研究目的
Investigating the feasibility of applying wavelet techniques to gait signals for automatically classifying the stance and swing phases in Parkinson’s Disease (PD) patients and healthy subjects using a Kinect camera.
研究成果
The study demonstrated that wavelet decomposition can automatically detect walking phases and analyze gaits using a Kinect RGB-D sensor with 93% accuracy compared to clinical judgment. This approach offers a fast, low-complexity method for gait analysis in PD patients, supporting its use in telemedicine contexts.
研究不足
The study was conducted on a small sample size (12 volunteers). The accuracy of gait phase classification might vary with more diverse or larger populations. The method requires further validation against other motion-capture systems.
1：Experimental Design and Method Selection:
The study used a Kinect RGB-D sensor for capturing gait signals and applied wavelet-based digital signal processing to classify gait phases automatically.
2：Sample Selection and Data Sources:
Twelve volunteers participated, including six PD patients and six healthy subjects. Gait signals were captured in a corridor using the Kinect sensor.
3：List of Experimental Equipment and Materials:
Microsoft’s Kinect sensor and MATLAB for signal processing.
4：Experimental Procedures and Operational Workflow:
Volunteers walked towards the Kinect three times. The Kinect’s skeleton frame was used to represent joint locations, focusing on ankle joints. Wavelet transforms were applied to the signals to classify gait phases.
5：Data Analysis Methods:
The accuracy of wavelet decompositions was assessed by comparing automatically classified gait phases with expert clinical judgment using the Hamming distance metric.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
