Data Sciences Institute - Aging and AI Workshop on September 28-29, 2026

Multi-site neuroimaging studies offer unprecedented opportunities to investigate aging and neurodegenerative disease in large, diverse populations. However, differences among sites—including scanner hardware, acquisition protocols, software versions, participant populations, and image-processing pipelines—can introduce variability that complicates the interpretation, integration, and reproducibility of neuroimaging results.
This two-day symposium and workshop will examine best practices for designing, conducting, and analyzing multi-site neuroimaging studies of aging and neurodegeneration. Speakers will consider challenges that arise throughout the research lifecycle, from initial study planning and protocol development to scanner upgrades, unexpected acquisition changes, data processing, statistical analysis, and clinical translation. Particular attention will be given to distinguishing meaningful biological variability from variability introduced by sites, scanners, and analytical methods.

The program will introduce prospective and retrospective approaches for improving the comparability of data collected across institutions, scanner platforms, and software environments. Examples will span diffusion MRI, structural MRI and measures of vascular brain injury, magnetic resonance spectroscopy, high-field vascular imaging, and amyloid positron emission tomography. Speakers will also discuss statistical and AI-based harmonization methods, including their assumptions, limitations, and effects on downstream analyses.

The event will further highlight the value of community-developed standards and harmonized analysis protocols, including the STRIVE standards for vascular brain injury and the Hippocampal Subfields Group harmonized segmentation protocol. Through lectures, panel discussions, and practical workshops, participants will explore how harmonization, interoperability, open science, and emerging AI tools can support more reliable, scalable, and clinically meaningful neuroimaging research.

This event is part of the Advancing Aging and Neurodegeneration Research through Data Science Emergent Data Science Program.

For more information, please visit

By the end of this event, participants will be able to:

1. Identify common sources of variability in multi-site neuroimaging studies, including differences in scanner hardware, acquisition protocols, software versions, and image-processing pipelines.
2. Evaluate prospective and retrospective strategies for improving the comparability and reproducibility of neuroimaging data collected across sites and platforms.
3. Describe how community-developed standards and harmonized protocols can support consistent measurement of structural, vascular, functional, metabolic, and molecular imaging markers.
4. Assess the opportunities, assumptions, and limitations of emerging statistical and AI-based methods for harmonizing neuroimaging data while preserving meaningful biological and clinical variability.

This is a two-day event, and you will be asked as part of your registration to indicate which day(s) you will attend:

• September 28, 9:00 am-5:20 pm
• September 29, 9:00 am-4:00 pm