In studies of multiple sclerosis tissues, iSCALE uncovered immune features like activated microglia and lesion rim boundaries that traditional methods failed to detect, with findings confirmed through immunostaining and pathology review.

iSCALE employs machine learning trained on small molecular capture samples and histological images to infer gene expression at subcellular resolution of approximately 8 microns, enabling comprehensive molecular profiling of clinical tissue samples.

The platform addresses limitations of existing spatial transcriptomic technologies, such as high costs, low resolution, limited gene coverage, and small tissue areas, by reconstructing gene activity across entire pathology slides with near single-cell resolution.

Researchers plan further validation across various tissue types, including cancer, organ transplants, and neuroinflammatory tissues, and are exploring collaborations for clinical deployment.

This innovative technology has received funding from NIH, the Department of Defense, the Mayo Clinic Foundation, and the American Association for Cancer Research, with tissue samples provided by the Multiple Sclerosis Society Tissue Bank.

A study published in Nature Methods validates iSCALE's effectiveness for high-resolution gene expression imaging across large tissue sections, demonstrating its potential for broad biomedical applications.

A new AI platform called iSCALE has been developed at Vanderbilt University Medical Center to enable detailed cellular landscape mapping, identifying cell types, microenvironments, and spatial features often missed by traditional methods.