Comprehensive electrocardiographic diagnosis based on deep learning

Abstract

Cardiovascular disease (CVD) is the leading cause of death worldwide, and coronary artery disease (CAD) is a major contributor. Early-stage CAD can progress if undiagnosed and left untreated, leading to myocardial infarction (MI) that may induce irreversible heart muscle damage, resulting in heart chamber remodeling and eventual congestive heart failure (CHF). Electrocardiography (ECG) signals can be useful to detect established MI, and may also be helpful for early diagnosis of CAD. For the latter especially, the ECG perturbations can be subtle and potentially misclassified during manual interpretation and/or when analyzed by traditional algorithms found in ECG instrumentation. For automated diagnostic systems (ADS), deep 2 learning techniques are favored over conventional machine learning techniques, due to the automatic feature extraction and selection processes involved. This paper highlights various deep learning algorithms exploited for the classification of ECG signals into CAD, MI, and CHF conditions. The Convolutional Neural Network (CNN), followed by combined CNN and Long Short-Term Memory (LSTM) models, appear to be the most useful architectures for classification. A 16-layer LSTM model was developed in our study and validated using 10-fold cross-validation. A classification accuracy of 98.5% was achieved. Our proposed model has the potential to be a useful diagnostic tool in hospitals for the classification of abnormal ECG signals.

Myocardial infarction detection using artificial intelligence and nonlinear features based on synthesis of the standard 12-lead and Frank XYZ leads

“…The reported accuracy for the automated detection of normal and MI ECG beatsby using CNN was 96.1%. Lih et al [93] proposed combined 16-layer CNN and Long Short-Term Memory (LSTM) models for the detection of MI with automatic feature extraction and selection. The reported accuracy was 98%.…”

Metabolic Imaging Based Sub-Classification of Lung Cancer

“…However, there is no rule on this issue yet. In this study, the dataset was randomly split for stratified 10fold cross validation which was used in many deep and conventional machine learning based bioimage and biosignal studies [29]. CNN and MLP models were trained from scratch starting with random weights using NVIDIA DGX-1 with NVIDIA Tesla V100 GPUs at Abdullah Gül University, High Performance Computing (HPC) Laboratory.…”

Machine learning as a tool for diagnostic and prognostic research in coronary artery disease

“…Методы ГО являются базовой платформой для приложений распознавания изображений, которые планируются или уже используются для визуализации: ангиография, ЭхоКГ, компьютерная томография, внутрисосудистое ультразвуковое исследование, оптическая когерентная томография и др. [3,5,51]. Наиболее распространенными вариантами ГО являются сверточные (CNN) и рекуррентные нейронные сети (RNN) [51].…”

“…[3,5,51]. Наиболее распространенными вариантами ГО являются сверточные (CNN) и рекуррентные нейронные сети (RNN) [51]. В 2016г была впервые опубликована статья, в которой представлены результаты применения CNN для автоматической верификации патологических признаков ЭКГ [52].…”