Methods for imaging and analysis of hystological tissue slides via Fourier Ptychographic Microscopy

Digital pathology indicates the ensemble of emerging methods, analysis techniques, and diagnostic protocols that aim to improve and automatize the inspection of biological samples such as slides of histological tissues and blood smears. From the microscopy methods standpoint, accessing a wide field of view (FOV) is highly required to look at the big picture of tissue organization, structures and connections. At the same time, high resolution is required to inspect the slide at the single cell, organelle and membranes level. Here, we show the use of Fourier Ptychographic Microscopy (FPM) to image effectively large tissue slides portions, over a wide mm² FOV, while attaining submicron resolution. Resolution is improved up to a six-fold factor by probing the specimen from multiple directions and exploiting a synthetic aperture principle. An iterative phase retrieval algorithm estimated the high-resolution phase-contrast from a set of low-resolution intensities. Remarkably, FPM is a marker-free and quantitative method. Neural and kidney tissue slides, both unstained and marked with different stains are imaged by FPM to investigate the effect of staining on the phase-contrast images. We show novel tissue analysis and classification techniques working on FPM images. These are specifically tailored for exploiting the multi-scale imaging capabilities of FPM and can be used to classify tissue portions working across the scales, i.e. analysing the phase-contrast signatures at the tissue organization level up to the single membrane level. We apply this novel approach to classify tissue slides from patients with breast cancer and fibroadenoma tissue.