Mobile health system to assist Parkinson’s patients


Changes in the regulation of vegetative bodily functions and sleep frequently occur even before Parkinson’s patients experience movement disorders. However, a broadly available screening and monitoring system for corresponding early diagnosis does not exist at this time. The project closes a gap in the screening and monitoring of Parkinson’s disease. Benefits include early diagnosis, diagnostic accuracy, improved treatment quality, and most of all a better quality of life for Parkinson’s patients.


The challenge

A body-hugging sensor is being developed, for example to wear on the wrist. The data can be evaluated with an app that is intuitive to use. Predictive algorithms are being developed for this purpose in the sense of a digital bio-marker, enabling automated data evaluation and intelligent data management. The system is supported by telemedical video monitoring. To ensure the best possible use of the subsequent system by patients, a user evaluation is being carried out in the project in order to optimize ergonomics and the human-machine interface.


Our contribution

Fraunhofer ISST is researching and developing the digital bio-marker for the prediction of conspicuous neurodegenerative parameters. Methods from the field of machine learning are being used to provide support with early diagnosis and in predicting the course of the illness. Data about sleep and the neuro-vegetative state of patients are collected using mobile sensors (polysomnography). These are pre-processed and analyzed for patterns in order to identify at-risk patients and construct a prediction model that can be used as predictive support by doctors in evaluating patients. A study is being conducted in the course of the project, serving as a data source in addition to the retrospective data from the Aachen University Hospital. The sleep data are pre-processed by manually annotating and windowing the activities during the REM phase in the raw signals. Characteristics are generated and standardized in order to train a model using classification methods such as support vector machines, k-nearest neighbor and decision trees. A k-fold cross-validation is used for validation.



The project objective is to develop the first patient-friendly, mobile screening and monitoring system for the early diagnosis of Parkinson’s disease. Here the focus in on the early diagnosis of REM sleep disturbances and other underlying changes of body functions such as slower movements or postural instability..



  • Aachen University Hospital
  • Fraunhofer ISST, Dortmund
  • SOMNOmedics GmbH, Randersacker
  • MVB Parkinson GmbH, Koblenz
  • RWTH Aachen University



  • Subsidized by: Federal Ministry of Education and Research (BMBF)
  • Project number: 16SV7857
  • Term: September 2017 to August 2020