Real-time approach for gait analysis using the Kinect v2 sensor for clinical assessment purpose

• Alexandre de Queiroz Burle UFPE
• Thiago Buarque de Gusmão Lafayette UFPE
• José Roberto Fonseca UFPE
• Veronica Teichrieb UFPE
• Alana Elza Fontes da Gama UFPE

Resumo

Gait analysis is often performed in clinical environments to evaluate the patient's walk capability and help the rehabilitation process. In a walking test, real-time feedback is fundamental for the health professional evaluate the gait quality, but the high cost of golden standard equipment is a limiting factor. After Microsoft Kinect v2 was released, many studies were published suggesting that this RGB-D sensor could be used for clinical purpose. In this research we analyzed 75 different gait samples with a real-time approach to compare the results with the real measurements of each parameter. Results showed that step length outcomes are accurate and stride length outcomes are precise. Regarding kinematic parameters, knee and hip angles indicate that these joints may be used for gait analysis, but feet angles were not reliable.

Palavras-chave: Kinect v2, Gait Analysis, Real-time, Clinical Environment, RBG-D sensor