Event Date
Event Date
We present a label-free optical phenotyping platform that integrates multiphoton microscopy with spatial transcriptomics for pancreatic cancer analysis. By co-registering autofluorescence and second harmonic generation images with gene expression maps, we characterize structural and molecular heterogeneity in intact tissue. Our deep learning model achieves over 89% accuracy across six tissue types, with ROC-AUC values nearing 1.0, demonstrating robust, non-destructive classification. This approach enables scalable, stain-free tissue analysis and holds promise for advancing clinical diagnostics and precision oncology.
Presenter
The Univ. of Arizona (United States)
Shuyuan Guan is a Ph.D. candidate in Optical Sciences at the University of Arizona, specializing in computational optical imaging and biomedical photonics. Her research focuses on the design and development of optical imaging systems, and leveraging deep learning techniques to enhance the outcomes of biomedical imaging. She carries a strong background in diffractive optics, spatial light modulators, holography, biomedical imaging system design, and wavefront shaping, she bridges the gap between physical optics and AI-driven image analysis. Her recent work integrates spatial transcriptomics with label-free imaging to investigate tumor heterogeneity at the cellular and molecular level. Shuyuan aims to advance non-invasive optical diagnostics through the fusion.