Sequence-based dynamic handwriting analysis for Parkinson’s disease detection with one-dimensional convolutions and BiGRUs

作者：

Highlights：

• Sequence-based dynamic handwriting analysis for Parkinson’s Disease identification.

• Combination of one-dimensional convolutions and Bi-GRU layers for classification.

• Robust feature learning through convolutions on features derived from handwriting.

• Experiments on the PaHaW and NewHandPD datasets achieve state-of-the-art results.

摘要

•Sequence-based dynamic handwriting analysis for Parkinson’s Disease identification.•Combination of one-dimensional convolutions and Bi-GRU layers for classification.•Robust feature learning through convolutions on features derived from handwriting.•Experiments on the PaHaW and NewHandPD datasets achieve state-of-the-art results.

论文关键词：Parkinson’s disease,Dynamic handwriting analysis,Recurrent neural networks,Computer-aided diagnosis

论文评审过程：Received 4 August 2020, Revised 5 November 2020, Accepted 27 November 2020, Available online 1 December 2020, Version of Record 9 December 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114405