Deep Neural Network for Respiratory Sound Classification in Wearable Devices Enabled by Patient Specific Model Tuning

• Published in: IEEE transactions on biomedical circuits and systems Vol. 14; no. 3; pp. 535 - 544
• Main Authors: Acharya, Jyotibdha, Basu, Arindam
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.06.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Acoustics; Anomalies; Artificial neural networks; Automation; Biological system modeling; Breathing; Classification; CNN; Data models; Deep learning; Diseases; Feature extraction; Humans; LSTM; Lung diseases; Machine learning; Measurement; Neural networks; Neural Networks, Computer; patient specific model; Patient-Specific Modeling; Quantization (signal); Respiration; respiratory audio analysis; Respiratory Sounds - classification; Signal Processing, Computer-Assisted; Sound; Spectrograms; Strategy; Training; Tuning; Wearable technology; weight quantization
• ISSN: 1932-4545; 1940-9990; 1940-9990
• DOI: 10.1109/TBCAS.2020.2981172

The primary objective of this paper is to build classification models and strategies to identify breathing sound anomalies (wheeze, crackle) for automated diagnosis of respiratory and pulmonary diseases. In this work we propose a deep CNN-RNN model that classifies respiratory sounds based on Mel-spectrograms. We also implement a patient specific model tuning strategy that first screens respiratory patients and then builds patient specific classification models using limited patient data for reliable anomaly detection. Moreover, we devise a local log quantization strategy for model weights to reduce the memory footprint for deployment in memory constrained systems such as wearable devices. The proposed hybrid CNN-RNN model achieves a score of <inline-formula><tex-math notation="LaTeX">\text{66.31}\%</tex-math></inline-formula> on four-class classification of breathing cycles for ICBHI'17 scientific challenge respiratory sound database. When the model is re-trained with patient specific data, it produces a score of <inline-formula><tex-math notation="LaTeX">\text{71.81}\%</tex-math></inline-formula> for leave-one-out validation. The proposed weight quantization technique achieves <inline-formula><tex-math notation="LaTeX">{\approx}4 \times</tex-math></inline-formula> reduction in total memory cost without loss of performance. The main contribution of the paper is as follows: Firstly, the proposed model is able to achieve state of the art score on the ICBHI'17 dataset. Secondly, deep learning models are shown to successfully learn domain specific knowledge when pre-trained with breathing data and produce significantly superior performance compared to generalized models. Finally, local log quantization of trained weights is shown to be able to reduce the memory requirement significantly. This type of patient-specific re-training strategy can be very useful in developing reliable long-term automated patient monitoring systems particularly in wearable healthcare solutions.