• algorithms
• motor patterns
• parkinson

The diagnosis of Parkinson's disease through neurological evaluation considers the hesitations during the execution of the motor tasks determinant to assign a judgment of severity. The experience of neurologists in daily clinical practice has shown that hesitation can also be found in healthy subjects. However, there is no universally recognized convention among neurologists to identify and quantify hesitations. This thesis aims to introduce both ways of trying to quantify hesitations and broaden the horizon on why to look for these parameters and enrich the collection of objective characteristics of motor patterns on both healthy and patient with Parkinson’s disease at all stages of pathology.
A motor exercise agreed with the neurologist and a traditional exercise for evaluation were administered to the cohort of recruits. The participants wore a high precision wearable system during the execution. The exercises were recorded on video and the signals of the wearable device were extracted. An algorithm has been developed to identify movement anomalies and to objectively characterize the motor pattern. Hesitations were detected for each test and placed in the time duration of each exercise. Finally, the outputs of the algorithm were compared with the video visual inspection. The study showed that evaluation with the naked eye of the neurologist must be supported by a precise instrument to make the analysis of the movement objective and automatic. Further researches are needed to gather normative values of traditional and new parameters and build a decision support system for the neurologist in the motor assessment of PD.