Objective: To classify step-by-step gait patterns of Parkinson’s disease (PD) patients and older adults from the assessment of short episodes of gait with body-worn-sensors (BFS) placed on the low-back.

Background: Quantitative assessment of gait in patients with PD has the potential to improve differential diagnosis and to monitor treatment and disease progression. However, effective clinical evaluation is dependent on fast and valid assessment. Thus the implementation of quantitative analysis of short episodes of gait, e.g. with BFS, present a promising addition to clinical routine assessment of PD patients.

Methods: Following an acoustic signal, 24 PD patients (mean age: 60 ± 9 y, H&Y stage 1–2.5; ie, without relevant gait impairment) and 14 older adults (mean age: 61 ± 1 y) walked a distance of 5 meters at self-selected (SS) and fast speed (FS) wearing a BFS (triaxial accelerometer and triaxial gyroscope) at the low-back (L5) level. Nine Spatiotemporal parameters (duration, displacement, range of anterior-posterior velocity, root-mean-square of acceleration and angular velocity in the 3-axes) were calculated for 7 different gait phases (from the acoustic signal to the start of movement, from the acoustic signal to the first heel-strike, second step, third step, intermediate step, pre-last step, and last step). Additionally, the same parameters were calculated relatively to the mean value across steps (marked with *R). Unpaired t-tests were applied for all calculated parameters between cohorts. For the significantly different parameters, sensitivity and specificity were calculated by choosing the best classification threshold. In addition, the harmonic mean of sensitivity and specificity (F1) was obtained.

Results: In both trials, most of the significantly different parameters between cohorts were found at *R initial phases: 13 out of 38 significantly different parameters from the SS trial, and 13 out of 39 from the FS trial. The *R displacement from the acoustic signal to the start of movement had the highest F1 in both trials, 0.842 in SS, and 0.828 in FS.

Conclusions: This study suggests that gait assessment over a distance of 5 meters and instrumented with BFS at the low-back level allows to differentiate early PD (with non obvious gait deficits) from controls with acceptable accuracy. These results may motivate further investigation of the usefulness of BFS in clinical routine assessment.

« Back to 2016 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/is-the-assessment-of-5-meters-of-gait-with-body-worn-sensors-enough-to-recognize-parkinsonian-gait/