AI and rehabilitation: evaluate meaningful recovery
By ChatGPT Health. Educational synthesis prepared with AI assistance. Sources checked September 28, 2026. Clinically reviewed by Harvey Castro, MD, MBA, on September 28, 2026. Approved as educational content. This page is not a clinical protocol.
Rehabilitation technology should be judged by function, participation and the person’s goals. A sophisticated device or accurate movement score does not by itself show better recovery.
Separate robotics from AI
Robotic gait devices, motion sensors, exercise reminders and adaptive algorithms are different interventions. A study of a device may not isolate an AI component. Record exactly what the technology adds to therapy and whether the comparison includes the same therapist contact and training time.
What the stroke evidence supports
A 2025 Cochrane review included 101 trials and 4,224 adults. Electromechanical or robotic gait training plus physiotherapy probably improved independent walking at treatment completion compared with physiotherapy or usual care. Walking speed and distance did not clearly improve; sustained benefit remained uncertain. These findings concern rehabilitation devices, not proof of an AI specific benefit. Read the review.
Choose outcomes with the person
Define meaningful goals such as walking safely at home, using an arm for daily tasks, communicating or participating in work and family life. Match the measure to the goal. Record baseline function, assistance required, fatigue, pain, falls and adherence where relevant to the intervention.
Questions for a rehabilitation program
- Does the study include the same condition, severity and recovery stage as the intended population?
- What supervision and hands on support are required?
- Can people with cognitive, sensory or communication differences use the system?
- Does a home program have a clear route to therapist review when performance worsens?
- Are equipment access, maintenance, travel and caregiver burden included in the evaluation?
Use the evidence within a care plan
Ask the rehabilitation team how a technology fits the existing plan and which changes require professional review. Do not use an automated score alone to increase exercise intensity or remove supervision. A local evaluation should define who can pause the tool and how therapy continues without it.
Sources
Explore related topics
Put the evidence to work
Download the evidence and safety evaluation worksheets or compare the study designs and their limits.
Editorial standards · Suggest a correction. Please do not send identifiable patient information.