Parkinson's disease (PD) is a progressive neurodegenerative disease primarily affecting elderly and is considered the second most common neurodegenerative disease after Alzheimer's disease. It is mainly characterized by motor and non-motor characteristics affecting the patient's movement, gait, balance and swallowing. Current research methods lack a comprehensive understanding of PD progression, as multiple clinical and non-clinical factors are involved in disease manifestation and progression, which often leads to heterogenicity. Parkinson's disease presents manifestations that are much more complex to predict than just motor/non-motor symptoms. One of the main concerns in predicting early-stage PD is that PD symptoms overlap with those of other diseases such as multiple sclerosis and Alzheimer's disease. Say no to plagiarism. Get a tailor-made essay on “Why violent video games should not be banned”?Get the original essayThe Unified Parkinson's Disease Rating Scale (UPDRS) is a widely used clinical symptom rating scale for PD and serves as a basic assessment tool to detect the presence and assess its severity. Although current medical approaches can alleviate symptoms, there is no standard therapy to cure PD. Therefore, early diagnosis is of crucial importance to predict disease progression to help patients survive and improve their quality of life. Additionally, monitoring the progression of Parkinson's disease is often cumbersome because it involves a patient-directed care management plan in which disease symptoms and UPDRS scores were determined clinically every 3, 6 or 12 months. Since most of the affected patients were elderly and the methods are time-consuming, this is not logistically and financially feasible for both doctors and patients. Additionally, most current methods used to track PD rely on expert clinical raters, from whom assessment of PD symptoms can be difficult due to interindividual variability. Therefore, for all these reasons, periodic remote monitoring of PD progression and UPDRS scores was implemented. has emerged as an alternative solution for monitoring Parkinson's patients with low-cost, non-invasive methods. Recognizing the importance of remote patient monitoring in PD, healthcare providers have developed and implemented several novel organizational-level approaches to improve remote monitoring of disease progression in PD patients. In addition to motor and non-motor symptoms, speech disorders are prevalent in 70 to 90% of PD subjects. Therefore, voice recordings would be useful in identifying PD and tracking disease progression. Additionally, subjects with PD exhibit characteristic voice and speech patterns, which could serve as potential indicators for early detection. However, it is difficult for a physician to manually characterize these speech patterns and voice recordings for the same patient or multiple patients over months. Machine learning (ML) methods programmed into wearable devices have proven to be a promising technology for automatically evaluating scoresseverity of illness using the UPDRS scale. These methods apply statistical or mathematical algorithms on the input data to draw arbitrary patterns, common structures, or data points in the data set to make predictions for new input data (outcome). In this context, a systematic review of the available literature is warranted to understand the development, application and validity of various machine learning algorithms in order to perform comparative analysis that helps clinicians predict disease progression in MP and designers to design wearable devices. for remote monitoring of PD patients. Although there is no specific test to diagnose PD, traditional diagnosis involves a neurologist reviewing the brain's medical history and assessing the subject's motor skills through various methods, knowing that traditional methods are prone to variability. interindividual. Another problem is that early symptoms of Parkinson's disease often overlap with symptoms of other diseases such as Alzheimer's disease, multiple sclerosis, Huntington's disease and dementia with Lewy bodies, leading to diagnostic errors. Due to the lack of laboratory tests or standard methods to diagnose PD, early diagnosis of PD has become difficult when most motor symptoms are not severe in the early stages of PD, requiring regular monitoring of PD. motor symptoms in clinical settings. PD often occurs in older people. adults, and constant monitoring of disease progression is justified by regular visits to the clinic. Remote patient monitoring is receiving increasing attention to track disease progression using various non-invasive methods such as monitoring speech patterns in PD subjects. Since speech disorders are prevalent in 70–90% of PD subjects, voice recordings would be useful in identifying PD and tracking disease progression. Additionally, subjects with PD exhibit characteristic voice and speech patterns, which may serve as potential indicators for early detection, low-cost, non-invasive and time-consuming diagnostic tools for PD. The reason behind the increased attention to PD diagnosis using speech models is mainly due to the rapid development of telediagnosis and telemonitoring in the medical field. Additionally, these methods are less expensive and the devices are often easy for subjects to self-monitor, reducing patient visits to clinics, allowing them to monitor the progression of their disease on their own. Although medications and surgeries can control the progression of Parkinson's disease by alleviating motor symptoms, there is no method to cure Parkinson's disease. Current research methods lack a comprehensive understanding of PD progression, as multiple clinical and non-clinical factors are involved in disease manifestation and progression, often leading to heterogenicity. Therefore, early diagnosis is of crucial importance to predict disease progression and help patients improve their quality of life and survive. The Unified Parkinson's Disease Rating Scale (UPDRS) is the universal and widely used clinical rating scale to assess the clinical spectrum of PD and serves as the baseline assessment for PD. Understanding the relationship between UPDRS scores andPatient voice signal characteristics have been widely studied to predict early-stage PD. However, clinicians cannot use UPDRS scores to manually evaluate and score patient voice recordings in a large dataset. Since patient voice recordings usually occupy a large space, it is almost impossible for clinicians to evaluate them manually as it is time-consuming. Machine learning systems have been shown to be a promising technology to automatically assess disease severity scores using the UPDRS scale. Therefore, computer-aided systems using machine learning algorithms are developed to objectively detect and monitor disease progression. Most studies used advanced machine learning algorithms to extract the relevant or most significant features (feature extraction) from the database that contributes to PD (UPDRS scores). Speech and non-speech segments were extracted from the model constructed using the Gaussian Mixture Model Universal Background Model (GMM-UBM) using the support vector regression algorithm. The model predicted PD with a Pearson correlation of 0.60 for MDS-UPDRS scores. Remote monitoring of Parkinson's disease progression was carried out using regression methods such as support vector machines (SVM), least squares support vector machines (LS-SVM), the multi-layer perceptron neural network (MLPNN) and the general regression neural network (GRNN) to predict the observed evolution. UPDRS scores. The results indicated that LS-SVM outperforms all other regression methods tested for the dataset. Minimum redundancy, maximum relevance feature selection algorithm tested on speech PD signals resulted in 90.3% accuracy and 90.2% accuracy and Mathews correlation value of 0.73 in using a random forest model. This finding showed that simple random forest was better than other methods such as bagging, boosting, SVM, and decision tree methods. To support incremental data updates, an incremental support vector regression (ISVR) approach was implemented to predict UPDRS scores. The study confirms that self-organizing map (SOM), non-linear iterative partial least squares (NIPALS) and ISVR techniques are effective in predicting total UPDRS and motor UPDRS (Nilashi et al. 2018). . Although clustering was not the primary objective of this study, advanced methods such as principal component analysis (PCA) and expectation maximization (EM) were used to cluster multicollinear PD speech data. Novel regression techniques such as adaptive neuro-fuzzy inference system (ANFIS) and SVR have been carried out to predict PD progression, which has remarkably improved the accuracy of PD prediction. However, this model was limited by the number of training samples in each cluster constructed by the PCA and EM algorithms. UPDRS evaluation for remote monitoring data was tested using linear and nonlinear regression techniques such as least squares (LS), iteratively reweighted LS (IRLS), and shrinkage and selection operator the least absolute (LASSO). Computational approaches such as neural networks could develop more accurate models for disease prediction. Artificial neural networks (ANN) and ANFIS were used to predict PD using voice data. The advantages of.