Study on ECG automatic analysis technology and its clinical usefulness based on artificial intelligence and cardiac simulation

2020 Fiscal Year Final Research Report

• Project/Area Number: 18K11532
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 62010:Life, health and medical informatics-related
• Research Institution: The University of Aizu
• Principal Investigator: ZHU XIN 会津大学, コンピュータ理工学部, 上級准教授 (70448645)
• Co-Investigator(Kenkyū-buntansha): 野呂 眞人 東邦大学, 医学部, 臨床教授 (10366495) ; 中村 啓二郎 東邦大学, 医学部, 助教 (20366181)
• Project Period (FY): 2018-04-01 – 2021-03-31
• Keywords: 心電図 / 深層学習 / 不整脈 / 心臓モデル

Outline of Final Research Achievements

In this research, we propose two method to synthesize simulation ECG for the development of automatic ECG interpretation algorithms. At first, we constructed cardiac models to synthesize simulation ECG. Secondly, we used clinical ECG data to synthesize ECG using anniversary neural networks. Then, we added Physionet Open database to construct the database for the training and testing of deep neural networks. Based on deep learning, we proposed ECG noise recognition algorithm, atrial fibrillation detection algorithm, QT interval measurement algorithm, and obstructive sleep apnea detection algorithm. These algorithms demonstrate better performance compared with traditional methods.

Free Research Field

生体医工学

Academic Significance and Societal Importance of the Research Achievements

本研究は、心臓モデル、対抗ニューラルネットワークを用い、深層ニューラルネットワークの学習・テスト用の心電図を合成し、より少ないデータで心電図自動解析アルゴリズムを開発できた。開発した心電図からノイズの識別アルゴリズム、心房細動心電図波形の検出アルゴリズム、QT間隔の自動計測アルゴリズム、閉塞性睡眠時無呼吸の自動検出アルゴリズムはいずれも高い臨床価値があり、心疾患の早期診断・治療に役立て、国民の健康増進、医療費の低減に貢献できると考えられる。