Singapore's Ng Teng Fong General Hospital (NTFGH) is known for pushing technological boundaries, and their approach to gait training is no exception. In 2023, they became the first hospital in Southeast Asia to adopt an AI-powered gait training wheelchair system that learns from each patient's movements and tailors exercises in real time.
The system, developed by a local startup, uses 12 built-in sensors to track joint angles, muscle activity, and balance with millisecond precision. An AI algorithm then analyzes this data to identify weaknesses—say, a patient favoring their right leg—and adjusts the wheelchair's resistance or guidance to target those areas. Over time, the machine "learns" the patient's progress, gradually increasing difficulty as they get stronger.
"It's like having a physical therapist who never sleeps," laughs 45-year-old Eliza Ong, who used the AI system after a spinal cord injury left her partially paralyzed. "On day one, the wheelchair was doing most of the work—pulling my legs forward, keeping me upright. By week three, it started giving me 'quizzes': making my left leg work harder, throwing in unexpected stops to test my balance. One time, I even 'failed' a session because I leaned too much, and the screen flashed, 'Let's try that again—you've got this!' It sounds silly, but that little push made me want to prove the machine wrong."
NTFGH's results have been striking. In a pilot study with 100 stroke patients, 92% showed significant improvement in gait symmetry (how evenly they distribute weight on each leg), compared to 68% in the control group. Perhaps more importantly, patient satisfaction scores soared—many cited the AI's personalized feedback as a key motivator to stick with therapy.