Computerized identification of motor complications in Parkinson's disease

• Objective: To investigate whether spirography-based objective measures of motor dysfunctions are able to discriminate between Parkinson’s disease (PD) patients with different motor states (Off and Dyskinesia) and healthy elderly (HE) subjects.Background: Sixty-five advanced PD patients and 10 HE subjects performed repeated assessments of spirography, using a touch screen telemetry device. On each test occasion, they were asked to trace a pre-drawn Archimedes spiral using dominant hand and repeating the test three times. The clinical assessment was only performed in the patient group by animating the three spirals in a web interface, allowing a clinician (DN) to observe accelerations and spa-tial changes during the drawing process. A scale ranging from 0 (normal) to 4 (extremely severe) was used for the assessment of kinematic properties of speed, irregularity and hesitation. Finally, the momentary motor state of the patient was marked using two classes: - 1 (Off) and 1 (Dyskinesia). The HE samples were assigned a 0 (On) class and used in subsequent analysis.Methods: After time series analysis, 13 quantitative measures were calculated for representing the severity of symptoms in each individual kinematic property. Principal Component Analysis was then used to reduce their dimensions by retaining the first 4 principal components (PC). To investigate differences in mean PC scores across the three classes a one-way ANOVA test followed by Tukey multiple comparisons was used. An ordinal logistic regression model, using 10-fold cross-validation, was used to map the 4 PC to the corresponding motor state classes.Results: The agreements between computer and clinician ratings were very good with a weighted area under the receiver operating characteristic curve (AUC) coefficient of 0.91 (Table 1). The mean PC scores were different across the three classes, only at different levels (Fig 1). The Spearman’s rank correlations between the first two PC and visually assessed kinematic properties were: speed (PC1, 0.34; PC2, 0.83), irregularity (PC1, 0.17; PC2, 0.17) and hesitation (PC1, 0.27; PC2, 0.77).Conclusions: These findings suggest that spirography-based objective measures are valid measures of spatial- and time-dependent deficits in PD. The differences among the three classes imply that these measures can be used to assess changes in the motor states in response to therapeutic interventions.

Ämnesord

SAMHÄLLSVETENSKAP  -- Medie- och kommunikationsvetenskap -- Systemvetenskap, informationssystem och informatik med samhällsvetenskaplig inriktning (hsv//swe)

SOCIAL SCIENCES  -- Media and Communications -- Information Systems, Social aspects (hsv//eng)

MEDICIN OCH HÄLSOVETENSKAP  -- Klinisk medicin -- Neurologi (hsv//swe)

MEDICAL AND HEALTH SCIENCES  -- Clinical Medicine -- Neurology (hsv//eng)

TEKNIK OCH TEKNOLOGIER  -- Elektroteknik och elektronik -- Datorsystem (hsv//swe)

ENGINEERING AND TECHNOLOGY  -- Electrical Engineering, Electronic Engineering, Information Engineering -- Computer Systems (hsv//eng)

Nyckelord

spirography

Parkinson's disease

objective measures

motor complications

machine learning

individualized medication

PAULINA - Uppföljning av Parkinsonsymptom från hemmet

Publikations- och innehållstyp

vet (ämneskategori)

kon (ämneskategori)