Review 2 Mentions
Artificial intelligence in rehabilitation: a review of clinical effectiveness, real-world performance, safety, and equity across modalities and settings
Nafisa Abdalla•2026
Rabie Adel El ArabAmany Ahmed Abdrbo
Top 1% · 99th Percentile
9 citations · Rehabilitation
Open Access

Summary

1 Sample Definition And Size

This paper is an umbrella review of reviews (i.e., a review of systematic reviews) using a Population–Exposure–Outcome framework. It synthesizes findings across multiple reviews of AI-enabled (ML/DL) and technology-assisted (non‑ML) rehabilitation interventions. The exact number of reviews included is not specified in the abstract or metadata.

2 Study Type

Umbrella review (review of reviews) synthesizing evidence across AI‑enabled and technology‑assisted rehabilitation modalities.

3 Conflicts Of Interest

One author, JL, serves as a consultant for Sword Health, Inc. The remaining authors declared no commercial or financial relationships that could be construed as potential conflicts of interest. ([frontiersin.org](https://www.frontiersin.org/journals/digital-health/articles/10.3389/fdgth.2026.1737957/full?utm_source=openai))

4 Results Summary

Key findings: Technology‑assisted training (robotics with or without VR) shows the most reproducible clinical signal for improving activity in post‑stroke upper limb rehabilitation, though effects on impairment and independence are inconsistent when dose is matched and assessors are blinded. Claims of non‑inferiority lack prespecified margins and confidence‑interval testing. AI‑enabled interventions show a development‑to‑deployment performance drop, notably in brain–computer‑interface classifiers and computer‑vision movement evaluation. Imaging‑based decision support (radiomics/CNN) is closer to practice but requires local calibration and impact evaluation. Reported adverse events are generally mild, but usability, adherence, equity, and cost are under‑measured, especially in home and hybrid settings. Prediction‑model and trial reporting often fall short of AI standards; representation skews toward high‑income settings, and subgroup performance is seldom reported. ([frontiersin.org](https://www.frontiersin.org/journals/digital-health/articles/10.3389/fdgth.2026.1737957/full?utm_source=openai))

Abstract

Background: Rehabilitation faces a scale problem: millions who could benefit lack timely, effective services. Artificial intelligence (AI) and device-based modalities (e.g., robotics and VR) can extend reach and personalise care when validated, yet decision-makers lack a consolidated view of clinical usefulness, translation to practice, safety, equity, and cost. Methods: We conducted an umbrella review of reviews using a Population-Exposure-Outcome framework. Searches span biomedical, allied health, and engineering databases from inception to September 1, 2025. We distinguished AI-enabled (ML/DL) interventions from technology-assisted (no ML demonstrated) modalities and synthesised outcomes across impairment, activity, independence, usability/safety, equity, and economics. Findings: The most reproducible clinical signal is activity improvement for post-stroke upper limb with technology-assisted training (robotics with or without VR) that increases task-specific practice; effects on impairment and independence are inconsistent once dose is matched and assessors are blinded. Claims of non-inferiority are not established when prespecified margins and confidence-interval testing are absent, so parity is interpreted as no between-group advantage under those conditions. Across AI-enabled domains, a development-to-deployment performance drop is evident most notably for brain-computer-interface classifiers and computer-vision movement evaluation limiting immediate clinical impact. Imaging-based decision support (radiomics/CNN) is closer to practice but varies by software and site, requiring local calibration and impact evaluation before pathway change. Reported adverse events are generally mild, yet usability, adherence, equity, and cost are under-measured, particularly in home and hybrid delivery. Prediction-model and trial reporting frequently fall short of contemporary AI standards; representation skews toward high-income settings, and subgroup performance is seldom reported. Conclusion: An adjunct-first posture is warranted. Adoption should be gated by minimum clinically important difference-anchored benefit under dose symmetry and blinded assessment; external, multi-site validation with declared lab-to-clinic performance loss; subgroup fairness with mitigation; decision-grade economic value; interoperability; and readiness for regulation, change control, and cybersecurity. Priorities include pragmatic, multi-site, assessor-blinded, dose-matched trials; standardised safety/usability capture for home use; and a public, living evidence atlas. AI can expand rehabilitation when held to clinical standards that matter to patients and services. With clear adoption gates and continuous post-market monitoring, systems can extend access and independence without sacrificing rigour, safety, equity, or fairness.

Referenced In