Abstract：This study proposes a method for classifying cardiac diseases by extracting 3D matrix features from multi-channel heart sound signals, addressing the limitations of traditional methods that primarily utilize one-dimensional features from single-channel signals, which may overlook pathological correlations across different channels and cycles. First, a Butterworth filter was applied for noise reduction on the heart sound signals from each channel. The R-wave peaks were then located to segment the heart sounds, from which 15 effective time-frequency features, including Welch method power spectral energy, are extracted. Subsequently, these features were stacked into a 3D matrix with dimensions corresponding to the number of channels, cycles, and features, with the optimal cycle number determined to be 4. This 3D matrix was directly used as input for a CNN classifier. The method was tested on a dataset comprising 126 normal and 185 abnormal heart sounds, achieving an accuracy of 98.9%. Additionally, the method was validated on 126 clinical cases of congenital heart disease sounds and normal sounds, resulting in a classification accuracy of 93.9%. These experimental results indicated that the 3D matrix features could effectively capture pathological characteristics in the heart sound signals, improving the classification accuracy by 2.7% compared to single-channel features, providing valuable assistance for clinical cardiac treatment.
[1] Juan PDM, Angel FJF, Manuel JDM, et al. deep neural networks for the recognition and classification of heart murmurs using neuromorphic auditory sensors [J]. IEEE Transactions on Biomedical Circuits and Systems ,2017,12(1),24-34.
[2] Badlaoui OE, Benba A, Hammouch A. Novel PCG analysis method for discriminating between abnormal and normal heart sounds [J]. IRBM, 2020, 41(4): 223-228.
[3] Barma S, Chen BW, Ji W, et al. Detection of the third heart sound based on nonlinear signal decomposition and time-frequency localization [J]. IEEE Trans Biomed Eng, 2016, 63(8),1718-1727.
[4] 张磊邦, 唐荣斌, 蒋建波, 等. 心音信号的预处理与包络提取算法研究 [J]. 生物医学工程学杂志, 2014, 31(4): 734-741.
[5] Deng Muqing, Meng Tingting, Cao Jiuwen, et al. Heart sound classification based on improved mfcc features and convolutional recurrent neural networks [J]. Neural Networks, 2020, 130: 22-32.
[6] Deperlioglu O. Heart sound classification with signal instant energy and stacked autoencoder network [J]. Biomedical Signal Processing and Control, 2021, 64(1): 102-211.
[7] 周酥, 朱蒂, 吴效明, 等. 基于小波变换的心音包络提取算法及应用 [J]. 中国组织工程研究与临床康复, 2011, 15(30): 5615-5619.
[8] Ajitkumar SS, Ashinikumar SS, Dinita ND, et al. Heart abnormality classification using PCG and ECG recordings [J]. Instituto Politecnico Nacional/Centro de Investigacion en Computacion, 2021, 25(2):381-391.
[9] 向家伟, 刘怡, 司泽田. 基于差分平方包络非局部均值的滚动轴承故障诊断方法及系统 [P].中国专利:CN202111111764.4,2022-01-18.
[10] Wang Shouyan, Zhang Lifan. Time-frequency analysis of heart rate variability [J]. 中国生物医学工程学报(英文版), 1997,16(5):119-126.
[11] 邢晓晴, 朱根民. Welch功率谱估计中窗函数的选择与算法分析 [J]. 计算机时代, 2018,30(2): 1-4.
[12] 王佳明,张安琪,郝宏燕,等. 基于功率谱密度相似性的最佳心音信号选择 [J]. 中国医学物理学杂志, 2022, 39(11): 1401-1406.
[13] 张琦, 泮思林, 刘凯. 先天性肺静脉曲张合并室间隔缺损一例[J]. 放射学实践, 2014, 29(8): 905.
[14] 刘伟伟, 桑胜波, 张宏鹏. 基于CNN+LSTM的改进心音分类模型研究 [J]. 电子设计工程, 2022, 30(2): 38-42.
[15] Zhang Wenjie, Han Jiaqing, Deng Shiwen. Heart sound classification based on scaled spectrogram and tensor decomposition [J]. Expert Systems with Applications, 2017, 84: 220-231.
[16] Metlapalli CA, Muthusamy T, Battula PB. Classification of Social Media Text Spam Using VAE-CNN and LSTM Model[J]. 2020, 25(6): 747-753.
[17] 陈尧. 心音心电信号处理的神经网络方法 [D].成都:四川大学, 2021.
[18] 陈成, 潘家华, 孙静, 等. 基于MFCC与GFCC混合特征的先心病心音分类研究 [J]. 计算机仿真, 2022, 39(7): 263-269.
