Cracking the Code: AI and Alzheimer’s Disease

For UC Noyce Initiative researchers Brittany Dugger, Ph.D. and Chen-Nee Chuah, Ph.D., their passion for Alzheimer’s research is deeply personal. 

In her late teens, Dugger watched as both her grandmothers— Grandma Morenski and Grandma Dugger—suffered from the devastating effects of the disease. Despite sharing the same diagnosis, their symptoms were strikingly different. That inconsistency left her with lingering questions: 

How could two people experience the same disease in such different ways? 

Finding the answer to that question has since guided her career, led her to focus on neuropathology research at UC Davis and enabled her to become the leader of the university's neuropathology core. But tackling a disease as complex as Alzheimer’s required an unconventional approach—and a powerful collaborator. 

Enter Chen-Nee Chuah, a professor in electrical and computer engineering at UC Davis and a leading expert in data science and artificial intelligence. Like Dugger, Chuah had a personal motivation for delving into medical research and becoming the principal investigator of this UCNI-funded project. 

“When I was dealing with my own health issues, I was frustrated by the gaps in applying cutting-edge computational methods to the medical domain,” she recalls. “I realized the data-driven AI techniques my team develops for networking research could potentially make a real difference in understanding complex neurological diseases like Alzheimer’s.” Chuah continued, “My mom also suffered from Alzheimer's, and did not recognize me as her daughter towards the end, while her physical health continued to decline and she passed away in 2014. So this work is personal to me as well.”

Now, with support from UCNI, Chuah and Dugger are spearheading an interdisciplinary effort to revolutionize Alzheimer’s research using AI. Chuah and Dugger have teamed up with experts from UC San Francisco (UCSF) and UC Irvine to conduct research that is leading to new breakthroughs in diagnosing and understanding the disease. Together, they are enhancing the way science is done and shared. 

Harnessing AI to Decode Alzheimer’s 

The challenge Dugger and Chuah’s team faces is enormous. Traditional methods for studying Alzheimer’s rely on experts meticulously reviewing high-resolution whole slide images (WSIs) of brain tissue—an extremely time- consuming and labor-intensive process. AI, however, has the potential to revolutionize this workflow. 

“Our goal is to create a robust, end-to-end AI pipeline that can take WSIs as input and generate valuable insights that inform both Alzheimer’s research and clinical decision-making,” Chuah explained. 

By leveraging semi-supervised learning techniques, the team has developed a deep-learning framework capable of quantifying pathologies associated with Alzheimer’s Disease in different brain regions with remarkable precision without relying on large labeled datasets. The team’s model demonstrated a 19% improvement in pathology detection accuracy while requiring only 0.1% of labeled data—a breakthrough that significantly improves efficiency. This breakthrough alleviates neuroscientists from having to engage in the labor intensive process of annotation. 

“This work is truly transformative because it allows us to make the most of the limited labeled data available, while still achieving high accuracy in detecting key biomarkers of Alzheimer’s,” Dugger said. 

A Collaborative Effort Across the UC 

Dugger and Chuah’s efforts extend beyond UC Davis. Their research is part of a multi-institutional collaboration that includes UCSF and UC Irvine. Each university contributes a unique piece to the puzzle: UC Davis specializes in digital pathology and AI-driven analysis, UC Irvine provides neuroradiology imaging expertise, and UCSF focuses on neurofibrillary tangle (NFT) detection. The synergy across institutions is helping to create a comprehensive, AI-powered framework that can be shared across research centers worldwide. 

“By bringing together experts from different fields, we can tackle these complex problems from multiple angles and accelerate the pace of discovery,” Dugger said.

From Research to Real-World Impact

The implications of this work extend beyond the research lab. By making AI models more interpretable and accessible to domain scientists, Chuah and Dugger are paving the way for more precise diagnoses and early detection tools that could one day be used in clinical settings. With a $6 million NIH grant and additional support from the ChanZuckerberg Initiative, the team is scaling up their efforts. They are refining their models, expanding AI training capabilities to incorporate new neuropathology tasks, and exploring new avenues for detectingAlzheimer’s earlier and more accurately than ever before.“

The ultimate goal is to make these tools freely available to researchers worldwide, ensuring that no scientist is limited by computational resources or dataset constraints,” Dugger explained.

A New Era in Alzheimer’s Research

For Chuah and Dugger, this project represents the culmination of years of work at the intersection of AI and real-world problem-solving. “When you see the devastation that Alzheimer’s can bring to individuals and their families, it really puts things in perspective,” she reflected. “I’m honored to be part of this effort, and I’m hopeful that our work can make a meaningful difference in the lives of those affected by this disease.” 

UC Davis Health contributed to this story.