Label-free biomedical imaging is becoming increasingly popular, as these techniques provide visualization of tissues and biological phenomena by leveraging naturally occurring contrast mechanisms. Quantifying images from label-free imaging modalities is often pursued to establish diagnostic approaches for tissue classification; for example, for cancer detection. However, one confounding factor is that many sources of label-free contrast produce significantly overlapping signal, which makes it challenging to identify true biological sources of measured variations in image contrast. Furthermore, significant heterogeneity in patient samples can severely limit performance of quantitative approaches when using simple methods for validation (e.g. simple binary classification of tissues using histology). Thus, realizing the full potential of label-free imaging requires advancements in robust validation techniques that can elucidate underlying sources of contrast and account for biological variability. I will discuss ongoing work to introduce a new method for validation of label-free biomedical imaging techniques using spatial sequencing technology. I will discuss quantitative approaches for image feature extraction in label-free images, how spatial sequencing can bridge the gap to comprehensive validation of this imaging technology, and our ongoing work to develop mathematical methods to bridge the gap between sequencing and imaging information.

Math Building, Room 402  https://map.arizona.edu/89