Artificial Intelligence-Assisted Personalized Mixed Reality Rehabilitation for Older Adults With Mild Cognitive Impairment and Sarcopenia: A Randomized Controlled Trial

As people grow older, two quiet challenges often begin to erode independence at the same time. The first is a subtle decline in thinking skills, known as mild cognitive impairment, where memory and planning become slightly harder but daily life remains manageable. The second is sarcopenia, the gradual loss of muscle mass and strength that makes movement sluggish and increases the risk of falls. These two conditions frequently travel together, creating a difficult cycle: weak muscles make it harder to stay active, and a lack of activity accelerates the loss of muscle, while both factors can further cloud the mind. Traditional rehabilitation usually tackles one problem at a time, offering physical exercises for the body or mental puzzles for the brain, but rarely both together in a way that adapts to the person's changing needs.

Researchers have begun exploring whether new technology can break this cycle by combining physical movement with mental challenges in a single, immersive experience. They are testing systems that blend the real world with digital overlays, known as mixed reality, to create engaging tasks that require the user to think and move simultaneously. The question driving recent work is whether adding artificial intelligence to these systems can make the training more effective. By using computer programs that can analyze a patient's performance in real time, therapists hope to create a workout that automatically adjusts its difficulty, keeping the challenge just right to stimulate improvement without causing frustration. This approach aims to treat the whole person, addressing the linked decline of mind and muscle in one integrated session.

A recent study published by researchers from Funrehab Co., Ltd and Jeonju University put this idea to the test. They recruited forty older adults, all over the age of sixty-five, who had been diagnosed with both mild cognitive impairment and sarcopenia. These participants were divided into two groups. One group received a standard rehabilitation program led by physical therapists, which included exercises with resistance bands, balance training, and paper-and-pencil mental tasks. The other group used an artificial intelligence-assisted mixed reality system. This system presented interactive tasks on a screen that the participants controlled with a joystick. The unique feature of this group was that an artificial intelligence program constantly analyzed their health data and how well they performed in previous sessions. Based on this information, the system automatically adjusted the difficulty of the tasks, the distance the participants needed to reach, and the complexity of the mental challenges before every single session.

Both groups trained for four weeks, completing twenty sessions that lasted thirty minutes each. The researchers measured the participants' global cognitive function using a standard test called the Montreal Cognitive Assessment, which evaluates memory, attention, and language. They also measured physical strength by testing how hard participants could squeeze a handheld device, measured muscle mass using a specialized body scanner, and assessed how far participants could walk in six minutes. After the four weeks, the results showed a clear difference in how the two groups improved. The group using the artificial intelligence system showed a significantly greater improvement in their cognitive test scores compared to the group doing the standard exercises. Their scores on the thinking test rose more sharply, suggesting that the personalized, adaptive nature of the AI system helped stimulate the brain more effectively than the fixed routine.

However, the story was different for the physical outcomes. Both groups improved their walking distance and maintained their muscle strength, but the artificial intelligence group did not show a greater physical improvement than the standard group. The researchers found that while the high-tech system was excellent at boosting mental performance, it did not provide an extra physical benefit over the traditional therapy in this short timeframe. The participants in the AI group also completed exploratory measurements of their hand movements while using the joystick. These measurements showed that as the weeks went on, their movements became slower and more consistent, suggesting they were adapting their strategy to the task, though the researchers noted this was an observation of change rather than a definitive proof of better motor control.

The study also looked at how the participants felt about the training. They reported feeling confident in their ability to perform daily tasks without falling and rated their overall quality of life. Both groups saw improvements in these areas, but again, there was no significant difference between the high-tech group and the standard group. The artificial intelligence system was found to be safe and tolerable, with participants reporting low levels of frustration and feeling that they had enough time to complete the tasks. The mental effort required was high, which was expected given the dual nature of the training, but the system did not cause dizziness or nausea.

The researchers concluded that while artificial intelligence-assisted mixed reality is a promising tool for improving thinking skills in older adults with both cognitive and physical decline, it is not yet a magic bullet for physical strength. The study suggests that the ability of the AI to constantly tailor the difficulty of the tasks may be the key factor in helping the brain adapt and improve. The physical benefits, however, appeared to come from the act of rehabilitation itself, regardless of whether the system was high-tech or traditional. The authors emphasize that this was a small study with a short duration, and larger trials are needed to confirm these findings and to see if the benefits last over a longer period. For now, the work offers a hopeful glimpse into a future where rehabilitation can be as unique and responsive as the person receiving it, particularly for the complex task of keeping the mind sharp alongside the body.