FIGURE 2 from Accelerating Cancer Histopathology Workflows with Chemical Imaging and Machine Learning
Year: 2023
Type: other
Abstract: Recorded data and predicted images in workflow. A, An overview of the proposed approach, acquiring multidepth data from surgical resections and using DL frameworks to generate H&E-stained images from SRSM data from FF tissue (G1) and FFPE-equivalent images (G2). B, Concept diagram of the staining model (G1) that converts SRSM data to H&E images obtained from FF tissues (virtual FF). C, Concept diagram for the model (G2) to convert virtual staining images to FFPE-like images (virtual FFPE). D1, D2, and D3 are the discriminators. G3 is generating virtual FF samples given FFPE samples, which is part of the training algorithm.
Language: English
Topic: AI in cancer detection
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4386828969