Computer-aided discrimination method for Parkinson's disease symptoms based on KINECT bone data

Abstract

The invention discloses a non-contact detection method for Parkinson's disease. The problem of interference with the normal walking of the detected person that may be caused by the contact type is solved, and the hardware cost and the complexity of the device are reduced. The method includes: collecting bone data through kinect; extracting the central point, and performing low-pass filtering on the point coordinate sequence to obtain a new sequence; extracting "the shortest point of the human body in the walking cycle" to calculate the walking cycle; calculating kinematic parameters and stride parameters. One-way analysis of variance was performed on the six groups of parameter sequences, and the experimental data were selected according to the results and box plots to obtain the normal walking data of the tested subjects. A. From the new data, calculate the displacement sequence of the adjacent frames of the left and right feet, and calculate the correlation coefficient to judge the symmetry; if the symmetry conforms to PD, it is suspected PD; B. For the center point movement acceleration sequence, the observation sequence is calculated through the patient's HMM model parameters. probability to determine whether it is a PD patient. If A and B are satisfied at the same time, PD is diagnosed.

Description

technical field [0001] The present application relates to fields such as medical field, mechanics field, anthropometry field, computer vision, etc., and particularly relates to a method for distinguishing Parkinson's disease through depth visual signal monitoring. Background technique [0002] Parkinson's disease (PD) is a degenerative disease of the central nervous system. The clinical manifestations mainly include resting tremor, bradykinesia, muscle rigidity, and posture and gait disturbance. The diagnosis of Parkinson's disease mainly depends on medical history, clinical symptoms and signs. By monitoring the gait signs of the test subject, it can be judged whether the test subject suffers from PD. Specifically, it is judged by monitoring parameters such as kinematic parameters and spatiotemporal parameters in the walking process of the elderly. Monitoring Parkinson's disease can inform patients of the disease's progress in a timely manner, urge patients to seek medical ...

AI Technical Summary

Problems solved by technology

Contact and the combination of contact and non-contact detection of Parkinson's disease may have the problem of interfering with the normal walking posture of the monitored

Images