Till Örebro universitet

Objective: To process tapping test results collected by a test battery device using signal processing methods. To analyze the tapping performance of PD patients in relation to their self-assessed motor conditions.

Background: A series of tests in a mobile device test battery, consisting of self-assessments and motor tests, were carried out repeatedly in a telemedicine setting. The test battery has been used by 65 patients with advanced PD in Sweden and 33 Italian patients; including 17 patients experiencing on-off fluctuations and 16 who were clinically stable. On test occasions, patients were asked to self-assess their momentary motor condition, just before starting 20-seconds-long tapping tests. There were three types of tapping tests: alternate tapping (AT), tapping with increasing speed (IS) and random chasing (RC).

Methods: Three self-assessed motor states were used in the analysis: OFF (including slightly, moderately and very off), ON (on without dyskinesias) and DYS (slightly, moderately and very dyskinetic). Digital signal processing and multivariate analysis methods based on data from the 65 Swedish patients were employed to process and summarize tapping results into scores for tapping distance and time, where 0 is worst and 1 is best possible scores. The scoring methods were then applied on the Italian data set. Fifty random samples of single test occasions per patient were drawn and then used in subsequent analysis. One-way ANOVA test with Tukey comparisons test was used to test whether the mean tapping scores were different among the self-assessed motor states. Statistical significance was set at p<0.05.

Results: In the stable group, the tapping distance scores were not different across the motor states for any test. The tapping time scores during AT were lower in OFF than in ON and DYS. During RC, the time scores were lower in OFF han in ON. In the fluctuating group, the tapping distance scores during AT were lower in OFF than in ON and DYS. The tapping time scores during RC were not different (p50.096) whereas during AT, time scores were higher in ON than in OFF and DYS.

Conclusions: Tapping scores in patient-rated OFF states were detected as poor performance by the signal processing method. The AT test shows best performance in separating the motor conditions and tapping time is a more important variable than tapping distance.