Parkinson’s disease diagnosis using convolutional neural networks and figure-copying tasks

Abstract

Parkinson’s disease (PD) is a progressive neurodegenerative disorder that causes abnormal movements and an array of other symptoms. An accurate PD diagnosis can be a challenging task as the signs and symptoms, particularly at an early stage, can be similar to other medical conditions or the physiological changes of normal ageing. This work aims to contribute to the PD diagnosis process by using a convolutional neural network, a type of deep neural network architecture, to differentiate between healthy controls and PD patients. Our approach focuses on discovering deviations in patient’s movements with the use of drawing tasks. In addition, this work explores which of two drawing tasks, wire cube or spiral pentagon, are more effective in the discrimination process. With $$93.5\%$$ 93.5 % accuracy, our convolutional classifier, trained with images of the pentagon drawing task and augmentation techniques, can be used as an objective method to discriminate PD from healthy controls. Our compact model has the potential to be developed into an offline real-time automated single-task diagnostic tool, which can be easily deployed within a clinical setting.

A review of machine learning and deep learning algorithms for Parkinson's disease detection using handwriting and voice datasets

“…Imbalanced datasets, as mentioned by Alissa et al (2022) [ 82 ], pose a significant challenge in PD detection. Imbalanced datasets can lead to biased model predictions, as they favor the majority class.…”

“…Alissa et al (2022) [ 82 ] explored the use of DL, RNN, and CNN to dif-ferentiate between healthy individuals and PD patients, considering multiple datasets, including imaging and movement data. They aimed to determine which PD test, whether imaging or time series data, is more effective for diagnosis.…”

Diagnosing of Parkinson's disease based on hand drawing analysis using Bi‐Directional LSTM equipped with fuzzy inferential soft‐max classifier

“…In this regard, data mining/deep learning methods and algorithms can be useful in extracting meaningful information from data and identifying patients from healthy people. [18][19][20] This paper will propose an efficient deep learning architecture for diagnosing Parkinson's patients from normal ones. We tried to automatically review the hand drawing images of people who took the Parkinson's test using digital pen devices (as illustrated in Figure 1).…”

“…This way, patient data can be distinguished from healthy individuals' data. In this regard, data mining/deep learning methods and algorithms can be useful in extracting meaningful information from data and identifying patients from healthy people 18–20 …”

COVID-19 Detection on X-Ray Images using a Combining Mechanism of Pre-trained CNNs

“…In reality, the two subsets of AI are employed to health data analysis: the first subset is Machine Learning (ML) and the second is Deep Learning (DL) approaches, including radiography images or computed tomography scans, have been shown to be useful on detection of illness and monitoring [12]- [14], [15]- [17]. As a result, various types of human maladies, like as Parkinson's disease [18]- [21], brain tumor segmentation [22], [23], breast cancer [24], diabetes [25], medical image segmentation [26], and heart disease prediction [27]- [30], atherosclerosis diseases [31], could be identified using such techniques. AI advancements have also contributed in the development of a wide range of other scientific fields [32]- [34], [35]- [39].…”