Objective Longitudinal Monitoring of Postural Stability and Gait During Neurorehabilitation: Clinical Application of Instrumented Posturography and Three-Dimensional Motion Analysis

Background: Instrumented human-movement assessment can be integrated with neurological examination and clinical scales to objectify functional deficits and quantify rehabilitation-related change. We evaluated a diagnostic-integration pathway in which TecnoBody ProKin posturography and Xsens MVN Awinda gait analysis were embedded in routine hospital neurorehabilitation as objective functional-diagnostic and outcome-monitoring tools. Methods: This manuscript reports a secondary exploratory longitudinal analysis of prospectively acquired routine-care measurements. The publication dataset comprised 151 participants with baseline posturography (stroke, n = 90; Parkinson’s disease, n = 38; multiple sclerosis [MS], n = 23); baseline gait data were available for 107. Diagnosis-specific Gaussian random-intercept models were fitted by restricted maximum likelihood with categorical time, and global time effects and PRE-to-POST contrasts were controlled using Benjamini–Hochberg False Discovery Rate (FDR) correction. Results: In stroke, dynamic open-eye Total Stability Index (TSI) showed an FDR-significant global time effect (Wald χ2(4) = 18.07, p = 0.00120, q = 0.00717), with lower model-estimated TSI at POST 2 (−1.46°, q = 0.0121; standardized effect = −0.50) and POST 3 (−1.92°, q = 0.0121; standardized effect = −0.65). Stroke cadence showed a global longitudinal signal in the full model (q = 0.0443); however, this finding did not remain FDR-significant in a sensitivity analysis excluding the sparsely populated POST 4 milestone (q = 0.139) and was therefore interpreted as exploratory. In Parkinson’s disease, the POST 1 TSI contrast was supportive only because the endpoint-level global test did not survive FDR correction. MS follow-up was insufficient for strong longitudinal inference, particularly for gait, and is interpreted primarily as a feasibility/data-availability finding. Conclusions: Diagnostic integration of ProKin and Xsens into routine neurorehabilitation was feasible and generated objective patient-specific trajectories for balance and gait outcome monitoring. The measurements can objectify rehabilitation-related functional change, but the uncontrolled design does not establish causal treatment efficacy, and clinically meaningful individual-change thresholds remain to be validated.