MouseMapper: AI Foundation Model Maps Systemic Damage from Obesity at Whole-Body Scale
Key Takeaways
- ▸MouseMapper is a foundation model-based deep learning system that automatically analyzes whole-body 3D imaging data, segmenting 31 organs and tissues while mapping nerves and immune cells
- ▸Obesity causes widespread structural damage across multiple organ systems, including previously unknown remodeling of the trigeminal nerve with reduced sensory endings and impaired sensory function
- ▸Molecular signatures of obesity-related nerve damage are conserved between mouse models and human tissue, validating preclinical findings and suggesting translational relevance for human disease
Summary
Researchers at Helmholtz Munich's Institute for Biological Intelligence have developed MouseMapper, an AI-powered suite of deep learning algorithms that automatically analyzes whole-body biological imaging data at unprecedented resolution. Built on foundation model architecture, MouseMapper can segment 31 different organs and tissue types while quantitatively mapping nerves and immune cells throughout intact organisms, enabling comprehensive multi-system disease analysis without manual region-of-interest selection. This breakthrough addresses a critical research gap: scientists previously lacked tools to study disease-associated changes across entire bodies in living organisms at high resolution.
The team applied MouseMapper to investigate obesity's systemic effects by imaging fluorescently-labeled mice using light-sheet microscopy, revealing widespread structural and organizational changes across multiple tissues including fat, liver, muscle, and peripheral nerves. Most strikingly, obesity caused significant damage to the trigeminal nerve—a major facial nerve—reducing nerve endings and branches by far fewer endings and branches. Behavioral tests confirmed this structural damage correlated with impaired sensory responses in obese mice.
Crucially, the researchers validated findings across species using spatial proteomics: the same molecular signatures of nerve remodeling detected in obese mice were identified in human trigeminal tissue samples. This cross-species conservation suggests obesity's neurological damage occurs in humans as well, opening new avenues for understanding obesity-related complications and identifying therapeutic targets.
- The foundation model architecture enables MouseMapper to generalize beyond its training data, suggesting broad applicability to other whole-body imaging research
Editorial Opinion
MouseMapper represents a transformative approach to systems-level biomedical research, harnessing AI foundation models to unlock insights previously impossible at whole-body scale. The cross-species validation—demonstrating that obesity-related neural damage occurs in both mice and humans—exemplifies AI's power to bridge the notoriously difficult preclinical-to-clinical translation gap. By automating analysis that would require teams of specialists for manual annotation, this work democratizes access to cutting-edge discovery and sets a template for AI-driven investigation across multiple disease areas. We anticipate foundation model approaches like MouseMapper will fundamentally accelerate our understanding of systemic disease mechanisms.
