Comparison of window shapes and lengths in short-time feature extraction for classification of heart sound signals

2026-04-15

AI Summary

The non-stationary nature of heart sound signals renders feature extraction highly sensitive to the shape and length of the analysis window, thereby limiting the performance of automated cardiovascular disease classification. This work proposes a sliding-window-based short-time statistical feature extraction approach combined with a bidirectional long short-term memory (biLSTM) network for classification. For the first time, it systematically evaluates the impact of Gaussian, triangular, and rectangular windows across varying window lengths. Experimental results demonstrate that a 75-ms Gaussian window yields the best classification performance, significantly outperforming baseline methods; the triangular window achieves the second-best results, while the rectangular window performs worst. These findings provide clear guidance for optimal window parameter selection in heart sound signal analysis.

Technology Category

• Machine Learning: Kernel Methods
• Natural Language Processing: Learning & Optimization for NLP
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Heart sound signals, phonocardiography (PCG) signals, allow for the automatic diagnosis of potential cardiovascular pathology. Such classification task can be tackled using the bidirectional long short-term memory (biLSTM) network, trained on features extracted from labeled PCG signals. Regarding the non-stationarity of PCG signals, it is recommended to extract the features from multiple short-length segments of the signals using a sliding window of certain shape and length. However, some window contains unfavorable spectral side lobes, which distort the features. Accordingly, it is preferable to adapt the window shape and length in terms of classification performance. We propose an experimental evaluation for three window shapes, each with three window lengths. The biLSTM network is trained and tested on statistical features extracted, and the performance is reported in terms of the window shapes and lengths. Results show that the best performance is obtained when the Gaussian window is used for splitting the signals, and the triangular window competes with the Gaussian window for a length of 75 ms. Although the rectangular window is a commonly offered option, it is the worst choice for splitting the signals. Moreover, the classification performance obtained with a 75 ms Gaussian window outperforms that of a baseline method.

Problem

• heart sound signals
• window shape
• window length
• feature extraction
• classification performance

Innovation

• window shape
• short-time feature extraction
• heart sound classification
• biLSTM
• phonocardiography