Determination of Sputum Suction Timing in Mechanical Ventilation Based on Transfer Learning and Breath Sounds Recognition
10.16156/j.1004-7220.2025.05.031
- VernacularTitle:基于迁移学习与呼吸音特征识别的机械通气吸痰时机判别
- Author:
Shuai WANG
1
;
Jiangzhen GUO
1
;
Chunjing TAO
1
Author Information
1. 北京航空航天大学医学科学与工程学院;北京生物医学工程高精尖创新中心,北京 100191
- Publication Type:Journal Article
- Keywords:
mechanical ventilation;
breath sound;
transfer learning;
determination of sputum suction time
- From:
Journal of Medical Biomechanics
2025;40(5):1318-1324
- CountryChina
- Language:Chinese
-
Abstract:
Objective To propose a transfer learning-based method for breath sound feature recognition and autonomous determination of sputum suction timing.Methods An electronic stethoscope was used to collect breath sounds from the main airways of clinically ventilated patients before and after sputum suction,with pre-suction breath sounds labeled as requiring suction.The collected data underwent high-pass filtering and wavelet soft-threshold denoising,followed by the extraction of log-Mel spectrograms.A VGGish model pretrained on the Audio Set dataset was then employed to extract feature vectors from these spectrograms,which were subsequently classified using a support vector machine to determine whether suction was required.Results The precision,recall and F1 score for recognition of breath sounds requiring sputum suction were 86.73％,93.06％and 89.78％,respectively.Conclusions The proposed breath sound recognition method based on transfer learning effectively determines the timing of sputum suction and shows a significant clinical potential.
