Diagnostics and treatment of the musculoskeletal system using Kinex is based on artificial intelligence, machine learning, and computer vision.
Each video of a user performing a diagnostic exercise is analyzed frame by frame using a pretrained high-precision neural network. The data from the neural network output characterize the individual's physical features and form their digital twin. The obtained digital twin of the athlete is stored in the decision support system database.
The product is developed using Flutter to make the service available on all mobile platforms for end users. The technology stack includes Postgresql for data management, PHP and Phalcon for server-side development, Python for machine learning algorithms and Dart for mobile application development.
Computer vision algorithms recognize over 50 key points of the musculoskeletal system and analyze over 600 body parameters. This allows for a digital representation of the individual's physical features, comparison with reference values in the system, and detection of deviations from the norm in the functioning of the musculoskeletal system components.
Based on machine learning models and data analysis, the system generates training plans for the user, recommends exercise routines, tracks exercise execution, and analyzes the dynamics of changes in the user's digital twin. The results are stored in the database for subsequent use and display in the system interface.
The product is developed using Flutter to ensure service accessibility on all mobile platforms for end users.