|
Currently available pathology techniques for obtaining a rapid tissue diagnosis, or for determining the adequacy of specimens intended for downstream analysis, are too slow, labor-intensive, and destructive for point-of-care (POC) applications. We previously demonstrated video-rate structured illumination microscopy (VR-SIM) for accurate, high-throughput, non-destructive diagnostic imaging of fluorescently-stained prostate biopsies in seconds per biopsy, with an area under the ROC curve of 0.82-0.88 after pathologist review. In addition, we have demonstrated that it is feasible to use VR-SIM to routinely image very large gross pathology specimens, such as entire prostate resection surfaces, in relatively short timeframes at subcellular resolution. However, our prior work has focused on applications in prostate cancer; the utility in other organ sites has not been explored.
Here we extended our technology to varying size kidney, liver, and lung biopsies. We conducted a validation study of VR-SIM against histopathology on a variety of human tissues, including both small biopsies and large slices of tissue. We conducted a blinded study in which the study pathologist accurately identified the organs based on VR-SIM images alone. The results were then used to create a clinical atlas between VR-SIM and H and E images for the different tissues of interest. This clinical atlas will be used to aid in pathologist interpretation in future POC clinical applications of VR-SIM in kidney, liver, and lung. Such applications could include on-site identification of the presence of kidney glomeruli for to ensure successful downstream IHC analysis, or determination of the adequacy of lung cancer biopsies for genomic analysis.
|
