Abstract: Asthma, a prevalent chronic respiratory disorder, affects millions worldwide, necessitating accurate and timely diagnosis for effective management. Recent advancements in artificial intelligence (AI) have facilitated the development of non-invasive diagnostic tools, particularly in analyzing respiratory sounds. In this paper, we propose a novel asthma recognition system utilizing Convolutional Neural Networks (CNNs) and Long Short-Term Memory networks (LSTMs) for analyzing respiratory sound data.
The proposed system begins by preprocessing raw respiratory sound recordings to extract relevant features. These features are then fed into a CNN-LSTM architecture, which effectively captures both spatial and temporal dependencies in the data. The CNN component learns hierarchical feature representations from spectrogram-like input representations of respiratory sound, while the LSTM component learns temporal patterns and dependencies. We conduct experiments on publicly available respiratory sound datasets to evaluate the performance of the proposed system. We compare our approach with existing methods and demonstrate its superior performance in terms of accuracy, sensitivity, and specificity in asthma recognition.
Furthermore, we analyze the interpretability of the CNN-LSTM model to provide insights into its decision-making process, enhancing its trustworthiness in clinical applications. Additionally, we discuss the scalability and deployment feasibility of the proposed system in real-world healthcare settings.
Our findings suggest that the CNN-LSTM-based asthma recognition system offers a promising avenue for accurate and automated asthma diagnosis using respiratory sound data. By leveraging deep learning techniques, this system has the potential to improve diagnostic accuracy, reduce healthcare costs, and enhance patient care in asthma management. patients.
Key Words: Asthma, Respiratory Sound, Healthcare, Machine learning
|
DOI:
10.17148/IJARCCE.2024.13573
