Phonocardiogram transfer learning-based CatBoost model for diastolic dysfunction identification using multiple domain-specific deep feature fusion

作者全名："Zheng, Yineng; Guo, Xingming; Yang, Yang; Wang, Hui; Liao, Kangla; Qin, Jian"

作者地址："[Zheng, Yineng] Chongqing Med Univ, Dept Radiol, Affiliated Hosp 1, Chongqing 400016, Peoples R China; [Liao, Kangla; Qin, Jian] Chongqing Med Univ, Dept Cardiol, Affiliated Hosp 1, Chongqing 400016, Peoples R China; [Guo, Xingming; Yang, Yang; Wang, Hui] Chongqing Univ, Coll Bioengn, Key Lab Biorheol Sci & Technol, Minist Educ, Chongqing 400044, Peoples R China; [Zheng, Yineng] Chongqing Med Univ, State Key Lab Ultrasound Med & Engn, Chongqing 400016, Peoples R China; [Zheng, Yineng] Chongqing Med Univ, Med Data Sci Acad, Chongqing 400016, Peoples R China; [Guo, Xingming] 164 Shazheng Rd, Chongqing 400044, Peoples R China; [Zheng, Yineng] 1 Youyi Rd, Chongqing 400016, Peoples R China"

通信作者："Guo, XM (通讯作者)，164 Shazheng Rd, Chongqing 400044, Peoples R China.; Zheng, YN (通讯作者)，1 Youyi Rd, Chongqing 400016, Peoples R China."

来源：COMPUTERS IN BIOLOGY AND MEDICINE

ESI学科分类：COMPUTER SCIENCE

WOS号：WOS:000962541000001

JCR分区：Q1

影响因子：7

年份：2023

卷号：156

期号：　

开始页：　

结束页：　

文献类型：Article

关键词：Phonocardiogram; Heart sounds; Diastolic dysfunction detection; Transfer learning; Computer-aided diagnosis

摘要："Left ventricular diastolic dyfunction detection is particularly important in cardiac function screening. This paper proposed a phonocardiogram (PCG) transfer learning-based CatBoost model to detect diastolic dysfunction noninvasively. The Short-Time Fourier Transform (STFT), Mel Frequency Cepstral Coefficients (MFCCs), S-transform and gammatonegram were utilized to perform four different representations of spectrograms for learning the representative patterns of PCG signals in two-dimensional image modality. Then, four pre-trained convolutional neural networks (CNNs) such as VGG16, Xception, ResNet50 and InceptionResNetv2 were employed to extract multiple domain-specific deep features from PCG spectrograms using transfer learning, respectively. Further, principal component analysis and linear discriminant analysis (LDA) were applied to different feature subsets, respectively, and then these different selected features are fused and fed into CatBoost for classification and performance comparison. Finally, three typical machine learning classifiers such as multilayer perceptron, support vector machine and random forest were employed to compared with CatBoost. The hyperparameter optimization of the investigated models was determined through grid search. The visualized result of the global feature importance showed that deep features extracted from gammatonegram by ResNet50 contributed most to classification. Overall, the proposed multiple domain-specific feature fusion based CatBoost model with LDA achieved the best performance with an area under the curve of 0.911, accuracy of 0.882, sensitivity of 0.821, specificity of 0.927, F1-score of 0.892 on the testing set. The PCG transfer learning-based model developed in this study could aid in diastolic dysfunction detection and could contribute to non-invasive evaluation of diastolic function."

基金机构："National Natural Science Foundation of China [31800823, 31870980]; Joint project of Chongqing Health Commission and Science and Technology Bureau [2022QNXM015]; Natural Science Foundation of Chongqing [cstc2019jcyj-msxmX0395]; Intelligent Medicine Research Project of Chongqing Medical University [ZHYX202102]"

基金资助正文："This study was supported by the National Natural Science Foundation of China (No. 31800823 and No. 31870980), Joint project of Chongqing Health Commission and Science and Technology Bureau (2022QNXM015), the Natural Science Foundation of Chongqing (cstc2019jcyj-msxmX0395), and Intelligent Medicine Research Project of Chongqing Medical University (ZHYX202102)."