive measurement of myocardial fractional flow reserve (FFR<sub>CT</sub>) may help to correctly distinguish which of these are associated with ischemia<sup><xref ref-type="bibr" rid="B8">8</xref></sup> . Recent studies have shown that the CTA is an accurate test to identify ischemia through non-invasive quantification of fractional flow reserve (FFR<sub>CT</sub>) when compared to the gold standard, the invasive FFR by cardiac catheterization (iFFR).<sup><xref ref-type="bibr" rid="B8">8</xref> - <xref ref-type="bibr" rid="B10">10</xref></sup></p>
        <p>The use of FFR<sub>CT</sub>in clinical practice has been mostly restricted by its low availability, especially due to the need for specific software that would run only on supercomputers at large international centers, substantially increasing the cost and delaying the whole diagnostic process.<sup><xref ref-type="bibr" rid="B8">8</xref></sup> More recently, a non-commercial software prototype (available for standard configuration personal computers) that uses artificial intelligence tools &#8211; convolutional neural network (deep learning) &#8211; to evaluate FFR<sub>CT</sub>was tested by Rother et al.<sup><xref ref-type="bibr" rid="B11">11</xref></sup> When compared to iFFR, the FFR<sub>CT</sub>calculated by this software has shown high accuracy to detect ischemia, with a significant reduction in the calculation time when compared to existing models that use supercomputers.<sup><xref ref-type="bibr" rid="B8">8</xref></sup> However, it should be noted that their study used only images acquired by a latest generation CT scanner (Siemens Somatom Force &#8211; 384 slices). Since this software is capable of calculating the FFR<sub>CT</sub>from images acquired using CT scanners that employ different technologies, our aim was to use it to investigate the diagnostic accuracy of the FFR<sub>CT</sub>using previous generation CT scanners, compared to iFFR, with imaging quality that can potentially affect the results of the algorithm used in the software. This study also compared the diagnostic accuracy of FFR<sub>CT</sub>to the isolated anatomical assessment by CTA.</p>
      </sec>
      <sec sec-type="methods">
        <title>Methods</title>
        <sec>
          <title>Study Population</title>
          <p>Retrospectively, this study included symptomatic patients that were referred to CTA for investigating significant CAD and, after the test findings, upon clinical decision, were referred to cardiac catheterization (within less than 30 days) and underwent iFFR analysis at S&#237;rio-Liban&#234;s Hospital, S&#227;o Paulo, Brazil, between January 2014 and February 2018. There was a total of 17 exclusions: 14 owing to factors that limited the FFR<sub>CT</sub>calculation, as described by the tool manufacturer (8 due to left main coronary artery (LMCA), ostia, or bifurcation lesions, 6 owing to the presence of stent); and 3 due to bad quality image as a result of excessive calcification and significant motion artifacts. During image post-processing, no other patient was excluded as a result of the software technical inability to measure FFR<sub>CT</sub>. It is noteworthy that the manufacturer&#8217;s recommendations regarding LMCA, ostia, or bifurcation lesions have also been followed by other authors,<sup><xref ref-type="bibr" rid="B11">11</xref></sup> and they seem to be related to the limited recognition of the anatomical borders in these scenarios. This study was approved by the S&#237;rio-Liban&#234;s Hospital Research Ethics Committee.</p>
        </sec>
        <sec>
          <title>Acquisition of the CTA images</title>
          <p>The images were obtained using a 256-slice Siemens Somatom Definition Flash CT scanner (temporal resolution - TR - of 75 ms; spatial resolution - SR - of 0.30 mm) and a 128-slice Somatom Definition AS CT scanner (TR of 150 ms; SR of 0.30mm) (Siemens Healthineers, Forchheim, Germany). The patients were prepared according to current guideline recommendations, including four-hour fasting, using an IV peripheral venous access catheter (18 Gauge), preferably on the right antecubital vein, and continuous electrocardiographic monitoring.<sup><xref ref-type="bibr" rid="B12">12</xref></sup> Whenever necessary, a beta-blocker was administered (50-100mg oral metoprolol tartrate 1 hour before the test and/or 5-20mg intravenously a few minutes before image acquisition) to obtain heart rate (HR) control (aim 55-60 bpm). All patients also received sublingual nitrate (Isordil 2.5mg) a few minutes before acquisition, except in cases of symptomatic hypotension or use of phosphodiesterase type 5 inhibitors (according to each drug action onset).</p>
          <p>Image acquisition was planned after bolus testing to calculate the contrast peak time in the aorta, with a 10-15 mL volume, followed by 30-50 mL of saline solution at 4.5-5.5 mL/s. The images were acquired using Flash mode (in the Definition Flash CT scanner) or retrospectively (in both CT scanners) with electrocardiographic gating in late diastole (55-75% of RR), tube voltage of 100-120 kVp (adjusted to the individual BMI), rotation time 0.28 (Flash) / 0.33 (AS+) seconds, 160-320 mAs and slice thickness of 0.6/0.3 mm. The Optiray 350 (Ioversol 350 mg/mL-Mallinckrodt-USA) iodinated contrast media infusion used the same parameters of the bolus testing (60-90ml).</p>
        </sec>
        <sec>
          <title>Analysis of the CTA and FFR CT images</title>
          <p>CTA image analysis was carried out using the Syngo.via imaging software (Siemens Healthineers, Forchheim, Germany). After choosing the images with best technical quality, the coronary tree evaluation was performed by three-dimensional, curved, multiplanar reformation (Vessel Probe), with quantification of stenosis degree and prevalent plaque composition, when present (non-calcified, calcified, and mixed). Luminal reduction quantification was carried out according to the Society of Cardiovascular Computed Tomography recommendations:<sup><xref ref-type="bibr" rid="B13">13</xref></sup> normal, minimal (&lt;25%), mild (25-49%), moderate (50-69%), severe (70-99%) and occlusion (100%). CTA-assessed stenosis was also classified as obstructive (&#8805;50%) and non-obstructive (&lt;50%).</p>
          <p>Post-processing FFR<sub>CT</sub>was performed in the same series in which visual anatomical analysis was carried out (as described above), using the Frontier platform and the non-commercial prototype of the cFFR software, version 3.0 (Siemens Healthineers, Forchheim, Germany), by a physician experienced in cardiovascular imaging (&gt;4 years).</p>
          <p>To calculate coronary artery FFR<sub>CT</sub>, the first step was to perform the automatic detection of the centerline and lumen contours, which were reviewed and corrected by the specialist, when necessary. Next, the upper and lower borders of all plaques in the vessels that had their iFFR values calculated during invasive catheterization were delimited.</p>
          <p>To precisely identify the site where the iFFR was calculated, the interventional cardiologist located the point of interest in the fluoroscopic images and documented it using anatomical references (coronary branches) and also a coronary segmentation model suggested by the Society of Cardiovascular Computed Tomography (SCCT). The CTA expert used this documentation to select the same anatomical location of the plaques that had the FFR<sub>CT</sub>calculated.</p>
          <p>The cFFR software calculates FFR values only for vessels &#8805;1.5mm in diameter and automatically calculates the value of the minimal lumen area (MLA) of each delimited plaque. The steps for FFR<sub>CT</sub>calculation are shown in <xref ref-type="fig" rid="f01001">Figure 1</xref> . This software was developed with new artificial intelligence tools, using machine-learning techniques (deep learning). All of them were installed in a standard computer used for radiology reports. The total post-processing time of all steps was around 10 minutes.</p>
          <p>
            <fig id="f01001">
              <label>Figure 1</label>
              <caption>
                <title>&#8211; Steps to calculate FFRCT using the cFFR software. A) Visual detection of the coronary lesion. Definition of centerline B) and lumen contours C) using the cFFR software D) definition by the operator of the lesion borders and the point of higher lumen reduction. E) FFR result at the point with higher lumen reduction shown in the coronary tree (after determining the centerline and lumen contours in the 3 main coronary arteries: LAD, LCx, and RCA).</title>
              </caption>
              <graphic xlink:href="1678-4170-abc-116-06-1091-90329-gf01-en.tif"/></fig>
          </p>
        </sec>
        <sec>
          <title>Cardiac catheterization and iFFR analysis</title>
          <p>Cardiac catheterization was performed through radial or femoral access, using 6 or 7 French (F) diagnostic catheters. To calculate the iFFR, the degree of stenosis was visually evaluated by the interventional cardiologist using at least two orthogonal projections. Intracoronary nitroglycerin (0.2 mg) was injected in all patients before the angiograms. A pressure monitoring guidewire was placed distal to the index lesion, and the mean pressures were recorded when they were stable. Intracoronary adenosine was manually injected through the guide catheter, through an 80 &#956;g bolus injection (left coronary artery) or 40 &#956;g bolus (right coronary artery) into 10 mL of saline solution. After the administration, the lowest stable FFR value during the hyperemic steady state was recorded. This value corresponds to the ratio between the mean coronary pressure distal to the stenosis and the mean aortic pressure at the time of the pharmacologically-induced hyperemia.</p>
          <p>The exact position of the iFFR measurement sensor was documented by the interventional cardiologist on the report, and this documentation was used by the CTA expert to measure FFR<sub>CT</sub>at the same anatomical location.</p>
        </sec>
        <sec>
          <title>Statistical Analysis</title>
          <p>Descriptive analyses were expressed as frequency (percentage) for categorical variables, and as mean &#177; standard deviation for continuous variables. The distribution of the continuous variables was visually assessed using QQ plots and checked using the Shapiro-Wilk test. Comparisons between the continuous variables found in the CTA and the catheterization (degree of coronary stenosis, FFR) were made using Student&#8217;s t-test for paired samples. Likewise, the correlation between these variables was made using Pearson&#8217;s correlation.</p>
          <p>The agreement between FFR<sub>CT</sub>and iFFR was determined by Bland-Altman analysis. Assuming iFFR &#8804;0.8 as the gold standard for the presence of ischemia, the diagnostic performance of FFR<sub>CT</sub>and other CTA anatomical parameters were evaluated by calculating sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). In addition, the area under the ROC curve (AUC) to detect coronary lesions associated with ischemia was calculated, and comparisons between the AUCs were performed according to the method described by DeLong et al.<sup><xref ref-type="bibr" rid="B14">14</xref></sup> The best FFR<sub>CT</sub>cut-off point for the detection of ischemia (iFFR &#8804;0.8) was calculated using the Youden index, which corresponds to the one with the highest value in the equation (sensitivity [Sens] + specificity [Spec] - 1).<sup><xref ref-type="bibr" rid="B15">15</xref></sup> It is noteworthy that the degree of stenosis was explored as a continuous and categorical variable (obstructive CAD, &gt;50%, or not). The choice to include the categorical form in the model was based on the fact that it is a clinical threshold for further decision-making.</p>
          <p>Considering the potential correlation between multiple vessels in the same individual, the generalized estimating equations (GEE) method with exchangeable correlation structure was used at a per-vessel level. Statistical analyses were carried out using the R software (R Foundation for Statistical Computing, Vienna, Austria). All reported <italic>P</italic> values are two-tailed, and when &lt;0.05, they were considered statistically significant.</p>
        </sec>
      </sec>
      <sec sec-type="results">
        <title>Results</title>
        <sec>
          <title>Characteristics of the patients and plaques</title>
          <p>Ninety-three patients were included in the study, with a total of 152 vessels. Fifty patients (54%) underwent CTA in the Flash CT scanner (256-detector rows), and 43 (46%) in the AS+ CT scanner (128-detector rows). The average HR during image acquisition was 58 &#177; 8 bpm.</p>
          <p>Seventy-four patients (80%) showed obstructive CAD (stenosis &gt;50%) in the CTA, 48 with moderate stenosis (50-69%) and 26 with severe stenosis (&gt;70%). In the per-vessel analysis, plaques were more often the mixed type (70%), and most commonly located in the left anterior descending coronary artery (LAD) (73%) and had a mean MLA of 3.2 &#177; 1.6 mm<sup>2</sup>. Clinical and CT characteristics of the patients are shown in <xref ref-type="table" rid="t1001">Tables 1</xref> and <xref ref-type="table" rid="t2001">2</xref> , respectively.</p>
          <p>
            <table-wrap id="t1001">
              <label>Table 1</label>
              <caption>
                <title>&#8211; Demographic Data</title>
              </caption>
              <table frame="hsides" rules="groups">
                <colgroup width="50%">
                  <col/>
                  <col/>
                </colgroup>
                <thead>
                  <tr>
                    <th align="left">Variables</th>
                    <th>n = 93</th>
                  </tr>
                </thead>
                <tbody>
                  <tr>
                    <td>Age, years*</td>
                    <td align="center">64 &#177; 11</td>
                  </tr>
                  <tr>
                    <td>Male, n (%)</td>
                    <td align="center">70 (75)</td>
                  </tr>
                  <tr>
                    <td>Hypertension, n (%)</td>
                    <td align="center">54 (58)</td>
                  </tr>
                  <tr>
                    <td>Dyslipidemia, n (%)</td>
                    <td align="center">45 (48)</td>
                  </tr>
                  <tr>
                    <td>Diabetes, n (%)</td>
                    <td align="center">24 (26)</td>
                  </tr>
                  <tr>
                    <td>Smoking, n (%)</td>
                    <td align="center">7 (8)</td>
                  </tr>
                  <tr>
                    <td>BMI, kg/m<sup>2</sup>*</td>
                    <td align="center">28 &#177; 4</td>
                  </tr>
                  <tr>
                    <td>HR, bpm*</td>
                    <td align="center">58 &#177; 8</td>
                  </tr>
                </tbody>
              </table>
              <table-wrap-foot>
                <fn id="TFN1001">
                  <p>
                    <italic>BMI: body mass index; HR: heart rate. *mean &#177; standard deviation.</italic>
                  </p>
                </fn>
              </table-wrap-foot>
            </table-wrap>
          </p>
          <p>
            <table-wrap id="t2001">
              <label>Table 2</label>
              <caption>
                <title>&#8211; Coronary plaque characterization</title>
              </caption>
              <table frame="hsides" rules="groups">
                <colgroup width="50%">
                  <col/>
                  <col/>
                </colgroup>
                <thead>
                  <tr>
                    <th align="left">Per patient</th>
                    <th>n = 93</th>
                  </tr>
                </thead>
                <tbody>
                  <tr>
                    <td>Stenosis &#8805; 50%, n (%)</td>
                    <td align="center">74 (80)</td>
                  </tr>
                  <tr>
                    <td>Stenosis 50-69%, n (%)</td>
                    <td align="center">48 (52)</td>
                  </tr>
                  <tr>
                    <td>Stenosis &#8805; 70%, n (%)</td>
                    <td align="center">26 (28)</td>
                  </tr>
                  <tr>
                    <td>FFR<sub>CT</sub>&#8804; 0.8, n (%)</td>
                    <td align="center">32 (34)</td>
                  </tr>
                  <tr>
                    <td>iFFR &#8804; 0.8, n (%)</td>
                    <td align="center">39 (42)</td>
                  </tr>
                  <tr>
                    <td>
                      <bold>Per vessel</bold>
                    </td>
                    <td align="center">
                      <bold>n = 152</bold>
                    </td>
                  </tr>
                  <tr>
                    <td>Location</td>
                    <td> </td>
                  </tr>
                  <tr>
                    <td>LAD, n (%)</td>
                    <td align="center">111 (73)</td>
                  </tr>
                  <tr>
                    <td>LCx, n (%)</td>
                    <td align="center">26 (17)</td>
                  </tr>
                  <tr>
                    <td>RCA, n (%)</td>
                    <td align="center">16 (10)</td>
                  </tr>
                  <tr>
                    <td>Stenosis &#8805; 50%, n (%)</td>
                    <td align="center">124 (82)</td>
                  </tr>
                  <tr>
                    <td>Stenosis 50-69%, n (%)</td>
                    <td align="center">95 (63)</td>
                  </tr>
                  <tr>
                    <td>Stenosis &#8805; 70%, n (%)</td>
                    <td align="center">29 (19)</td>
                  </tr>
                  <tr>
                    <td>MLA, mm<sup>2</sup>*</td>
                    <td align="center">3.2 &#177; 1.6</td>
                  </tr>
                  <tr>
                    <td>
                      <bold>Morphology</bold>
                    </td>
                    <td> </td>
                  </tr>
                  <tr>
                    <td>Calcified, n (%)</td>
                    <td align="center">16 (10)</td>
                  </tr>
                  <tr>
                    <td>Mixed, n (%)</td>
                    <td align="center">106 (70)</td>
                  </tr>
                  <tr>
                    <td>Non-calcified, n (%)</td>
                    <td align="center">30 (20)</td>
                  </tr>
                  <tr>
                    <td>
                      <bold>FFR<sub>CT</sub>*</bold>
                    </td>
                    <td align="center">
                      <bold>0.88 &#177; 0.08</bold>
                    </td>
                  </tr>
                  <tr>
                    <td>FFR<sub>CT</sub>&#8804; 0.8, n (%)</td>
                    <td align="center">32 (21)</td>
                  </tr>
                  <tr>
                    <td>
                      <bold>iFFR*</bold>
                    </td>
                    <td align="center">
                      <bold>0.86 &#177; 0.08</bold>
                    </td>
                  </tr>
                  <tr>
                    <td>iFFR &#8804; 0.8, n (%)</td>
                    <td align="center">47 (31)</td>
                  </tr>
                </tbody>
              </table>
              <table-wrap-foot>
                <fn id="TFN2001">
                  <p>
                    <italic>CAD: coronary artery disease; iFFR: invasive fractional flow reserve; FFR<sub>CT</sub>: fractional flow reserve derived from computed tomography; MLA: minimal lumen area. *mean &#177; standard deviation.</italic>
                  </p>
                </fn>
              </table-wrap-foot>
            </table-wrap>
          </p>
        </sec>
        <sec>
          <title>Comparison between FFR CT and iFFR</title>
          <p>There was a strong correlation between the FFR<sub>CT</sub>and iFFR values (r = 0.73, p&lt;0.001) ( <xref ref-type="fig" rid="f02001">Figure 2</xref> ). On average, FFR<sub>CT</sub>values were slightly higher than iFFR values (0.88 &#177; 0.08 vs. 0.86 &#177; 0.08, p = 0.02), a systematic error confirmed by the Bland-Altman analysis (bias of -0.02 with a confidence interval of -0.14 to 0.09) ( <xref ref-type="fig" rid="f02001">Figure 2</xref> ). The type of CT scanner used did not change the association between iFFR and FFR<sub>CT</sub>(p-value for interaction of 0.73).</p>
          <p>
            <fig id="f02001">
              <label>Figure 2</label>
              <caption>
                <title>&#8211; Correlation (A) and agreement using Bland-Altman analysis (B) between FFR CT and iFFR (per-vessel analysis): EQUIPMENT: <underline>AS</underline> refers to the 128-detector row CT scanner and <underline>FLASH</underline> to the 256-detector row CT scanner.</title>
              </caption>
              <graphic xlink:href="1678-4170-abc-116-06-1091-90329-gf02-en.tif"/></fig>
          </p>
        </sec>
        <sec>
          <title>Ischemia Detection</title>
          <p>For the identification of flow-limiting obstructive coronary lesions (iFFR &#8804;0.8 as the gold standard), FFR<sub>CT</sub>showed a significantly superior performance compared to the isolated visual classification of coronary obstruction (AUC 0.93 vs. 0.61, p&lt;0.001) and to MLA by CTA (AUC 0.93 vs. 0.75, p&lt;0.001) ( <xref ref-type="fig" rid="f03001">Figure 3</xref> ). The best cut-off point (with fewer false-positive results) for FFR<sub>CT</sub>, defined using the Youden index, to distinguish lesions with from those without ischemia was 0.85, which achieved the following accuracy values: 87% sensitivity, 86% specificity, 73% PPV, and 94% NPV at this cut-off point ( <xref ref-type="fig" rid="f04001">Figure 4</xref> ). These performance metrics using this cut-off point (0.85) were slightly higher when analyzing only the plaques with moderate lumen reduction (50-69%, n=95), with 89% sensitivity, 91% specificity, 74% PPV, 97% NPV. Out of the 152 evaluated lesions, three (2%) were false-positives and 18 (12%) were false-negatives using the traditional cut-off point (FFR<sub>CT</sub>&#8804; 0.80). Using the highest cut-off point (FFR<sub>CT</sub>&lt;0.85), 12 (7%) were false-positives and 9 (6%) false-negatives.</p>
          <p>
            <fig id="f03001">
              <label>Figure 3</label>
              <caption>
                <title>&#8211; FFR CT performance for the diagnosis of flow-limiting obstructive lesions (iFFR&lt;0.8).</title>
              </caption>
              <graphic xlink:href="1678-4170-abc-116-06-1091-90329-gf03-en.tif"/></fig>
          </p>
          <p>
            <fig id="f04001">
              <label>Figure 4</label>
              <caption>
                <title>&#8211; Diagnostic performance of FFR CT &lt;0.85. EQUIPMENT: AS refers to the 128-detector row CT scanner and FLASH to the 256-detecttor row CT scanner.</title>
              </caption>
              <graphic xlink:href="1678-4170-abc-116-06-1091-90329-gf04-en.tif"/></fig>
          </p>
          <p>When evaluating the lumen reduction degree, the plaques with visually moderate reduction (50-69%) showed 86 (56%) false-positive cases in relation to the gold standard (iFFR &#8804;0.8), while the plaques with visually severe lumen reduction (&#8805;70%) showed 23 (15%) false-positive cases, which for the latter represent a 50% higher magnitude when compared to the results of FFR<sub>CT</sub>&lt;0.85 (15% <italic>versus</italic> 7%).</p>
        </sec>
      </sec>
      <sec sec-type="discussion">
        <title>Discussion</title>
        <p>FFR<sub>CT</sub>analysis using a machine learning-based software showed good agreement with iFFR measurements, emphasizing that CTA image post-processing was carried out using standard computers in about 10 minutes. Regarding its diagnostic performance, even using previous-generation CT scanners, FFR<sub>CT</sub>performed better than the isolated anatomical evaluation, both in visual stenosis quantification and in the calculation of MLA, significantly reducing the number of false-positives.</p>
        <p>In agreement with Rother et al.,<sup><xref ref-type="bibr" rid="B11">11</xref></sup> which retrospectively studied a cohort of 71 patients using the same software version used in this study (cFFR version 3.0),<sup><xref ref-type="bibr" rid="B11">11</xref></sup> the FFR<sub>CT</sub>showed considerable agreement with the iFFR measurement, with minimal overestimation. These results are in disagreement with previous versions of this same software (cFFR version 1.4),<sup><xref ref-type="bibr" rid="B16">16</xref> - <xref ref-type="bibr" rid="B19">19</xref></sup> where an underestimation was described, which probably reflects algorithm changes with the software upgrade.</p>
        <p>Although comparable to the three major multicenter studies published to date (DISCOVER-FLOW, Defacto, and NXT),<sup><xref ref-type="bibr" rid="B8">8</xref> - <xref ref-type="bibr" rid="B10">10</xref></sup> it should be noted that the limit of agreement of our study was wider in the Bland-Altman analysis (~0.20), which means lower repeatability of the method, compared to what was observed by Rother et al.<sup><xref ref-type="bibr" rid="B11">11</xref></sup> As our patients had an HR &lt;60 bpm on average, which results in good image quality, we believe that the superior performance of that study can be explained, in part, by the use of a CT scanner with 20% higher spatial resolution (0.3 vs. 0.24mm), in addition to the use of a more recent and robust reconstruction algorithm (ADMIRE). These factors may have led to a better detection of coronary contours (centerline and lumen) by that study, with consequent better results. Another justification that cannot be ruled out would be the broader experience by the observer with the new version of cFFR at that center.</p>
        <p>Regarding the power to distinguish flow-limiting from non-flow-limiting coronary stenosis, the FFR<sub>CT</sub>was superior in comparison to the CTA isolated anatomical evaluation, both qualitatively (visual classification of obstructive CAD), and quantitatively (MLA). Using a cut-off point of 0.85 for the FFR<sub>CT</sub>, the NPV and PPV were comparable to those of other cohorts that used this software.<sup><xref ref-type="bibr" rid="B16">16</xref> - <xref ref-type="bibr" rid="B19">19</xref></sup> In addition, we point out the following aspects: 1) FFR<sub>CT</sub>performance was better in cases with moderate lesions (50-69%); 2) FFR<sub>CT</sub>led to a reduction of over 50% in the number of false-positive cases observed when only the anatomical evaluation of severe CAD (&#8805;70%) was used. These findings are highly relevant in clinical practice, since moderate lesions in CTA are relatively frequent and often these patients are referred to additional tests.<sup><xref ref-type="bibr" rid="B20">20</xref></sup> In fact, the opportunity of global reduction in unnecessary referrals to catheterization may be even greater using this new FFR<sub>CT</sub>tool, since only 42% of our patients had iFFR &lt;0.8.</p>
        <p>Finally, we point out this new software fast image post-processing based on machine-learning technology (deep learning). When using pioneering softwares,<sup><xref ref-type="bibr" rid="B8">8</xref> - <xref ref-type="bibr" rid="B10">10</xref></sup> which use computational fluid dynamic algorithms, the calculation of the FFR<sub>CT</sub>takes from 1 to 4 hours to be processed, and it is carried out by supercomputers located only in specific centers in the United States (whose headquarters is in California), the UK, and Japan. In addition to the high cost, in general, about 24 hours are required to obtain the results, and the DICOM images need to be sent out of the institution environment. Therefore, this new software could have a real impact on clinical practice for the care of patients with CAD.</p>
        <sec>
          <title>Limitations</title>
          <p>This is a retrospective, unicentric study, with a relatively small study population, which predominantly had obstructive CAD. When following the manufacturer&#8217;s recommendations regarding the tool appropriate use, patients with significant stenosis in the left main coronary artery, main coronary ostia, or in bifurcations; chronic arterial occlusions; previous history of revascularization surgery or stent implantation were excluded. Likewise, patients with typical symptoms were not always submitted to invasive functional testing (iFFR) by clinical decision. Therefore, this study should be interpreted while giving due attention to the clinical context of its population (less severe/complex CAD and/or clinical scenarios of greater diagnostic uncertainty).</p>
        </sec>
      </sec>
      <sec sec-type="conclusions">
        <title>Conclusion</title>
        <p>This new version of FFR<sub>CT</sub>, even when using previous-generation CT scanners, showed good diagnostic performance for the detection of flow-limiting obstructive coronary lesions, with a significant reduction in the number of false-positive cases, which can significantly decrease the number of patients who are referred to additional tests. The clinical importance of these findings needs to be validated by studies specifically designed to evaluate clinical outcomes. This software features innovative technology that uses machine learning, which enables greater accessibility, rapid performance and potential cost reduction.</p>
      </sec>
    </body>
    <back>
      <fn-group>
        <fn fn-type="other">
          <p>Study Association</p>
          <p>This study is not associated with any thesis or dissertation work.</p>
        </fn>
      </fn-group>
      <fn-group>
        <fn fn-type="supported-by">
          <p><bold>Sources of Funding</bold>:  There were no external funding sources for this study.</p>
        </fn>
      </fn-group>
    </back>
  </sub-article>
</article>
