Clinical Decision Support Systems with Game-based Environments, Monitoring Symptoms of Parkinson’s Disease with Exergames

• Main Author: Garcia-Agundez Garcia, Augusto
• Format: Others
• Language: en
• Published: 2020
• Online Access: https://tuprints.ulb.tu-darmstadt.de/14137/2/2020-10-12_Garcia-Agundez_Garcia_Augusto.pdf; Garcia-Agundez Garcia, Augusto <http://tuprints.ulb.tu-darmstadt.de/view/person/Garcia-Agundez_Garcia=3AAugusto=3A=3A.html> (2020): Clinical Decision Support Systems with Game-based Environments, Monitoring Symptoms of Parkinson’s Disease with Exergames.Darmstadt, Technische Universität Darmstadt, DOI: 10.25534/tuprints-00014137 <https://doi.org/10.25534/tuprints-00014137>, [Ph.D. Thesis]

Parkinson’s Disease (PD) is a malady caused by progressive neuronal degeneration, deriving in several physical and cognitive symptoms that worsen with time. Like many other chronic diseases, it requires constant monitoring to perform medication and therapeutic adjustments. This is due to the significant variability in PD symptomatology and progress between patients. At the moment, this monitoring requires substantial participation from caregivers and numerous clinic visits. Personal diaries and questionnaires are used as data sources for medication and therapeutic adjustments. The subjectivity in these data sources leads to suboptimal clinical decisions. Therefore, more objective data sources are required to better monitor the progress of individual PD patients. A potential contribution towards more objective monitoring of PD is clinical decision support systems. These systems employ sensors and classification techniques to provide caregivers with objective information for their decision-making. This leads to more objective assessments of patient improvement or deterioration, resulting in better adjusted medication and therapeutic plans. Hereby, the need to encourage patients to actively and regularly provide data for remote monitoring remains a significant challenge. To address this challenge, the goal of this thesis is to combine clinical decision support systems with game-based environments. More specifically, serious games in the form of exergames, active video games that involve physical exercise, shall be used to deliver objective data for PD monitoring and therapy. Exergames increase engagement while combining physical and cognitive tasks. This combination, known as dual-tasking, has been proven to improve rehabilitation outcomes in PD: recent randomized clinical trials on exergame-based rehabilitation in PD show improvements in clinical outcomes that are equal or superior to those of traditional rehabilitation. In this thesis, we present an exergame-based clinical decision support system model to monitor symptoms of PD. This model provides both objective information on PD symptoms and an engaging environment for the patients. The model is elaborated, prototypically implemented and validated in the context of two of the most prominent symptoms of PD: (1) balance and gait, as well as (2) hand tremor and slowness of movement (bradykinesia). While balance and gait affections increase the risk of falling, hand tremors and bradykinesia affect hand dexterity. We employ Wii Balance Boards and Leap Motion sensors, and digitalize aspects of current clinical standards used to assess PD symptoms. In addition, we present two dual-tasking exergames: PDDanceCity for balance and gait, and PDPuzzleTable for tremor and bradykinesia. We evaluate the capability of our system for assessing the risk of falling and the severity of tremor in comparison with clinical standards. We also explore the statistical significance and effect size of the data we collect from PD patients and healthy controls. We demonstrate that the presented approach can predict an increased risk of falling and estimate tremor severity. Also, the target population shows a good acceptance of PDDanceCity and PDPuzzleTable. In summary, our results indicate a clear feasibility to implement this system for PD. Nevertheless, long-term randomized clinical trials are required to evaluate the potential of PDDanceCity and PDPuzzleTable for physical and cognitive rehabilitation effects.