Campus News

USC neuroscientist delivers Talairach Lecture on worldwide data sharing and AI

Sidney Taiko Sheehan July 31, 2025
Photo shows Paul Thompson delivering the Talairach Lecture at OHBM 2025.

Paul M. Thompson, PhD, delivering the Talairach Lecture at OHBM 2025. Photo/Stevens INI

In a keynote at the 2025 Organization for Human Brain Mapping (OHBM) annual meeting, Paul M. Thompson, PhD, associate director of the USC Mark and Mary Stevens Neuroimaging and Informatics Institute (Stevens INI) at the Keck School of Medicine of USC, delivered the prestigious Talairach Lecture, outlining a bold vision for harnessing artificial intelligence and worldwide collaboration to combat brain diseases.

Speaking to a packed audience of neuroscientists and clinicians from across the globe, Thompson emphasized that solving the mysteries of the human brain requires scaling science beyond borders—and beyond traditional methods.

“We’re entering a new era where our ability to tackle brain diseases is no longer limited by geography or scale,” Thompson said. “Over the last decade, we have created a vast global alliance of researchers pooling their expertise, data, and technologies — working night and day in the quest to battle human brain diseases.”

Thompson, who leads the ENIGMA consortium—a global alliance of over 2,000 scientists spanning 47 countries—drew from more than a decade of experience in large-scale neuroimaging collaborations. His lecture showcased how uniting brain scans and clinical data from around the world can reveal common patterns in conditions such as Alzheimer’s disease, schizophrenia, depression, and ADHD—insights that would be impossible to detect in smaller, localized studies.

The talk also spotlighted the growing role of artificial intelligence in neuroscience. Thompson demonstrated how advanced machine learning models, including generative AI and vision-language networks, are beginning to decode subtle patterns in brain structure and function, automatically labeling abnormalities and predicting disease risk years before symptoms emerge.

He unveiled early results from a novel system that generates descriptive text summaries from brain MRI data—potentially transforming how radiologists and researchers interpret scans. These AI-powered tools, trained on tens of thousands of brain images, could dramatically improve diagnostic accuracy and speed while offering more accessible insights across languages and skill levels. The Stevens INI is a global leader in developing and applying these cutting-edge AI tools, driving forward innovation in brain mapping, diagnostics, and data sharing.

However, Thompson acknowledged that scientific progress must be matched by ethical responsibility. His lecture addressed persistent challenges in data privacy, global equity, and algorithmic fairness, especially when deploying AI in low-resource settings.

“If we want to build a truly global neuroscience,” Thompson said, “we need tools that work for every population—and protections that respect every individual. We have led the largest brain imaging studies of over 30 brain diseases for over 15 years; AI is now bringing a second revolution in the diagnosis and discovery of novel treatment types. Just as the brain is a network, so is the future of neuroscience—a mutually supportive effort across countries and disciplines, powered by AI, shared data, and unlimited human creativity.”

Thompson closed with a call for a new generation of open-access platforms, global studies led by teams from diverse cultures and backgrounds, and AI-driven diagnostics that are inclusive, scalable, and grounded in collaboration.

The Talairach Lecture, named after famed French neurosurgeon, Jean Talairach, is one of the highest honors at OHBM. Thompson’s selection underscores his long-standing impact on the field, including his pioneering work in computational brain mapping and data-sharing.

“Dr. Thompson’s lecture was a call to action for the global neuroscience community,” said  Arthur W. Toga, PhD, director of the USC Mark and Mary Stevens Neuroimaging and Informatics Institute. “At the Stevens INI, we are deeply committed to advancing collaborative science that spans continents, disciplines, and technologies. Paul Thompson’s work exemplifies our mission to harness the power of data, innovation, and partnership to understand the brain and fight disease. It’s a point of pride to see our faculty shaping the future from within our institute.”

A version of Thompson’s lecture is available here.