Motivation

Parkinson’s disease (PD), is a neurodegenerative disorder, whose symptoms (tremor, rigidity, and intentional movements) negatively affect the patient’s daily lives. While there is no cure for PD, early diagnosis and proper treatment are of great importance in easing the patient from symptoms and, in turn, improving the quality of daily life. Analysis of drawing and handwriting skills is one of the procedures widely used in the diagnostics of PD. Deep learning has achieved remarkable results in the medical field, such as the diagnosis, segmentation, prediction, and detection of various anatomical regions of interest (ROI). Artificial intelligence-based approaches could help specialists and physicians in supporting the diagnosis of PD.

Name of Researcher: Xuechao Wang, Junqing Huang, Marianna Chatzakou

Scientist in Charge: Prof. Michael Ruzhansky, Dr. Sven Nõmm

Host: Ghent University, Tallinn University of Technology

In this project, our objective is to introduce deep learning to assist traditional methods of diagnosing Parkinson’s disease, specifically, to transform the collected handwriting exam data into different dimensions, to combine it with the appropriate deep learning models. The research project is divided into three parts:

Part I time-series data

Part II color-image data

Part III point-cloud data

Publications:

• Wang X., Huang J., Chatzakou M., Nomm S., Valla E., Medijainen K., Taba P., Toomela A., Ruzhansky M., Comparison of one- two- and three-dimensional CNN models for drawing-test-based diagnostics of the Parkinson’s disease. Biomedical Signal Processing and Control, 87 (2024) 105436. doi, arxiv
• Wang X., Huang J., Chatzakou M., Medijainen K., Taba P., Toomela A., Nomm S., Ruzhansky M., A Light-weight CNN Model for Efficient Parkinson’s Disease Diagnostics. arxiv
• Wang X., Huang J., Nomm S., Chatzakou M., Medijainen K., Toomela A., Ruzhansky M., LSTM-CNN: An efficient diagnostic network for Parkinson’s disease utilizing dynamic handwriting analysis. arxiv

References：

• Mei J, Desrosiers C, Frasnelli J. Machine learning for the diagnosis of Parkinson’s disease: a review of literature[J]. Frontiers in aging neuroscience, 2021, 13: 633752.
• Loh H W, Hong W, Ooi C P, et al. Application of deep learning models for automated identification of Parkinson’s disease: a review (2011–2021)[J]. Sensors, 2021, 21(21): 7034.
• Landolfi A, Ricciardi C, Donisi L, et al. Machine Learning Approaches in Parkinson’s Disease[J]. Current Medicinal Chemistry, 2021, 28(32): 6548-6568.
• Rana A, Dumka A, Singh R, et al. Imperative Role of Machine Learning Algorithm for Detection of Parkinson’s Disease: Review, Challenges and Recommendations[J]. Diagnostics, 2022, 12(8): 2003.
• Pereira C R, Pereira D R, Weber S A T, et al. A survey on computer-assisted Parkinson’s disease diagnosis[J]. Artificial intelligence in medicine, 2019, 95: 48-63.
• Valla E, Nõmm S, Medijainen K, et al. Tremor-related feature engineering for machine learning based Parkinson’s disease diagnostics[J]. Biomedical Signal Processing and Control, 2022, 75: 103551.
• Kamran I, Naz S, Razzak I, et al. Handwriting dynamics assessment using deep neural network for early identification of Parkinson’s disease[J]. Future Generation Computer Systems, 2021, 117: 234-244.
• Gazda M, Hireš M, Drotár P. Multiple-fine-tuned convolutional neural networks for Parkinson’s disease diagnosis from offline handwriting[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2021, 52(1): 78-89.
• Lamba R, Gulati T, Al-Dhlan K A, et al. A systematic approach to diagnose Parkinson’s disease through kinematic features extracted from handwritten drawings[J]. Journal of Reliable Intelligent Environments, 2021, 7(3): 253-262.
• Deharab E D, Ghaderyan P. Graphical representation and variability quantification of handwriting signals: New tools for Parkinson’s disease detection[J]. Biocybernetics and Biomedical Engineering, 2022, 42(1): 158-172.