Please use this identifier to cite or link to this item:
Title: Visual representation of online handwriting time series for deep learning Parkinson’s disease detection
Authors: Taleb, Catherine
Khachab, Maha 
Mokbel, Chafic 
Likforman-Sulem, Laurence
Affiliations: Faculty of Medicine 
Faculty of Engineering 
Keywords: CNN
Gramian Angular Field images
Parkinson’s Disease
PDMultiMC dataset
Spectrogram images
Issue Date: 2019-01-01
Part of: International Conference on Document Analysis and Recognition Workshops, ICDARW 2019
Conference: International Workshop on Arabic and Derived Script Analysis and Recognition, ASAR 2019 - ICDAR 2019 ( 3rd : 22-25 Spt, 2019 : Sydney, Australia )
Parkinson’s disease (PD) is a neurological disorder associated with a progressive decline in motor skills, speech, and cognitive processes. Since the diagnosis of Parkinson’s disease is difficult, researchers have worked to develop a support tool based on algorithms to separate healthy controls from PD patients. Online handwriting dynamic signals can provide more detailed and complex information for PD detection task. Existing techniques often depended on handcrafted features that required expert knowledge of the field. In this paper, it is suggested to learn pen-based features by means of deep learning for automatic classification of PD. For this purpose, a visual representation of the time series can be computed and used at the input of a convolutional neural network (CNN) as in [4]. Classically, the time series is transformed into a fixed dimension image applying normalization on the time dimension. In this work we have experimented several visual representations, including the spectrogram where normalization of the time scale is applied after short term information has been extracted locally. We have been able to show that considering the local short term information allows the deep learning models to provide better classification results compared to a globally normalized fixed dimension visual representation. For validation purpose, a CNN-BLSTM was directly applied on the time series, without any normalization of the time scale which led to best performance equivalent to the one obtained on spectrogram representation.
ISBN: 9781728150543
DOI: 10.1109/ICDARW.2019.50111
Ezproxy URL: Link to full text
Type: Conference Paper
Appears in Collections:Faculty of Medicine
Department of Computer Engineering

Show full item record


checked on May 27, 2023

Record view(s)

checked on May 28, 2023

Google ScholarTM


Dimensions Altmetric

Dimensions Altmetric

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.