Applications of artificial intelligence for image enhancement in pathology

Published in Artificial Intelligence and Deep Learning in Pathology (Elsevier), 2020

Pathology is finally profiting from the recent development of promising novel imaging techniques, as well as exploration of computational methods for extracting increased information from existing slides and conventional microscopy. This chapter focuses largely on the latter, looking at application of artificial intelligence (AI) methods for denoising, sharpening, super-resolution, and color mapping from unstained slides to generate virtually stained images. In addition, software-based tools have been developed to replicate histochemical and immunohistochemical results without having to perform the actual procedures. Finally, an example of an application of novel slide-free histology using stimulated Raman microscopy for intraoperative guidance during brain cancer surgery is presented remarkably; this method includes AI classification tools (cancer/not cancer) that surgeons can rely on directly.

View book chapter here