I-Corps: Chemometric fluorescence microscopic imaging and virtual staining for rapid label-free histopathology

Award ID: 2017396

The broader impact/commercial potential of this I-Corps project is to enable robust, rapid, and label-free biological sample evaluation. Cancer is a worldwide public health problem, and its diagnosis currently depends on evaluation of specimens of biological material. Time-efficient and objective alternative methods are urgently needed to address several shortcomings of the existing process, which is time-consuming, labor-intensive, and subject to interpreter variations. The proposed project is a validated method using light on unprocessed or minimally processed cell and tissue specimens. The proposed method particularly meets the demands of rapid and accurate diagnosis in surgical suites and telepathology laboratories. This I-Corps project is to advance the translation of chemometric fluorescence microscopic imaging and virtual staining (CFM-VS) on unstained cell and tissue specimens. Using endogenous cellular fluorescence, CFM produces 2D images revealing both subcellular morphology and function, visually differentiating specific cell properties including structure, cellular metabolism, and protein production. One unique advantage of CFM is the quantification of the absolute concentration of the endogenous fluorescent biomolecules, enabling reliable and accurate diagnosis. CFM and the derived virtual staining (CFM-VS) have been successfully applied to differentiate and diagnose lung and prostate cancers. The virtually stained images for unstained histological slides not only share the morphology of traditional hematoxylin and eosin (H&E) stained image counterparts, but also indicate the biochemical alterations due to cancer. Attractive features of CFM-VS include: ability to image unprocessed or minimally processed cell and tissue sections in close to real-time; yields virtual H&E stained images familiar to pathologists; eliminates distortions introduced in tissue processing; and robust diagnosis is achieved and may be improved through adaption of the algorithm from learning with the accumulation of data. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Program Director: Ruth Shuman