Objective: To develop and evaluate the accuracy of a passive smartphone-based method for detecting and quantifying essential tremor (ET) severity.

Background: ET is among the most prevalent movement disorders, yet monitoring relies on episodic assessments which are subjective and may miss tremor fluctuations, precluding objective assessment of treatment response. We developed a passive tremor assessment method that detects and estimates tremor amplitude, utilizing smartphone accelerometer data enabling continuous quantification without dedicated hardware.

Method: 42 ET patients and 7 controls were recruited at a neurology clinic in southern Sweden. Participants performed standardized motor tasks holding their smartphone (iOS or Android) while simultaneously recorded with depth and regular cameras. Accelerometer data (50 Hz) were processed via principal component analysis, time-frequency persistence and spectral quality gating to 1) detect tremor segments, and a spectral-domain kinematic model to 2) compute displacement (mm). Detection was validated against time-labelled video and acceleration magnitude was compared to depth camera motion. Clinical Rating Scale for Tremor (CRST) Part A ratings (0–4) were assessed by two independent clinical raters to determine displacement cutoffs for each CRST category. Participants were subsequently monitored passively using smartphones for a mean of 32 days.

Results: Filming of patients during the day of enrolment yielded 2,829 video analysis windows from 42 participants. The pipeline detected 76.3% of video-confirmed tremor episodes. Smartphone acceleration showed agreement with depth camera measurements (Pearson r = 0.975). Over the monitoring period, subjects experienced an average of 104 (95%CI, 72-137) tremor events per day. The mean composite tremor score, calculated as the weighted proportion of smartphone detected tremor events scored CRST 1–4, was 1.76 (95%CI, 1.66-1.87).

Conclusion: A passive smartphone approach reliably detects tremor and quantifies severity in ET with substantial agreement to clinical ratings. These results support feasibility of continuous objective tremor monitoring using consumer smartphones, that could be utilized in clinical trials and advise treatment decisions.

« Back to 2026 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/passive-smartphone-based-tremor-quantification-in-essential-tremor/