iHand - Intelligent therapy control in hand rehabilitation

The aim of the "iHand" innovation project is to develop an AI-based assistance system for controlling hand therapy using a rehabilitation robot. Two partners from the AIMECA innovation network - LIME medical GmbH and the Neuromotor and Training Department of the Institute of Sports Science at the University of Münster - are working together on the data-based solution. Thanks to artificial intelligence, the robotic system should be able to react independently to changes in the therapy process and make appropriate adjustments. In the best-case scenario, hand therapy can be carried out largely at home in the future and the success of the therapy can be monitored via an app. The "iHand" project is being funded with a total volume of around 445,000 euros from the Central Innovation Program (ZIM). 

Hand therapy is a form of treatment for restoring and maintaining hand function after operations, injuries, chronic illnesses, osteoarthritis and pain in the hand. In hand therapy, gripping, mobility, coordination, strength and fine motor skills are trained. Supporting traditional physiotherapy with assistance systems such as therapy robots represents a great opportunity - not least to counteract the shortage of skilled workers in Germany and enable more flexibility in hand therapy.

Self-learning therapy robot
By implementing a classification algorithm, a self-learning control and planning software is to be developed within the "iHand" R&D project. By continuously recording sensor data during the use of a therapy robot, the system constantly receives new information about the relationships between therapy settings, therapy progress, the state of health of the hand and the well-being of the patient. The decision for or against an adaptation of the parameters and the extent of the adaptation is made by a knowledge-based clinical decision support system (CDSS). A CDSS is computer software that supports medical professionals in decision-making tasks. The planning, implementation and adaptation of robotic hand-motion therapy is to be automated through the use of decision support methods and, in particular, machine learning (ML) methods. 

The development phase comprises the majority of the R&D project. While the Neuromotor and Training department at the University of Münster identifies suitable indicators for training the classification algorithm and develops the CDSS, LIME medical GmbH is responsible for developing the ML model and the real-time capable control system.In the implementation and validation phase, all components are integrated into a complete system and the final prototype is commissioned and validated. 

The idea for the "iHand" project was developed as part of the innovation network AIMECA - Artificial Intelligence in Medical Care, which is funded by the Central Innovation Program for SMEs (ZIM). As part of the membership, the partners are actively supported in the realization of R&D projects and in securing funding. AIMECA is managed by IWS GmbH, which also handles the application management for cooperation projects and provides intensive support to members in the development of new technologies.

Research Team Münster:
Prof. Dr. Claudia Voelcker-Rehage
Matthias Hendricks
Dr. Julian Rudisch
Lena Kopnarski, M.Sc

Sponsor:
Central Innovation Program Münsterland (ZIM)

Duration:
2023-2025

Contact:
Matthias Hendricks