High-Resolution Kidney Lipid Atlas Uses Imaging AI for Disease Insights
June 17, 2025
Vanderbilt researchers created the first high-resolution kidney lipid atlas using advanced imaging and machine learning, mapping over 100,000 tissue units.
Key Details
- Study published in Science Advances in June 2025 by Vanderbilt and Delft University teams.
- Mapped lipid profiles across more than 100,000 functional units in 29 donor kidneys using MALDI imaging mass spectrometry and microscopy.
- Applied multimodal registration and interpretable machine learning to co-register and analyze datasets.
- Identified molecular signatures tied to specific kidney structures and disease risk factors such as BMI and sex.
- Data and analysis tools are publicly available through NIH's HuBMAP program.
- Project funded by multiple NIH institutes and the Chan Zuckerberg Initiative.
Why It Matters
This molecular atlas advances the integration of molecular imaging and AI, offering a reference framework for kidney health and disease, which could lead to new diagnostics, therapies, and biologically-informed segmentation for renal pathology and radiomics.