More than one in three people is affected by a neurocognitive disorders such as Alzheimer’s disease (AD), depression and schizophrenia at some point in their lives². In another study conducted in 2024, this number is projected to increase by 22% by 2050³.

Racing against time, Assoc Prof Zhou and team are bringing together human and machine intelligence to push the limits of neuroimaging and AI to deepen our understanding of how the human brain develops, ages and becomes vulnerable to disease. Their goal: treat illness, and promote resilience and brain health across the human lifespan.

“By 2030, the cost burden of brain disorders for Singapore is estimated to reach $30 billion,” she said. “Given our knowledge that neuropsychiatric disorders target specific neural networks linked to their symptoms and are associated with distinct ages—from childhood to late adulthood, we are using various brain imaging techniques like MRI and EEG, together with AI, to understand the brain’s structure, how different regions communicate and coordinate to influence behaviour and how they change functionally over time—essentially, the brain network phenotype.”

Understanding brain functions

The human brain is complex. It controls most of the activities of the body by processing, integrating and coordinating information received from the senses, and initiating verbal, physical and behavioural responses accordingly.

That is why, instead of focusing on a single brain region, Assoc Prof Zhou explores interactions between different regions. “I am a strong believer in studying brain circuits and brain-behaviour relationships to spot early signs of brain changes and to better track longitudinal trajectories, including disease progression and treatment response.”

Her interest led to her groundbreaking work that demonstrates the differences between AD brain network phenotypes and cerebrovascular disease in Asian populations. “We noted that almost 40% of people with dementia in Singapore have both AD and cerebrovascular disease. But on closer look, we saw that their brain network phenotypes are different,” she said. “What’s even more exciting is that we managed to verify the same findings with the Korean cohorts through collaboration with our South Korean colleagues.”

But the brain is not the only part that Assoc Prof Zhou’s lab has interest in. “We are equally interested in the brain-heart axis, and exploring the broader brain-body connection, among other things,” she said. “For example, we recently discovered that early microvascular changes in the brain are actually linked to cardiac dysfunction—and together, they can contribute to cognitive decline, indicating a possible heart-brain connection. We’ve also found that the brain is connected to our physical health, such as hand grip strength, which in turn is linked to cognition.”

Genetic and environmental influences on the brain

The work of understanding the human brain just got more interesting for Assoc Prof Zhou recently. “In the last two decades, both brain imaging technology and AI have progressed very rapidly,” she explained. “Together, they present a unique opportunity to better understand the human brain, and to develop personalised, precise treatments.”

Harnessing the two technologies, her team is currently developing brain foundation models that use deep learning and generative AI to address meaningful questions in neuroscience and Medicine. One example is the Brain-JEPA foundation model unveiled last year. “Most AI models require large amounts of labelled data. But by integrating self-supervised learning into the Brain-JEPA, our model learns abstract patterns in brain activity and can effectively predict demographics, understand personality traits, and diagnose brain disorders across different ethnic groups.”