Best Paper Award for Research on Movement Analysis in Rehabilitation
The team of Assoc. prof. Jan Sedmidubsky from the Faculty of Informatics of Masaryk University in collaboration with VisionCraft was successful at the 17th International Conference on Similarity Search and Applications (SISAP), which took place 4-6 November 2024 in Providence, USA. They bring home the Best Paper Award for a publication introducing new approaches for analyzing rehabilitation exercises that can significantly improve the use of telemedicine in home care. For the study's first author, Ph.D. student Andrej Cernek, this is the first ever scientific publication. Congratulations!
The article titled REHAB24-6: Physical Therapy Dataset for Analyzing Pose Estimation Methods focuses on the issue of telemedicine in rehabilitation. The aim is to allow patients to exercise at home in front of a camera and get feedback based on automated motion analysis. The author team presents a new dataset with recorded movements of typical rehabilitation exercises and an approach to evaluate the quality of existing methods that detect significant points on the human body from conventional video. Indeed, accurate detection of such points is important for subsequent analysis. Presentation of the results at the SISAP 2024 conference in the USA brought recognition in the form of a Best Paper Award.
"The interest not only in the scientific community and reviewers indicates the potential for further research," said Assoc. Sedmidubský from the FI MU Department of Machine Learning and Data Processing, who collaborated on the publication with Ph.D. student Andrej Černek and FI MU graduate Petra Budíková, representing the company VisionCraft.
The resulting article is the first scientific publication by Andrej Černek, who is listed as the first author. "I am pleasantly surprised to have achieved such success at my first conference. I appreciate all the more the advice and guidance of both my supervisor and my colleagues from the Data Oriented Systems and Applications Laboratory (DISA), who helped me polish my presentation," says the student, who has been working with Assoc. prof. Sedmidubský since his master's thesis, for which he won the Dean's Award. In addition, when the opportunity arose to work on his PhD research on motion estimation from video data in a practical application for rehabilitation purposes, he did not hesitate. "Thanks also go to Petra Budíková and Miriam Jánošová, with whom we are collaborating on the project," he adds.
The research is part of the project VisioTherapy: Supporting physiotherapy treatment using computerised motion analysis, supported by the Technology Agency of the Czech Republic. The aim is to develop VisioTherapy software to support patients and physiotherapists during home treatment. Using AI, it analyses movements from videos, compares the patient's exercises with the correct pattern and highlights errors. At the same time, it will track the progress of therapy and provide statistics for better therapy planning. The research team is looking forward to completing the prototype of the rehabilitation app in the near future, especially expanding it to a larger number of exercises. In the long term, there are also opportunities for further collaboration with industry, specifically, for example, in finding ways to improve the detection of significant points of the human body from conventional video.
Congratulations to the whole team for a successful conference and we look forward to seeing what happens next!
Author: Marta Vrlová, Office for External Relations and Partnerships at FI MU
Photo: Jan Sedmidubský
- Attachments
