Quantitative Evaluation of Postural SmartVest's Multisensory Feedback for Affordable Smartphone-Based Post-Stroke Motor Rehabilitation

Abstract

Accessible tools for post-stroke motor rehabilitation are critically needed to promote recovery beyond clinical settings. This pilot study evaluated the impact of a posture correction intervention using the Postural SmartVest, a wearable device that delivers multisensory feedback via a smartphone app. Forty individuals with post-stroke hemiparesis participated in a single supervised session, during which each patient completed the same four-phase functional protocol: multidirectional walking, free walking toward a refrigerator, an upper-limb reaching and object-handling task, and walking back to the starting point. Under the supervision of their therapists, each patient performed the full protocol twice-first without feedback and then with feedback-which allowed within-subject comparisons across multiple metrics, including upright posture duration, number and frequency of posture-related events, and temporal distribution. Additional analyses explored associations with demographic and clinical variables and identified predictors through regression models. Wilcoxon signed-rank and Mann-Whitney U tests showed significant improvements with feedback, including an increase in upright posture time (p<0.001), an increase in the frequency of upright posture events (p<0.001), and a decrease in the total task time (p=0.038). No significant subgroup differences were found for age, sex, lateralization, or stroke chronicity. Regression models did not identify significant predictors of improvement.

Keywords: digital intervention; mHealth; multisensory feedback; postural balance; smartphone application; stabilization; stroke rehabilitation; wearable technology.

Figure 1

Postural SmartVest: athletic compression tank top with a smartphone pocket and a customizable Android app that delivers multisensory feedback [43]. The screen displays the message “System calibrated” and contains buttons to start, pause, and exit.

Figure 2

Postural SmartVest configuration panel: (a) full view and (b) close-up view [43].

Figure 3

Therapy room layout [43].

Figure 4

Examples of activities: (a) Ambulant activity and (b) upper limb activity [43].

Figure 5

Patients’ characteristics: (a) Histogram of time since stroke in months (tM). (b) Histogram of Functional Independence Measure (FIM) scores.

Figure 6

Three-dimensional visualization of patients’ best-at-the-time (BAAT) posture and the ideal reference point.

Figure 7

Distribution of patients’ deviation distances. (a) Histogram of Euclidean distances between each patient’s best-at-the-time (BAAT) and the ideal reference point. (b) Q-Q plot of the deviation distances relative to the BAAT posture.

Figure 8

Comparison of total duration spent in baseline and intervention phases. Wilcoxon signed-rank test W = 255.5, p = 0.038, with Cohen’s d = 0.31, indicating a statistically significant difference with a small effect size.

Figure 9

Comparison of number of posture-related events between baseline and intervention phases. Significant improvements were observed in “User is up straight” events, with a very large effect (Wilcoxon signed-rank test W = 13, p < 0.001; d = −1.45), as well as in “User leaned to the left” (W = 8, p < 0.001), “User leaned to the right” (W = 57, p < 0.001), and “User leaned forward” (W = 199.5, p < 0.05), all showing moderate to large effects. No significant difference was found in “User leaned back” events (W = 148, p > 0.05). Asterisks indicate statistical significance: * p < 0.05, *** p < 0.001.

Figure 10

Comparison of mean frequency of “User is up straight” events per minute in baseline and intervention phases. Wilcoxon signed-rank test (W = 28, p < 0.001) indicates a significant improvement in posture maintenance with a very large effect size (Cohen’s d = −1.71).

Figure 11

Temporal distribution of ”User is up straight” events. Wilcoxon signed-rank test W = 46, p < 0.001, with Cohen’s d = −1.40, indicating a large effect size and improved posture maintenance.