Medical Speech Symptoms Classification via Disentangled Representation
Preprint
Abstract
Intent is defined for understanding spoken language in existing works. Both textual features and acoustic features involved in medical speech contain intent, which is important for symptomatic diagnosis. In this paper, we propose a medical speech classification model named DRSC that automatically learns to disentangle intent and content representations from textual-acoustic data for classification. The intent representations of the text domain and the Mel-spectrogram domain are extracted via intent encoders, and then the reconstructed text feature and the Mel-spectrogram feature are obtained through two exchanges. After combining the intent from two domains into a joint representation, the integrated intent representation is fed into a decision layer for classification. Experimental results show that our model obtains an average accuracy rate of 95% in detecting 25 different medical symptoms.Keywords
All Related Versions
This publication has 8 references indexed in Scilit:
- Automatic illness prediction system through speechComputers and Electrical Engineering, 2022
- A comprehensive survey on multimodal medical signals fusion for smart healthcare systemsInformation Fusion, 2021
- Medical Image Generation Using Generative Adversarial Networks: A Review2021
- Multimodal speech emotion recognition and classification using convolutional neural network techniquesInternational Journal of Speech Technology, 2020
- Clinical Reasoning Assessment Methods: A Scoping Review and Practical GuidanceAcademic Medicine, 2019
- Speech recognition for clinical documentation from 1990 to 2018: a systematic reviewJournal of the American Medical Informatics Association, 2019
- Speech Intention Classification with Multimodal Deep Learning2017
- Determining the initial states in forward-backward filteringIEEE Transactions on Signal Processing, 1996