Robust automated Parkinson disease detection based on voice signals with transfer learning

作者：

Highlights：

• The CNN model was developed for automated PD identification based on voice characteristics.

• The developed CNN model can successfully identify the PD with an accuracy of 91.17%.

• It will be possible to detect PD by implementing the proposed CNN model to a smart device for personal usage.

• The physician can utilize the developed model for PD detection during the in-clinic assessment.

摘要

•The CNN model was developed for automated PD identification based on voice characteristics.•The developed CNN model can successfully identify the PD with an accuracy of 91.17%.•It will be possible to detect PD by implementing the proposed CNN model to a smart device for personal usage.•The physician can utilize the developed model for PD detection during the in-clinic assessment.

论文关键词：Parkinson's disease (PD),Acoustic sensing,Convolutional neural network (CNN),Transfer Learning,Voice signal

论文评审过程：Received 7 March 2021, Revised 30 March 2021, Accepted 7 April 2021, Available online 13 April 2021, Version of Record 21 April 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115013