Search for a command to run...
Breast cancer is the second most common cancer diagnosed in women in the US with almost 280,000 new cases anticipated in 2023. Currently, on-site pathology for location guidance is not available during the collection of breast biopsies or during surgical intervention procedures. This shortcoming contributes to repeat biopsy and re-excision procedures, increasing the cost and patient discomfort during the cancer management process. Both procedures could benefit from on-site feedback, but current clinical on-site evaluation techniques are not commonly used on breast tissue because they are destructive and inaccurate. Ex-vivo microscopy is an emerging field aimed at creating histology-analogous images from non- or minimally-processed tissues, and is a promising tool for addressing this pain point in clinical cancer management. We investigated the ability structured illumination microscopy (SIM) to generate images from freshly-obtained breast tissues for structure identification and cancer identification at a speed compatible with potential on-site clinical implementation. We imaged 47 biopsies from patients undergoing a guided breast biopsy procedure using a customized SIM system and a dual-color fluorescent hematoxylin & eosin (H&E) analog. These biopsies had an average size of 0.92 cm2 (minimum 0.1, maximum 4.2) and had an average imaging time of 7:29 (minimum 0:22, maximum 37:44). After imaging, breast biopsies were submitted for standard histopathological processing and review. A board-certified pathologist returned a binary diagnostic accuracy of 96% when compared to diagnoses from gold-standard histology slides, and key tissue features including stroma, vessels, ducts, and lobules were identified from the resulting images.