Comprehensive Histopathology Imaging in Pancreatic Biopsies: High Definition Infrared Imaging with Machine Learning Approach

Infrared (IR) based histopathology offers a new paradigm in looking at tissues and can provide a complimentary information source for more classical histopathology, which makes it a noteworthy tool given possible clinical application. This study aims to build a robust, pixel level machine learning model for pancreatic cancer detection using IR imaging. In this article, we report a pancreatic cancer classification model based on data from over 600 biopsies (coming from 250 patients) imaged with IR diffraction-limited spatial resolution. To fully research model's classification ability, we measured tissues using two optical setups, resulting in Standard and High Definitions data. This forms one of the largest IR datasets analyzed up to now, with almost 700 million spectra of different tissue types. The first six-class model created for comprehensive histopathology achieved pixel (tissue) level AUC values above 0.95, giving a successful technique for digital staining with biochemical information extracted from IR spectra.

Automated summary

Infrared (IR) based histopathology is a valuable tool for studying tissues and has potential clinical applications. This study develops a robust machine learning model for pancreatic cancer detection using IR imaging. By analyzing a large dataset of over 600 biopsies, this model achieved high accuracy in classifying pancreatic cancer at a pixel level, demonstrating successful digital staining with biochemical information extracted from IR spectra.