Student Researchers Share Projects on Alzheimer Disease, Autism, and Long COVID, From the Regeneron Science Talent Search

Emerging high school researchers presented findings spanning neurodevelopmental diagnostics, long COVID neuroimaging, and Alzheimer disease vascular injury at the Regeneron Science Talent Search competition.^1,2

Edward Kang investigated whether machine learning models applied to retinal fundus images could diagnose autism spectrum disorder and attention-deficit/hyperactivity disorder. His computational models achieved at least 89% diagnostic accuracy, with explainable AI techniques identifying the macula and fovea as regions of particular discriminatory importance. On the biological side, Kang identified dozens of candidate genes involved in retinal development that had not previously been linked to autism spectrum disorder, offering a potential mechanistic basis for retina-based neurodevelopmental screening. He noted that "these large statistical models, these machine learning tools, are often so complex that no human can fully understand what is actually going on inside, and how the models are actually making their diagnostic decisions," underscoring the importance of model interpretability in clinical translation.

Jashvi Desai examined neuroimaging data from long COVID patients who reported persistent cognitive symptoms despite normal performance on standardized cognitive tests. Using a brain age prediction model, Desai found that COVID-19 patients paradoxically appeared younger neurologically than their chronological age—an artifact she attributed to abnormal cortical thickening consistent with neuroinflammatory processes rather than healthy aging. She further observed compensatory prefrontal cortex recruitment, concluding that "long COVID doesn't just affect the brain structurally—it does change the brain function and how the brain compensates for that loss of function."

Leon Wang investigated repurposing US Food and Drug Administration-approved lung fibrosis drugs—pirfenidone and nintedanib—to mitigate blood-brain barrier vascular damage in Alzheimer disease using stem cell-derived endothelial and pericyte cell cultures. Both drugs significantly reduced a vascular damage biomarker, with pirfenidone additionally promoting healthier cell expression, suggesting potential for both preventive and rescue therapeutic applications.

Ms Desai is a top 40 finalist of the Regeneron Science Talent Search and a high school senior at Yorktown High School.

Mr Wang is the 8th place winner of the Regeneron Science Talent Search and a high school senior at King School.

Mr Kang is the 2nd place winner of the Regeneron Science Talent Search and a high school senior at Bergen County Academies.

References

1. Regeneron Science Talent Search 2026 recognizes America’s top young scientists, awarding more than $1.8 million to high school seniors for innovative research in computational mathematics, neural science, and blood cancer treatment. Society for Science. March 10, 2026. Accessed April 15, 2026. https://www.societyforscience.org/press-release/regeneron-sts-2026-top-awards/

2. Regeneron science talent search. Society for Science. Accessed April 15, 2026. https://www.societyforscience.org/regeneron-sts/