For their case studies, the research groups used sensors built into clothing that register the body movements of diseased individuals during their normal daily lives. Algorithms process the signals transmitted by the sensors in their overall context. This new AI system is not only capable of identifying the movement patterns characteristic of a neurological disease, which are so small that they remain invisible even to experienced neurologists. It can also determine the stage of a patient's disease and predict with high accuracy in each individual case what further course the disease is likely to take without therapeutic intervention. The sensor-based algorithms function as digital biomarkers that enable precise and continuous monitoring of patients for the first time. In diagnostic terms, these biomarkers are superior to established clinical methods for detecting neurodegenerative diseases: from the onset of a disease to the detection of characteristic symptoms, only about half as much time elapses as when using traditional methods.
The two case studies now published on Friedreich's ataxia and Duchenne muscular dystrophy show that the underlying new technology can in principle be applied to all diseases that cause disorders or changes in movement behavior. In particular, it can provide valuable diagnostic and therapeutic support in diseases for which a gradual or highly variable course is characteristic. "The systematic linking of wearables and artificial intelligence enables medicine for the first time to develop therapy concepts for rare neurodegenerative diseases that are tailored to the individual physical condition of the patients. Once a therapy has been started, our biomarkers can help to monitor its effectiveness and make any necessary adjustments," says Prof. Dr. Aldo Faisal. Important research contributions to the new technology have been made under his leadership at Imperial College in London in collaboration with other British partner institutions. As Professor of Digital Health at the University of Bayreuth, he will continue to develop it at the Campus in Kulmbach in a new "Quantitative Living Lab (QLiLa)" that is currently under construction.
A case study of Friedreich's ataxia: Measuring gene activity based on movement data only
Friedreich's ataxia, named after its discoverer Nicolaus Friedreich (1825-1882), is due to a genetically determined disturbance in the body's production of the protein frataxin. This disorder can damage the nervous system in very different ways. Diagnosis is often complicated in the early stages by the fact that the same or similar symptoms can occur in other neurological disorders. The new biomarkers are able to monitor the genetic control of frataxin production over time. For the first time, it is now possible to measure the activity of genes in humans using only movement data, without taking blood or tissue samples. This enables long-term prognoses that established clinical methods are not capable of. Patients are therefore spared lengthy series of examinations, and the healthcare system is relieved of the corresponding costs. The new AI system thus makes it possible for the first time to develop effective therapies that are precisely tailored to individual patients.
Duchenne muscular dystrophy case study: Timely and precise monitoring of therapeutic measures
Duchenne muscular dystrophy was first described by the physiologist Guillaume-Benjamin Duchenne (1806-1875). It is a genetic muscle disease that begins in early childhood. The life expectancy of patients with the disease is often very limited, especially due to severe impairment of respiratory function. Researchers have now succeeded in developing a biomarker – "KineDMD" – that provides a reliable overall picture of the current motor skills of a person with the disease. The effects of therapeutic measures are recorded promptly and precisely. "Our research results contain numerous starting points for extending this technology to other neurodegenerative diseases, but also to cardiological and orthopedic diseases – including damages to the nervous system caused, for example, by a stroke or a heart attack. These two papers show how far we can get when AI researchers, engineers, life scientists and clinicians closely collaborate in a team," explains Prof. Dr. Aldo Faisal.
Future research at the University of Bayreuth
The Quantitative Living Lab (QLiLa), which Prof. Faisal is setting up in Kulmbach – the location of the Faculty of Life Sciences at the University of Bayreuth – is a unique project worldwide. The focus of the research will be on using AI to solve health-related issues while integrating them into everyday life. The interdisciplinary team led by Prof. Faisal includes researchers from engineering, computer science, behavioral sciences and neuroscience. The common goal is the analysis of human behavior and the development of new technologies that support a long, healthy and independent life. The idea of the "living lab" – also called "real lab" – transfers the scientific and technical laboratory concept to a place of everyday life, specifically to two apartments. Instead of conducting science in abstract experiments in a laboratory or hospital, people will in future be able to be examined and treated in their daily lives using digital procedures.
Contact for scientific information:
Prof. Dr. Aldo Faisal
Digital Health with Data Science in Life Sciences
University of Bayreuth
B. Kadirvelu et al.: A wearable motion capture suit and machine learning predict disease progression in Friedreich’s ataxia. Nature Medicine (2023), DOI: https://dx.doi.org/10.1038/s41591-022-02159-6
V. Ricotti et al.: Wearable full-body motion tracking of activities of daily living predicts disease trajectory in Duchenne muscular dystrophy. Nature Medicine (2023), DOI: https://dx.doi.org/10.1038/s41591-022-02045-1