The Applied Bioinformatics Laboratories can train deep neural networks to automatically classify images obtained from histopathology, MRI, microscopy, and other sources. 

Request a Histopathology Service

To request this service from us, please provide the following:

• as many images as possible (100-plus images for each class)
• a sample sheet with the different labels associated with each case/patient

You receive the following from us:

• a study on how deep-learning tool (Inception-v3 architecture) performs on the selected images 
• the probabilities associated with images on pretrained network

Additional or Customized Analyses

We can also study the possibility of training a convolutional neural network for the automatic classification and/or diagnosis or mutation predictions from other types of images.

Additional Resources

Below is a link to a computational pipeline that we typically use for our analyses:

• DeepPATH pipeline

The following select publication has used our analyses:

• Coudray N, Ocampo PS, Sakellaropoulos T, Narula N, Snuderl M, Fenyö D, et al. Classification and mutation prediction from non-small cell lung cancer histopathology images using deep learning. Nature Medicine. 2018;24(10):1559-1567.