An IoT Integrated Machine Learning Approach for Consistent Monitoring and Treatment of Parkinson’s Disease

Parkinson’s disease is a long degenerative disorder which affects the central nervous system. About 2% of the world’s population more than 65 years get affected by this ailment [1]. As the aged population increases, the number of patients affected by Parkinsons’s disease is likely to increase. In the current medical practice, assessment and treatment of Parkinson’s disease is done based upon the clinical parameters that are examined during patient’s periodic consultation visit and symptoms recorded on home diaries. Data available in the diaries may not be always appropriate and correct as they depend on the memory of the patient and perception of symptoms of his own. Costs spent on transportation of the patient, the consultation fees of the clinician and the expenditure for hospital infrastructure seem to burden the patient. An automatic and procedure at – home to devise methods to detect and quantify tremors, bradykinesia and postural instability can pave the way for quality treatment at affordable costs. The emerging technologies of wearable sensors and Internet-of – Things, can be used for continous monitoring of PD under indoor environments. However, the development of such a system needs (i) a real-time data acquisition module for continous monitoring and (ii) an appropriate technique for the analysis of data to calculate the severity of the PD and provide appropriate treatment. Machine learning algorithms have been applied extensively to Parkinson’s symptoms [24]. However, they are mostly based on general datasets taken from global databases and individuals’ historical data have not been considered. Thus this paper aims to develop an IoT Integrated Machine Learning Approach for Consistent Monitoring and Treatment of PD.