Parkinson’s disease diagnosis using recurrent neural network based deep learning model by analyzing online handwriting

Kumar, Kaushal; Ghosh, Rajib

doi:10.1007/s11042-023-15811-1

Cited by 11 publications

(1 citation statement)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine learning algorithms are applied increasingly to healthcare datasets. Machine learning and deep learning algorithms have been used for various types of information such as handwritten pattern [11], speech [10], cerebrospinal fluid (CF) [12], and electronic healthcare records for PD subject classification [9]. Using the handwritten pattern for discrimination healthy against PD subjects is popular.…”

Section: Related Workmentioning

confidence: 99%

Late fused multi-modal neural network with votingclassifier for Parkinson’s Disease detection

aljohani

2024

Preprint

View full text Add to dashboard Cite

Parkinson’s Disease (PD) is categorized as a neurodegenerative progressivedisease caused by the destruction of the cells in the midbrain posterior.Detecting PD in its early stages will help physicians alleviate the complications of the disease. Artificial Intelligence (AI) is considered as the groupof trained models that can be used for classification and regression. Differentmodalities such as text, speech, and picture, can be used for detecting PD.This research proposes a multi-modal deep learning recognition technique forPD classification. To improve the quality of PD detection in the early stages,the proposed method is composed of three main sections. These sections are:feature extracting, merging, and classifying. As feature extractors a combination of Convolutional Neural Network (CNN) and attention mechanisms isdeveloped. To extract features from related motion signals a combination ofCNN and Long-Short Term Memory (LSTM) model is used. Finally, Random Forest (RF), Logistic Regression (LR), Support Vector Machine (SVM),Extreme Boot Classifier (XGB), and voting classifier are used to distinguishbetween healthy and PD subjects. The experimental result indicates 99.95%accuracy, 99.99% precision, 99.98% sensitivity, and 99.95% F1-score usingthe proposed CNN with attention and voting classifier on PD handwritingand corresponding motion datasets. The achieved results show the proposedmethod of extracting features from both handwriting pictures and correlatedmotor symptoms followed by fusing the features and finally using voting asthe classifier can achieve perfect performance for PD classification

show abstract

Section: Related Workmentioning

confidence: 99%