Machine Learning-Based Framework for Early Detection of Distinguishing Different Stages of Parkinson’s Disease

Main Article Content

Archana Panda, Dr Prachet Bhuyan

Abstract

Parkinson’s disease (PD) is caused by a disruption in the brain cells that produce dopamine, a substance that allows brain cells to communicate with one another. Dopamine-producing cells in the brain are in charge of movement control, adaptation, and fluency. When 60–80% of these cells are lost, there is insufficient dopamine production, and Parkinson's motor symptoms appear. It is believed that the disease begins many years before the movement-related symptoms appear, so researchers are looking for ways to identify the non-movement symptoms that appear early in the disease as early as possible, thereby halting the disease's progression. Accurately detecting Parkinson’s disease during an early stage is unquestionably critical for delaying its steady progress and providing patients with access to disease-modifying therapy. To that end, the premotor stage of Parkinson’s disease should be closely monitored. Based on the results of the test, a technique is introduced to determine whether an individual has Parkinson's disease or not. Premotor characteristics Particularly, several indicators have been considered to detect Parkinson’s disease at an early stage. A comparison of the proposed different Machine Learning models will be used based on relatively small data sets of healthy individuals and early Parkinson’s disease patients reveals. The designed model should achieve the highest accuracy on average. In this study, we provided the feature importance of the Parkinson's disease (PD) detection process based on the Machine Learning Process. Decision Trees, Random Forest, Neural Networks, Deep Learning, Gradient Boosted Tree, and Support Vector Machines, algorithms were used to classify Parkinson’s patients. These algorithm’s feature elimination outperformed all other methods. With the fewest number of voice features, 87.18% accuracy was achieved for Parkinson’s diagnosis. We find that these techniques perform well in classifying early Parkinson’s disease and healthy normal people, with high accuracy. The analysis of non-invasive biological markers for disease detection is critical for accurate clinical diagnosis. As a result, the analysis can be used to detect Parkinson’s disease at an early stage.

Article Details

Section
Articles