Real-time intraoperative diagnosis by deep neural network driven multiphoton virtual histology

摘要

Proposed workflow for real-time intraoperative diagnosis during breast cancer surgeries. Instead of waiting days for histological and histochemical processing, we propose to image the tissue immediately after it is resected, or even directly in the cavity, and then analyze the images in real-time with a DNN-based classifier, thus generating results (probability of being cancerous) within minutes.