Singapore neurologist harnesses AI to predict individual ALS progression patterns

Huang Jing Han was thirty years old when her body began to betray her. After the birth of her third daughter, she noticed her speech growing slurred, her legs weakening on the stairs. It would take two more years before a diagnosis arrived: amyotrophic lateral sclerosis, a disease that destroys the nerve cells controlling movement, leaving patients progressively paralyzed. She spent those two years in denial, trying acupuncture and traditional remedies, hoping the symptoms would reverse. They did not.

ALS kills slowly and unpredictably. Between 250 and 400 people in Singapore live with the disease at any given moment, and the National Neuroscience Institute diagnoses thirty to fifty new cases each year. The condition offers no cure, no way to stop its advance. What makes it especially cruel is that it does not advance the same way in everyone. Two patients might walk into a clinic with identical symptoms, and one will deteriorate rapidly while the other remains stable for years. Doctors have no reliable way to know which will be which, and families have no way to plan.

This unpredictability is what drew neurologist Crystal Yeo to artificial intelligence. Working at the National Neuroscience Institute, Yeo recognized that ALS is not actually one disease but a collection of diseases with different underlying biology. Some patients experience onset in their hands, others in their feet or tongue. Asian patients, Yeo's research found, tend to develop the disease five to ten years earlier than Caucasian populations, yet live roughly twice as long. The variation is profound, and it demands precision.

Yeo's team built machine learning models trained on clinical data from international sources, teaching the algorithms to recognize which factors best predicted how quickly a patient's disease would progress. The models examined blood test results, clinical monitoring scores that measure muscle strength and breathing capacity, nerve conduction studies, and proteins and genes found in motor neurons grown from patients' own stem cells. The team then validated these predictions not just in Western populations but specifically in Singapore's patient population, ensuring the findings would apply globally.

The work, published in 2025 in the peer-reviewed journal Muscle & Nerve, represents a shift toward what Yeo calls precision medicine. Rather than treating all ALS patients with the same approach, doctors can now use AI to analyze which drug combinations are most likely to work for a particular patient's biology, allowing for earlier intervention and more targeted care. "It is about finding the right treatment for the right patient at the right time," Yeo explained.

For Huang, now forty-one, the disease has reshaped her life in ways both devastating and unexpected. Nine years after diagnosis, she uses a wheelchair and an eye tracker to conduct technical analysis for day trading. She runs a small business, answers customer inquiries through WhatsApp, and meets with clients with the help of caregivers. The woman who once feared her three daughters would grow up without her has instead become, in her own words, a role model for them. "I don't let what I can't do stop me from doing what I can do," she said. She has learned to see each day as a gift, to focus on what remains rather than what was lost. Her story, and the work of researchers like Yeo, point toward a future where ALS patients might know earlier what to expect, plan more clearly, and receive treatments tailored to their individual biology rather than generic protocols. The disease remains fatal and progressive. But the darkness is becoming less absolute.