Applying masked autoencoder-based self-supervised learning for high-capability vision transformers of electrocardiographies

• Published in: PloS one Vol. 19; no. 8; p. e0307978
• Main Authors: Sawano, Shinnosuke, Kodera, Satoshi, Setoguchi, Naoto, Tanabe, Kengo, Kushida, Shunichi, Kanda, Junji, Saji, Mike, Nanasato, Mamoru, Maki, Hisataka, Fujita, Hideo, Kato, Nahoko, Watanabe, Hiroyuki, Suzuki, Minami, Takahashi, Masao, Sawada, Naoko, Yamasaki, Masao, Sato, Masataka, Katsushika, Susumu, Shinohara, Hiroki, Takeda, Norifumi, Fujiu, Katsuhito, Daimon, Masao, Akazawa, Hiroshi, Morita, Hiroyuki, Komuro, Issei
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 14-08-2024; Public Library of Science (PLoS)
• Subjects: Adult; Aged; Algorithms; Artificial neural networks; Biology and Life Sciences; Cardiac arrhythmia; Clinical medicine; Computer and Information Sciences; Data analysis; Datasets; Deep Learning; Echocardiography; Echocardiography - methods; EKG; Electric transformers; Electrocardiogram; Electrocardiography; Electrocardiography - methods; Female; Heart; Hospitals; Humans; Image processing; Machine learning; Male; Medical imaging equipment; Medicine and Health Sciences; Middle Aged; Neural networks; Neural Networks, Computer; Patients; Performance enhancement; Performance evaluation; Physical Sciences; Research and Analysis Methods; ROC Curve; Self-supervised learning; Social Sciences; Supervised Machine Learning; Transformers; Ventricular Dysfunction, Left - diagnosis; Ventricular Dysfunction, Left - physiopathology; Vision; Japan
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0307978

The generalization of deep neural network algorithms to a broader population is an important challenge in the medical field. We aimed to apply self-supervised learning using masked autoencoders (MAEs) to improve the performance of the 12-lead electrocardiography (ECG) analysis model using limited ECG data. We pretrained Vision Transformer (ViT) models by reconstructing the masked ECG data with MAE. We fine-tuned this MAE-based ECG pretrained model on ECG-echocardiography data from The University of Tokyo Hospital (UTokyo) for the detection of left ventricular systolic dysfunction (LVSD), and then evaluated it using multi-center external validation data from seven institutions, employing the area under the receiver operating characteristic curve (AUROC) for assessment. We included 38,245 ECG-echocardiography pairs from UTokyo and 229,439 pairs from all institutions. The performances of MAE-based ECG models pretrained using ECG data from UTokyo were significantly higher than that of other Deep Neural Network models across all external validation cohorts (AUROC, 0.913-0.962 for LVSD, p < 0.001). Moreover, we also found improvements for the MAE-based ECG analysis model depending on the model capacity and the amount of training data. Additionally, the MAE-based ECG analysis model maintained high performance even on the ECG benchmark dataset (PTB-XL). Our proposed method developed high performance MAE-based ECG analysis models using limited ECG data.