An improved CNN for rehabilating parkinson’s freezing of gait

Michelle, Tang (2023) An improved CNN for rehabilating parkinson’s freezing of gait. Project Report. Melaka, Malaysia, Universiti Teknikal Malaysia Melaka. (Submitted)

	Text (24 Pages) An improved CNN for rehabilating parkinson’s freezing of gait.pdf - Submitted Version Download (376kB)
	Text (Full Text) An improved CNN for rehabilating parkinson’s freezing of gait.pdf - Submitted Version Restricted to Registered users only Download (2MB)

Abstract

As the freezing of gait symptoms worsen in a patient with Parkinson's disease, the motion of the extremities becomes increasingly sluggish, one of the most noticeable changes in the freezing of gait symptoms. The symptoms of freezing of gait are currently incurable, and the only method for delaying the progression of the disease is training. The cumulative number of falls rises, resulting in psychological and physical injuries. In addition, the majority of Parkinson's patients who visit the facility for training must make an appointment in advance because therapists can only accept one patient at a time, resulting in inefficiency. This results in a decline in training, which accelerates their condition, while freezing of gait causes their extremities to become increasingly sluggish. Most patients' family members frequently have their occupations and lack patient care training. To address these problems and develop this system, it must be able to predict patient falls, alert, and evaluate the patient's posture during training. During training, this system can assist the therapist in monitoring the patient's walking posture. A warning can be issued immediately when a fall is imminent, and the patient's posture can be evaluated for the therapist's reference. This would reduce therapists' burden and allow them to simultaneously treat multiple patients. This will enhance efficiency. This can reduce the incidence of falls and injuries caused by falls among Parkinson's patients. In addition, the system can be installed in the patient's residence to assist the patient's family members in monitoring and training the patient at home, which can significantly reduce the incidence of falls and monitor and evaluate the patient. In this study, CNN is used to educate the system to predict falls with an accuracy of 95%. This system is anticipated to be able to assist therapists who treat Parkinson's patients in the hospital as well as non-therapist family members.

Actions (login required)

View Item