Objective: To evaluate a computational analysis of gait, based on pressure sensors insoles data, by correlating it with clinical assessments of motor symptoms in Parkinson’s disease (PD), over disease progression.

Background: Wearable sensors, enabling precise and objective assessments, are becoming key tools in the management of PD. Their application may facilitate disease diagnosis and monitoring in clinical or research settings.

Method: Participants were PD patients (Hoehn & Yahr stages 1-4) of the Movement Disorders Clinic, Patras University Hospital, Greece. The Smart-Insole Gait Assessment Protocol, as previously described [1], was employed for gait analysis at baseline and at 2-years follow-up. All participants wore a validated sensor insole system, calculating 16 gait temporal and spatial characteristics, during OFF and ON medication conditions. Sensor obtained data were correlated with clinical parameters, i.e. Part-III MDS-UPDRS total scores, axial and bradykinesia sub-scores, and medication status, at the two follow-up time points. Mixed linear models and Pearson correlation coefficient were employed for statistical analysis (p<0.05).

Results: Fourteen patients (12 male), with mean age 61.7±11.5 years, disease duration 11.3±9.3 years and levodopa equivalent dose 729.7±342.2 met inclusion criteria and completed assessments at 2-years follow-up. Five patients of the pilot study (N=19) could not complete the reassessment protocol (1 passed away, 3 switched to H&Y 5, 1 lost). At 2-years follow-up, mean total UPDRS-III score was 44.6±10.9 and 33.9±15.1, during OFF and ON conditions. Baseline scores were 40.7±20.7 and 30.6±20.4, respectively. Statistical analysis revealed significant correlations between UPDRS-III total, bradykinesia and axial sub-scores and sensor features, confirmed over follow-up assessments. Further, sensor gait characteristics could separate between the two medication states.

Conclusion: The present study suggests that computational gait analysis using pressure sensors may sufficiently measure motor changes in PD during disease progression. Further investigations are required for the evaluation of prognostic potential of the current approach.

References: [1]Chatzaki, C.; Skaramagkas, V.; Kefalopoulou, Z.; Tachos, N.; Kostikis, N.; Kanellos, F.; Triantafyllou, E.; Chroni, E.; Fotiadis, D.I.; Tsiknakis, M. “Can Gait Features Help in Differentiating Parkinson’s Disease Medication States and Severity Levels? A Machine Learning Approach”. Sensors 2022, 22, 9937. https://doi.org/10.3390/s22249937

« Back to 2024 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/pressure-sensor-insole-gait-assessment-for-parkinsons-disease-patients-a-longitudinal-study/