|
A review of machine learning-based audio signal analysis methods in diagnosing Parkinson's disease
|
| |
|
|
|
Abstract: (18 Views) |
Parkinson's disease is known as a progressive neurodegenerative disease. Different areas of the human brain are damaged by this disease and there is still no definitive cure for it. Older people suffer more from this disease and statistics show that the number of people affected is increasing. Early diagnosis and assessment of the presence or severity of Parkinson's disease are very important in controlling the progression of the disease. Nowadays, real-time and non-invasive methods based on audio signal analysis, using machine learning methods, have received attention. Machine learning methods are usually characterized by a transparent pipeline that makes the results highly interpretable. In particular, acoustic features are used in many machine learning methods and can also act as indicators of the general condition of people's voices. The aim of this review is to identify the most applicable and promising machine learning methods in the diagnosis of the disease. This paper reviews the evaluation to identify the most commonly used and effective features in early detection of Parkinson's, as well as algorithms, free datasets, and practical toolkits in this field.
|
|
| Keywords: speech processing, speech analysis, feature extraction, machine learning, Parkinson's disease. |
|
|
Full-Text [PDF 550 kb]
(27 Downloads)
|
Type of Study: Scientific-extension |
Subject:
Special
Received: 2026/05/10 | Accepted: 2026/03/16 | Published: 2026/03/16
|
|
|
|
|
|
|
| Add your comments about this article |
|
|
|
