Virtual Staining Tutorial: KI67 Synthesis with Deep Learning

This tutorial demonstrates how to build and train a deep learning model to virtually stain hematoxylin and eosin (H&E) images into corresponding immunohistochemistry (IHC) stained images using the KI67 marker.

We will walk through:

• Setting up the dataset and preprocessing
• Defining the generator and discriminator models
• Training with a combination of adversarial, pixel-wise, and perceptual losses
• Evaluating and visualizing the results

Dataset

Data for this tutorial can be accessed from this Google Drive link. It contains:

• Testing images for evaluation
• Samples generated during validation
• Model checkpoints