A digital assessment system for evaluating kinetic tremor in essential tremor and Parkinson’s disease

Abstract

BackgroundSpiral drawing on papers is a common tremor evaluation tool for diagnosing patients with essential tremor (ET) or Parkinson’s disease (PD). No standard drawing methods and parameters that use graphic tablets are yet available for objective evaluation.MethodsThis study established a tremor assessment system for tremor severity by using graphic tablets. Twelve patients with ET and twelve patients with PD were tested to establish system algorithms, and six additional patients were tested with the developed system to evaluate its performance. The patients also performed spiral drawing with three guiding paradigms on a graphic tablet: traced along a given spiral (S1), performed freehand drawing (S2), and traced along a guiding point (S3). Three parameters were calculated to quantify tremor severity: the means of radial difference per radian (|dr/dθ|), the means of radial difference per second (|dr/dt|), and the area under curve (AUC) of the frequency spectrum of the velocity. Each patient’s drawing was also evaluated using a visual rating scale (VRS) by experienced physicians. The interrater reliability was examined to identify the most consistent test paradigm.ResultsThe parameter |dr/dθ| and AUC correlated well with the VRS (R > 0.8) in S1, S2 and S3 tests. The S1 test presented the best interrater reliability (Weighted Kappa coefficient, k = 0.80) among three tests. The Weighted Kappa coefficients are 0.67 and 0.71 in S2 and S3 tests, respectively.ConclusionsWe developed three different guiding paradigms for spiral drawing on a digital graphic tablet for clinical tests. Three parameters were calculated to represent the tremor severity in spiral drawing and used to quantify temporal and spatial characteristics of tremor, and provided good correlation with current clinical assessments. The test “traced along a given spiral” is recommended due to its good interrater reliability.Electronic supplementary materialThe online version of this article (10.1186/s12883-018-1027-2) contains supplementary material, which is available to authorized users.

A Multimodal Approach to the Quantification of Kinetic Tremor in Parkinson’s Disease

“…We refer to two studies, the first being the work by Lin et al [12], where the authors reported Pearson's correlations between the VRS and the investigated signal-derived parameters to be as high as 0.973. The models of VRS were trained and validated with digitized spiral drawings from ET and PD patients.…”

“…We recognize this generalizability issue as very important in the context of targeting the kinetic tremor, as this symptom is widespread in ET. Combining data from PD and ET patients would allow for better comparison with, e.g., work by Lin et al [12]. However, one should be aware that, during the model fitting, such a combination may result in discarding parameters that are specific for only one of the diseases, yielding models that are more general, albeit with inferior prediction quality when tested solely in PD or ET populations.…”

“…Teaching predictive models based on both of the aforementioned groups of parameters, being a multimodal approach, is a novelty in the field of the kinetic tremor diagnostics. The solutions found in literature are based exclusively on one of two types of parameters mentioned above, commonly with a disease diagnosis or a disease stage classification as an output [8][9][10][11][12][13][14].…”

“…As for the quantification of kinetic tremor, two of the cited studies are relevant. Lin et al [12] presented a system for assessment of the tremor severity through a digital analysis of spiral drawings by PD and essential tremor (ET) patients. The quantification was based on signal-derived parameters (no processing with ANNs was performed), and predictions correlated well with visual ratings provided by physicians.…”

Validation of a Digital Interface for Assessment of Motor Function Based on MFM

“…While not making direct measurements, dexterity can be judged by the quality of the line left on the screen. The speed, acceleration and execution time of the drawings / tasks can also be related to the degree of loss of functionality of the hands [16].…”