Automated Detection and Localization of Myocardial Infarction With Interpretability Analysis Based on Deep Learning

被引:15
作者
Han, Chuang [1 ]
Sun, Jiajia [2 ]
Bian, Yingnan [3 ]
Que, Wenge [4 ]
Shi, Li [4 ]
机构
[1] Zhengzhou Univ Light Ind, Sch Comp & Commun Engn, Zhengzhou 450000, Peoples R China
[2] Zhengzhou Univ Light Ind, Inst Polit Sci & Law, Zhengzhou 450000, Peoples R China
[3] Henan Coll Transportat, Sch Logist, Zhengzhou 450000, Peoples R China
[4] Tsinghua Univ, Beijing Natl Res Ctr Informat Sci & Technol, Dept Automat, Beijing 100000, Peoples R China
关键词
Lead; Electrocardiography; Feature extraction; Solid modeling; Residual neural networks; Kernel; Information representation; Class activation mapping with gradient (Grad-CAM); deep neural networks; electrocardiogram (ECG); interpretability; myocardial infarction (MI); CONVOLUTIONAL NEURAL-NETWORK; PHASE DISTRIBUTION PATTERN; DIAGNOSIS; CLASSIFICATION;
D O I
10.1109/TIM.2023.3258521
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Electrocardiogram (ECG) is a noninvasive, simplest, and fastest way to diagnose myocardial infarction (MI). Although different methods have been leveraged based upon deep learning covered by the existing studies, the spatial-temporal relationship in the lead and between leads has not been deeply analyzed. To address the issue, a novel multilead branch with the residual network integrated with squeeze and excitation networks and bidirectional long short-term memory (LSTM) model named MLB-ResNet-SENet-BL was presented. First, spatial features were exploited by the morphological information representation network in the lead based on MLB-ResNet. Then, these feature mappings among these spatial features based on SENet were strengthened and weakened by the importance analysis network of feature mapping in the lead, respectively. In addition, the temporal features were extracted by the dependency network between the leads based on BLSTM. Meanwhile, the model was evaluated using fivefold cross validation for MI detection and localization based on PTB and PTB-XL. The resulting model outperforms the state-of-the-art studies for intra-patient and inter-patient paradigms. The interpretability analysis using class activation mapping with gradient (Grad-CAM) was also leveraged for visualization of the specific QRS waves and ST-T segments of 12-leads ECG, which demonstrated that the highlighted parts of heat maps were completely in line with the diagnostic basis and strategy of doctors. Deployment of such models can potentially help ensure the life safety of patients and strive for the best treatment opportunity.
引用
收藏
页数:12
相关论文
共 50 条
  • [21] Derived vectorcardiogram based automated detection of posterior myocardial infarction using FBSE-EWT technique
    Khan, Sibghatullah, I
    Pachori, Ram Bilas
    BIOMEDICAL SIGNAL PROCESSING AND CONTROL, 2021, 70
  • [22] DEEP LEARNING-BASED FINE-GRAINED AUTOMATED PNEUMONIA DETECTION MODEL
    Rangasamy, Keerthana
    Fuzi, Nurul Shuhada Mohd
    As'ari, Muhammad Amir
    Rahmad, Nur Azmina
    Sufri, Nur Anis Jasmin
    JOURNAL OF ENGINEERING SCIENCE AND TECHNOLOGY, 2022, 17 (04): : 2373 - 2389
  • [23] An Interpretability Optimization Method for Deep Learning Networks Based on Grad-CAM
    Zhang, Yubo
    Zhu, Yong
    Liu, Junli
    Yu, Wei
    Jiang, Chuang
    IEEE INTERNET OF THINGS JOURNAL, 2025, 12 (04): : 3961 - 3970
  • [24] ECG Analysis Using Multiple Instance Learning for Myocardial Infarction Detection
    Sun, Li
    Lu, Yanping
    Yang, Kaitao
    Li, Shaozi
    IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 2012, 59 (12) : 3348 - 3356
  • [25] CardioNet: A Lightweight Deep Learning Framework for Screening of Myocardial Infarction Using ECG Sensor Data
    Gupta, Kapil
    Bajaj, Varun
    Ansari, Irshad Ahmad
    IEEE SENSORS JOURNAL, 2025, 25 (04) : 6794 - 6800
  • [26] Explainable detection of myocardial infarction using deep learning models with Grad-CAM technique on ECG signals
    Jahmunah, V.
    Ng, E. Y. K.
    Tan, Ru-San
    Oh, Shu Lih
    Acharya, U. Rajendra
    COMPUTERS IN BIOLOGY AND MEDICINE, 2022, 146
  • [27] Diagnostic Accuracy of the Deep Learning Model for the Detection of ST Elevation Myocardial Infarction on Electrocardiogram
    Choi, Hyun Young
    Kim, Wonhee
    Kang, Gu Hyun
    Jang, Yong Soo
    Lee, Yoonje
    Kim, Jae Guk
    Lee, Namho
    Shin, Dong Geum
    Bae, Woong
    Song, Youngjae
    JOURNAL OF PERSONALIZED MEDICINE, 2022, 12 (03):
  • [28] Deep Learning for Detecting and Locating Myocardial Infarction by Electrocardiogram: A Literature Review
    Xiong, Ping
    Lee, Simon Ming-Yuen
    Chan, Ging
    FRONTIERS IN CARDIOVASCULAR MEDICINE, 2022, 9
  • [29] Inferior Myocardial Infarction Detection From Lead II of ECG: A Gramian Angular Field-Based 2D-CNN Approach
    Yousuf, Asim
    Hafiz, Rehan
    Riaz, Saqib
    Farooq, Muhammad
    Riaz, Kashif
    Rahman, Muhammad Mahboob Ur
    IEEE SENSORS LETTERS, 2024, 8 (10)
  • [30] Automated Localization of Myocardial Infarction From Vectorcardiographic via Tensor Decomposition
    Zhang, Jieshuo
    Liu, Ming
    Xiong, Peng
    Du, Haiman
    Yang, Jianli
    Xu, Jinpeng
    Hou, Zengguang
    Liu, Xiuling
    IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 2023, 70 (03) : 812 - 823