ai
Researchers Develop MouseMapper AI Model for Whole-Body 3D Analysis
A new deep-learning framework called MouseMapper analyzes entire mouse bodies at the cellular level. The model maps nerves and immune cells across 31 organs and tissues. Researchers applied it to study diet-induced obesity and identified nerve changes linked to sensory deficits.
May 20, 11:01 AM·1m read1 source
ER
swissinfo.chAudio version
Tap play to generate a narrated version.
Researchers developed MouseMapper, a suite of foundation-model-based deep-learning algorithms for three-dimensional image analysis of entire mouse bodies. The system segments and quantifies peripheral nerves and immune cells while mapping structures to 31 organs and tissues.
MouseMapper was tested on mice fed a high-fat diet for 16 to 18 weeks. These animals showed increased body weight and impaired insulin response compared with chow-fed controls. Whole-body imaging revealed greater immune-cell infiltration in the liver and visceral adipose tissue of obese mice.
The model identified structural changes in the infraorbital branch of the trigeminal nerve in obese mice. These alterations were associated with functional sensory deficits in whisker sensing. Proteomic analysis showed changes in axon remodelling and complement pathways in the trigeminal ganglion.
Similar proteomic signatures were observed in trigeminal ganglia from obese humans. The framework also produced three-dimensional maps of immune-cell clusters across multiple tissues. MouseMapper demonstrated generalizability across different imaging resolutions and antibody-labelled datasets without retraining.
The study provides a scalable method for quantifying systemic changes in animal models of disease.
