Intro: What if we could anticipate falls during sport, at work, or in everyday life and thus avoid injuries? Here, test subjects stumble for all their worth in order to achieve this goal. Numerous measuring systems and artificial intelligence are used to detect early signs of falls. The new approach of this project is to use gait analysis as a preventative measure in the new laboratory for AI-based gait analysis, the Smart Gait Lab at Koblenz University of Applied Sciences. Here, we meet researchers who are getting to the bottom of the characteristics of human gait from different angles.
Prof. Lukas Scheef: You have to realize that medicine is moving more and more in the direction of prevention and running, walking, and all movements are very complex processes in the body, And even the smallest changes often indicate illnesses. Yes, let's take a look at running. Some runners don't run optimally. They have incorrect movement sequences. This then leads relatively quickly to overloading of the joints, ligaments and muscles. It would be much, much better if you could recognize these incorrect movements early on and then simply change your running style. Let's take a look at geriatrics. We know older people become increasingly wobbly in their gait over time. Let's Look at how pronounced this insecurity is. Are there any near-falls, and we can also look at this: Where does this happen?
Voice-over: The team wants to achieve these goals with the help of artificial intelligence, which can recognize conspicuous patterns from a large amount of movement data, sometimes even in real time. Tests and training sessions take place in the tripping course and on the treadmill. Dr. Annika Weber perhaps has the most fun of all because she literally gets to watch her test subjects who are of course well secured, stumble, and learn how to catch themselves again. She and her team measure many parameters in the process.
Prof. Anika Weber: We can then use the force plates to measure the forces acting on the body and then draw conclusions about the forces acting within the body. In our working group, we also do a lot of research into how this affects the muscles and tendons, particularly in the lower extremities. We have developed this gait course ourselves, and we can generate tripping, slipping, and miss-stepping events. These are very unpredictable for the participants who then walk across this course.
Voice-over: Occupational health and safety is a major area of application for the project.
Prof. Ulrich Hartmann: Last year we conducted a study with a major German parcel delivery company in the field of occupational safety. It was about the prevention of trip and fall accidents, and there was a training course that was VR based. In other words, you wore VR glasses and then walked on the treadmill and suddenly the horizon moved and you had to get to grips with that. The other training is mechanical. The employees were on the treadmill and their legs were plucked as they walked, and that's how you get tripped up. They get inoculated by this.
Voice-over: The project is working on making gait measurement systems available directly to people in everyday life, whether as wearables or simply via a smartphone. The first steps have already been taken with a self-developed algorithm that requires very little information.
Prof. Babette Dellen: In the AI based gait analysis sub-project, we use acceleration data recorded with a smartphone. We use the data to learn a mathematical model for a person's normal gait. We then identify anomalies as deviations from the normal gait pattern. The aim is to run the process in real time so that we can then transfer it to smartphones and smart watches.
Voice-over: The prevention of falls is desirable from an economic point of view alone, saving costs through prevention instead of rehabilitation. But above all, it's about providing people with better support.
Prof. Lukas Scheef: As a medical professional and also as a person with an affinity for sport, my dream would be for us to develop sensor systems in order to capture people in their entirety in the long term, the biological aspect of people, so to speak. We could record laboratory values and movement data virtually 24/7 in order to gain a better understanding of the development of diseases and to be able to work preventatively in the long term and also to continue to individualize therapy.
