A number of pathologies impact on the way a patient can either move or control the movements of the body. Traumas, articulation arthritis or generic orthopedic disease affect the way a person can walk or perform everyday movements; brain or spine issues can lead to a complete or partial impairment, affecting both muscular response and sensitivity. Each of these disorder shares the need of assessing patient’s condition while doing specific tests and exercises or accomplishing everyday life tasks. Moreover, also high-level sport activity may be worth using digital tools to acquire physical performances to be improved. The assessment can be done for several purpose, such as creating a custom physical rehabilitation plan, monitoring improvements or worsening over time, correcting wrong postures or bad habits and, in the sportive domain to optimize effectiveness of gestures or related energy consumption.
The paper shows the use of low-cost motion capture techniques to acquire human motion, the transfer of motion data to a digital human model and the extraction of desired information according to each specific medical or sportive purpose. We adopted the well-known and widespread Mocap technology implemented by Microsoft Kinect devices and we used iPisoft tools to perform acquisition and the preliminary data elaboration on the virtual skeleton of the patient.
The focus of the paper is on the working method that can be generalized to be adopted in any medical, rehabilitative or sportive condition in which the analysis of the motion is crucial. The acquisition scene can be optimized in terms of size and shape of the working volume and in the number and positioning of sensors. However, the most important and decisive phase consist in the knowledge acquisition and management. For each application and even for each single exercise or tasks a set of evaluation rules and thresholds must be extracted from literature or, more often, directly form experienced personnel. This operation is generally time consuming and require further iterations to be refined, but it is the core to generate an effective metric and to correctly assess patients and athletes performances. Once rules are defined, proper algorithms are defined and implemented to automatically extract only the relevant data in specific time frames to calculate performance indexes. At last, a report is generated according to final user requests and skills.