2023 Fiscal Year Final Research Report

Development of a new sudden cardiac death prediction support technology using a 12-lead ECG generation model

Research Project

PDF

Project/Area Number	21K08140
Research Category	Grant-in-Aid for Scientific Research (C)
Allocation Type	Multi-year Fund
Section	一般
Review Section	Basic Section 53020:Cardiology-related
Research Institution	Fujita Health University
Principal Investigator	Watanabe Eiichi 藤田医科大学, 医学部, 教授 (80343656)
Co-Investigator(Kenkyū-buntansha)	寺本篤司名城大学, 情報工学部, 教授 (00513780) 祖父江嘉洋藤田医科大学, 医学部, 准教授 (20724793)
Project Period (FY)	2021-04-01 – 2024-03-31
Keywords	心臓突然死 / 心電図 / 人工知能 / 敵対的生成ネットワーク / リスク層別化
Outline of Final Research Achievements	Many sudden cardiac deaths occur due to hemodynamic collapse caused by fatal tachyarrhythmia. In cases where the patient survives cardiopulmonary resuscitation, an implantable cardioverter defibrillator (ICD) is implanted as a secondary prevention of sudden cardiac death, but the damage caused by the electric shock treatment is significant. In addition, the similar device is implanted as a primary prevention in high-risk groups for sudden cardiac death, such as those with low cardiac function or hereditary arrhythmia, but there is no accurate risk stratification method based on electrocardiograms. In this study, we aimed to develop a novel technology to predict the risk of sudden cardiac death by converting the electrocardiograms of sudden cardiac death cases into normal electrocardiograms using a generative adversarial network (GAN), and then detecting and evaluating the difference between the electrocardiogram before conversion and the normal electrocardiogram.
Free Research Field	循環器内科
Academic Significance and Societal Importance of the Research Achievements	わが国では年間約８万例の心臓突然死が発生する。１２誘導心電図は心臓病の診断や病態評価に重要な役割を果たすが、ハイリスク症例を特定する目的で詳細に検討されることは少ない。今回は心臓突然死例と正常例の１２誘導心電図の間で敵対的生成ネットワークを用いて、互いの心電図を再生成することで、2者の心電図波形の相違性を明らかにして突然死ハイリスク心電図の特徴を明らかにすることを目的とした。心電図検査の際に、このアルゴリズムを使用することにより心臓突然死ハイリスク例の検出を行い、生命予後を改善したいと考えている。