Session Title: Phenomenology and Clinical Assessment of Movement Disorders
Session Time: 1:15pm-2:45pm
Location: Les Muses Terrace, Level 3
Objective: A medical evaluation protocol as gold standard for controlling iPrognosis application derived data of an international multilingual study.
Background: Smartphones as daily companions are the digital health tools of the future due to the potential of longitudinal, unobtrusive, remote real-life monitoring of people’s behavior. In the i-PROGNOSIS project, we developed the “iPrognosis” Android smartphone application for unobtrusive remote data collection in the general population, with the aim of evolving it into an early Parkinson`s disease (PD) detection tool. Early detection of behavioral changes being linked to motor and non-motor symptoms (NMS) of PD allows for timely clinical diagnosis which is an unmet need.
Method: The medical evaluation protocol is conducted as a prospective longitudinal multicenter study in Greece, Germany and United Kingdom involving early stage PD patients and healthy controls (HC), all using the iPrognosis app. The study encompasses physician and participant-based scales covering motor, NMS and health related QoL with baseline evaluations and six month follow-ups [table 1]. The MDS recommended probability methodology for the diagnosis of prodromal PD is estimated (1). The iPrognosis app passively collects participants’ behavioral data from the daily use of their smartphones, including voice features, accelerometer data for tremor, keystroke timing data (2) for brady-/hypokinesia and rigidity assessment.
Results: Of 2185 App installations, 1465 participants provided consent and 1266 (214 PD patients, 1052 HC) contributed data. So far 70 participants (47 PD patients, 66% male, mean age 61.91±7.63 and 23 HC, 43% male, 55.96±10.91) underwent a medical evaluation. The age- and sex-matched group of PD patients vs. HC were significantly different in, e.g. motor: mean UPDRS part III 17.96±10.45 vs. 2.70±5.17 (p<0.01); NMS: NMSQuest 6.42±4.35 vs. 3.18±3.29 (p<0.01), HrQoL: PDQ-8 5.64±4.09 vs. 1.86±2.82 (p<0.001). Comparison of clinical gold standard with iPrognosis data-based machine learning algorithms showed highest diagnostic accuracy for rigidity (0.89), followed by brady-/hypokinesia (0.83) and tremor (0.77).
Conclusion: Our findings indicate that the i-PROGNOSIS approach is promising as it differentiates PD patients from HC as the gold standard clinical evaluation by a movement disorders specialist.
References: (1) D. Berg et al. MDS Research Criteria for Prodromal Parkinson’s Disease. Mov Disord. 2015 Oct;30(12):1600-11. (2) D. Iakovakis et al. Motor impairment estimates via touchscreen typing dynamics towards Parkinson’s disease detection from data harvested in-the-wild. Front. ICT 2018;5:28.
To cite this abstract in AMA style:L. Klingelhoefer, S. Bostanjopoulou, D. Trivedi, S. Hadjidimitriou, S. Mayer, Z. Katsarou, V. Charisis, M. Stadtschnitzer, S. Dias, G. Ntakakis, N. Grammalidis, K. Kyritsis, H. Jaeger, D. Iakovakis, I. Ioakeimidis, F. Karayiannis, J. Diniz, A. Delopoulos, L. Hadjileontiadis, H. Reichmann, K. Chaudhuri. Medical evaluation as gold standard to control iPrognosis application derived data for early Parkinson’s disease detection [abstract]. Mov Disord. 2019; 34 (suppl 2). https://www.mdsabstracts.org/abstract/medical-evaluation-as-gold-standard-to-control-iprognosis-application-derived-data-for-early-parkinsons-disease-detection/. Accessed December 3, 2023.
« Back to 2019 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/medical-evaluation-as-gold-standard-to-control-iprognosis-application-derived-data-for-early-parkinsons-disease-detection/