Project Grant 2141157

This National Science Foundation Project Grant of $495,000 will support the development of biopsy-free, label-free three-dimensional virtual histology of intact skin through March 2025. Funded under the Engineering program (CFDA 47.041), the award to the University of California, Los Angeles aims to advance a machine learning-based technique to transform 3D skin images into virtually stained images that can facilitate accurate and rapid diagnosis of skin lesions without biopsy. Specifically, the project will utilize deep neural networks to train a virtual histology algorithm to obtain stained images of normal skin and lesions suspicious for basal cell carcinoma from intact skin samples. Tests will evaluate whether the virtual histology format can improve dermatologist diagnosis compared to the gold standard of histological processing. If successful, the studies could help transform clinical dermatology practice by enabling non-invasive biopsy-free histologic assessment of skin.