You are here

  • Home
  • Online First
  • Prediction of prognosis in patients with tetralogy of Fallot based on deep learning imaging analysis

Article Text

Article menu

other Versions

Download PDFPDF
Congenital heart disease
Original research
Prediction of prognosis in patients with tetralogy of Fallot based on deep learning imaging analysis
  1. Gerhard Paul Diller1,
  2. Stefan Orwat1,
  3. Julius Vahle1,
  4. Ulrike M M Bauer2,3,
  5. Aleksandra Urban3,
  6. Samir Sarikouch4,
  7. Felix Berger5,6,
  8. Philipp Beerbaum7,
  9. Helmut Baumgartner1
  10. German Competence Network for Congenital Heart Defects Investigators
    1. 1Department of Cardiology III - Adult Congenital and Valvular Heart Disease, University Hospital Muenster, Muenster, Germany
    2. 2Competence Network for Congenital Heart Defects, DZHK (German Centre for Cardiovascular Research), Berlin, Germany
    3. 3National Register for Congenital Heart Defects, DZHK (German Centre for Cardiovascular Research), Berlin, Germany
    4. 4Department of Heart- Thoracic- Transplantation- and Vascular Surgery, Hannover Medical School, Hannover, Germany
    5. 5Department of Congenital Heart Disease - Pediatric Cardiology, German Heart Institute Berlin, Augustenburger Platz 1, Berlin, Germany
    6. 6DZHK (German Centre for Cardiovascular Research), Partner Site Berlin, Augustenburger Platz 1, Berlin, Germany
    7. 7Department of Pediatric Cardiology and Pediatric Intensive Care, Hannover Medical School, Hannover Medical School, Hannover, Germany
    1. Correspondence to Professor Gerhard Paul Diller, Adult Congenital and Valvular Heart Disease Center, Department of Cardiology and Angiology, University Hospital Muenster, Muenster 48149, Germany; Gerhard.Diller{at}ukmuenster.de

    Abstract

    Objective To assess the utility of machine learning algorithms for automatically estimating prognosis in patients with repaired tetralogy of Fallot (ToF) using cardiac magnetic resonance (CMR).

    Methods We included 372 patients with ToF who had undergone CMR imaging as part of a nationwide prospective study. Cine loops were retrieved and subjected to automatic deep learning (DL)-based image analysis, trained on independent, local CMR data, to derive measures of cardiac dimensions and function. This information was combined with established clinical parameters and ECG markers of prognosis.

    Results Over a median follow-up period of 10 years, 23 patients experienced an endpoint of death/aborted cardiac arrest or documented ventricular tachycardia (defined as >3 documented consecutive ventricular beats). On univariate Cox analysis, various DL parameters, including right atrial median area (HR 1.11/cm², p=0.003) and right ventricular long-axis strain (HR 0.80/%, p=0.009) emerged as significant predictors of outcome. DL parameters were related to adverse outcome independently of left and right ventricular ejection fraction and peak oxygen uptake (p<0.05 for all). A composite score of enlarged right atrial area and depressed right ventricular longitudinal function identified a ToF subgroup at significantly increased risk of adverse outcome (HR 2.1/unit, p=0.007).

    Conclusions We present data on the utility of machine learning algorithms trained on external imaging datasets to automatically estimate prognosis in patients with ToF. Due to the automated analysis process these two-dimensional-based algorithms may serve as surrogates for labour-intensive manually attained imaging parameters in patients with ToF.

    • congenital heart disease
    • tetralogy of Fallot
    • cardiac magnetic resonance (CMR) imaging
    • advanced cardiac imaging

    https://doi.org/10.1136/heartjnl-2019-315962

    Statistics from Altmetric.com

      Request Permissions

      If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

      View Full Text

      Footnotes

      • GPD and SO contributed equally.

      • Collaborators German Competence Network for Congenital Heart Defects Investigators: Gunter Kerst, Jaime F. Vazquez-Jimenez, Dimitrios Gkalpakiotis, Andrea Schedifka, Gernot Buheitel, Joachim Streble, Kai Thorsten Laser, Eugen Sandica, Burkhard Trusen, Felix Berger, Oliver Miera, Stanislav Ovroutski, Björn Peters, Katharina Schmitt, Stephan Schubert, Joachim Photiadis, Felix Berger, Bernd Opgen-Rhein, Katja Weiss, Christoph Berns, Carl-Christian Blumenthal-Barby, Thomas Boeckel, Guido Haverkämper, Andreas Kästner, Heike Koch, Christian Köpcke, Frank Streichan, Jens Timme, Birgit Franzbach, Gabriela Senft, Frank Beyer, Klaus Winter, Johannes Breuer, Bahman Esmailzadeh, Martin Schneider, Boulos Asfour, Jens Bahlmann, Eberhard Griese, Trong Phi Lê, Joachim Hebe, Jan-Hendrik Nürnberg, Annette Magsaam, Ronald Müller, Ludger Potthoff, Renate Voigt, Tim Krüger, Hubert Gerleve, Ulrich Kleideiter, Dirk Schneider-Kulla, Jürgen Krülls-Münch, Elmo Feil, Thomas Menke, Martin Lehn, Antje Heilmann, Helge Tomczak, Otto N. Krogmann, Gleb Tarusinov, Michael Scheid, Ertan Mayatepek, Frank Pillekamp, Artur Lichtenberg, Christiane Terpeluk, Bruno Kolterer, Sven Dittrich, Ulrike Gundlach, Robert Cesnjevar, Ulrich Neudorf, Geert Morf, Anoosh Esmaeili, Stephan Backhoff, Brigitte Stiller, Friedhelm Beyersdorf, Johannes Kroll, Nicole Häffner, Jannos Siaplaouras, Antje Masri-Zada, Christian Jux, Andreas Böning, Hakan Akintürk, Thomas Paul, Matthias Sigler, Theodor Tirilomis, Gabriele Schürer, Johannes Hartmann, Ralph Grabitz, Uta Liebaug, Claudius Rotzsch, Rainer Kozlik-Feldmann, Jörg Sachweh, Arlindo Riso, Stefan Renz, Andreas Schemm, Bernd Friedrich, Otmar Schlobohm, Philipp Beerbaum, Dietmar Böthig, Alexander Horke, Johann Bauersachs, Mechthild Westhoff-Bleck, Matthias Gorenflo, Matthias Karck, Tsvetomir Loukanov, Hermann Schrüfer, Martin Wilken, Hashim Abdul-Khaliq, Tanja Rädle-Hurst, Axel Rentzsch, Hans-Joachim Schäfers, Hagen Reichert, Thomas Kriebel, Arnulf Boysen, Anselm Uebing, Joachim Thomas Cremer, Jens Scheewe, Regina Buchholz-Berdau, Peter Möller, Wolfgang Ram, Konrad Brockmeier, Gerardus B. W. E. Bennink, Alex Gillor, Tim Niehues, Peter Terhoeven, Steffen Leidig, Ingo Dähnert, Peter Kinzel, Martin Kostelka, Liane Kändler, Martin Bethge, Stefan Köster, Christoph Schröder, Jens Karstedt, Uwe Seitz, Christoph Kampmann, Christian-Friedrich Vahl, Frank Stahl, Mojtaba Abedini, Joachim Müller-Scholden, Peter Ewert, Alfred Hager, Harald Kaemmerer, Nicole Nagdyman, Jörg Schoetzau, Oktay Tutarel, Rüdiger Lange, Jürgen Hörer, Nikolaus A. Haas, Lale Rosenthal, Michael Hauser, Alexander Roithmaier, Hans-Gerd Kehl, Edward Malec, Helmut Baumgartner, Gerhard Diller, Roswitha Bahle, Gerald Hofner, Stefan Zink, Roland Reif, Helmut Singer, Christoph Parlasca, Matthias W. Freund, Michael Schumacher, Oliver Dewald, Christine Darrelmann, Reinald Motz, Olaf Willmann, Norbert Schmiedl, Peter Quick, Dirk Hillebrand, Stephan Michele Eiselt, Torsten Nekarda, Michael Eberhard, Georg Baier, Frank Uhlemann, Ioannis Tzanavaros, Alexander Beyer, Gudrun Binz, Steffen Hess, Thomas Teufel, Ronald-Peter Handke, Michael Hofbeck, Renate Kaulitz, Ludger Sieverding, Christian Schlensak, Christian Apitz, Michael Kaestner, Christoph Kupferschmid, Jürgen Holtvogt, Carl-Friedrich Wippermann, Andreas Heusch, Johannes Wirbelauer, Wolfgang Brosi.

      • Contributors GPD, SO and JV collected the data and performed the deep learning-based analysis. UMMB and AU collected the clinical data and reviewed the manuscript. SS, FB, PB and HB were involved in the study design and critical review of the manuscript. GPD wrote the manuscript and performed the statistical analysis.

      • Funding This study was supported by a research grant from the EMAH Stiftung Karla Voellm, Krefeld, Germany, and by the German Competence Network for Congenital Heart Defects (funded by the German Federal Ministry of Education and Research, BMBF, FKZ 01G10210, 01GI0601 until 2014 and the DZHK, German Centre for Cardiovascular Research, as of 2015).

      • Competing interests None declared.

      • Patient and public involvement Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.

      • Patient consent for publication Not required.

      • Ethics approval The study was registered with the National Registry for Congenital Heart Disease, and approval of the study protocol was obtained from the appropriate ethics committee. Patients were enrolled in the national registry as part of recruitment and gave appropriate informed consent before inclusion.

      • Provenance and peer review Not commissioned; externally peer reviewed.

      • Data availability statement Data are available upon reasonable request. Data is available for academic institutions from the German National Registry for Congenital Heart Disease subject to submission of a reasearch protocol and approval by the study board of the registry.

      Linked Articles