Detection of Parkinson's Disease Using Voice Changes and Hand-tremor

Parkinson's (PD) is a degenerative, progressive condition that disturbs nerve cells in the basal ganglia and the locus niger, which are deep areas of the brain. The neurotransmitter dopamine is produced by nerve cells in the complicated body part, and it is responsible for conveying messages that plan and govern bodily action. In certain people, the dopamine-producing nerve cells in the nucleus niger begin to migrate away. PD symptoms such as trembling, weak muscles, stiffness, and balance issues appear after 80% of dopamine is eliminated. Detection of PD from rapidly given cues, like as speech and handwriting, is crucial to various social behaviors, including as fast imitation and empathy. With technological improvements and, as a result, the prominence of audio collecting devices in everyday life, dependable models that transform audio data into a screening aid for healthcare providers might possibly deliver diagnoses that are both cheaper and more accurate. This experiment uses two modalities to identify Parkinson's disease. Voice and handwriting are the two modalities. Audio processing and machine learning methods in Matlab are used to identify PD via speech. The characteristics are derived from the audio of people with Parkinson's disease who have a local binary pattern. For classification, machine learning techniques are utilized. Machine learning classifiers are used in the classification process using neural networks, and embedded systems are used to identify PD via handwriting. The microcontroller is used to see the disease with the help of a gyroscope sensor. The principle of angle measurement is determined by calculating the angle of free vibration of the hanging mass with respect to the gyroscope casing. As a result, if the vibration passed the edge, the person is suffering from this disease.