Smartphone-basierte Untersuchung zu Stimmcharakteristik, Körperbewegungen und Geschicklichkeit bei Patienten mit REM-Schlaf-Verhaltensstörung, Parkinson-Krankheit und gesunden Kontrollpersonen

Zielsetzung: Wir versuchten mittels gängigen Smartphones motorische Symptome von Parkinson (PK)-Patienten, Patienten mit REM-Schlaf-Verhaltensstörung (RBD) und gesunden Kon- trollen zu erfassen und daraus abzuleiten, welcher der drei Gruppen diese angehörten und wie stark die Ausprägung der Symptom...

Full description

Saved in:
Bibliographic Details
Main Author: Beilharz, Moritz
Contributors: Oertel, W. (Prof. Dr. Dr. h.c.) (Thesis advisor)
Format: Doctoral Thesis
Published: Philipps-Universität Marburg 2024
Online Access:PDF Full Text
Tags: Add Tag
No Tags, Be the first to tag this record!

Objective: Our goal was to monitor motor symptoms of patients with Parkinson’s-Disease (PK), patients with REM-sleep-behavior-disorder (RBD) and healthy controls with conven- tional Smartphones and derive hereby to which of the three groups the monitored patient belongs and characterize the specific manifestation of the symptoms. Methods: Overall 43 RBD-, 19 PK- and 34 control-patients were conducting seven tests four times a day over a duration of one week to measure voice, balance, gait, dexterity (finger tap- ping), reaction time, rest tremor and postural tremor. The patients were conducting the seven tests once under supervision in the clinic and afterwards at home under uncontrolled conditions. The Parkinson- and RBD-patients were recruited in our specialized ambulances. We now tried to assign the patients to their cohort, using only the smartphone recordings. Furthermore, we tried to identify and characterize the most discriminative and salient features and tests for the assignment to the groups. Results: We were able to discriminate the three groups with good statistical certainty. For the three pairwise comparisons PK versus controls, RBD versus controls and PK versus RBD the AUCs were between 0,82 - 0,95 for 10 - 50 incorporated features. In an overall comparison postural tremor, rest tremor, voice and reaction time were the most important tests. We were able to show that especially postural tremor and reaction time are continuously changing with disease progression. Conclusion: Smartphones are able to detect subtle changes in motor function and quantify them objectively. They are able to distinguish between PK, RBD and healthy controls using Apps and statistical machine learning. To get a precise picture of the patient and his diseasestatus using only smartphones, further research and development of algorithms is needed. In the future the development of smartphone-based tools to measure the patient’s symp- toms will most likely be successful and improve the management of therapies.