Deeper insights into the speed of Parkinson’s disease
Researchers are increasingly distinguishing between two types of progression in Parkinson’s disease: a rapidly progressing type and a slowly progressing type. This finding is also based on a study conducted with the support of artificial intelligence (AI) at the Technical University of Dresden and presented at the 96th Congress of the German Neurological Society in Berlin(link). Such differences in the progression of the disease can complicate treatment and require different therapeutic approaches for different patient groups. The study results can also be read in detail here(page 189 / abstract 689).
AI in Parkinson’s research
The study used longitudinal data from Parkinson’s patients to identify the different Parkinson’s progression types. These data came from three large Parkinson’s cohorts: PPMI, ICEBERG and LuxPARK. By synchronising the patient data on a uniform time scale of disease progression and using AI technologies, it was possible to identify two different types of progression in all three Parkinson’s cohorts. This AI model, which was first trained in the PPMI cohort and then applied to the other cohorts, showed differences in motor and non-motor symptoms, survival rates, response to medication, DaTSCAN imaging and digital biomarkers of gait assessment between the two progression types.
Optimisation of Parkinson’s research
The research results have far-reaching implications for clinical trials by showing that adapting the study population to rapidly progressing patient types could significantly reduce the required size of the study group. This means that predictive models could enable clinical trials with fewer participants in the future by specifically including rapidly progressing patients. Simulations showed that increasing the proportion of rapidly progressing patients based on the predictive models can reduce the required cohort size of clinical trials by around 43%
Accelerating Parkinson’s research through global data donations
These advances in Parkinson’s research emphasise the importance of databases such as the kill parkinson initiative. By collecting and donating anonymised health data from Parkinson’s patients worldwide, researchers can access a much broader database. The integration of this data into AI-supported analysis tools can significantly accelerate basic research into Parkinson’s disease. Access to such extensive data and the application of advanced analytical methods and AI offer the opportunity to gain a deeper understanding of the disease and ultimately lead to more effective treatment strategies and ideally, one day, victory over Parkinson’s.
Authors and sources
Tom Hähnel, Tamara Raschka, Björn Falkenburger, Holger Fröhlich
Technische Universität Dresden, Klinik und Poliklinik für Neurologie, Dresden, Germany; Fraunhofer Institute for Algorithms and Scientific Computing (SCAI), Bioinformatics, Sankt Augustin, Germany; University of Bonn, Bonn-Aachen International Center for IT, Bonn, Germany; German Centre for Neurodegenerative Diseases (DZNE), Dresden, Germany