Analysis of Glioblastoma Whole Slide Images using deep learning computer vision methods

Project to use slide images of recessed tumours in training of deep learning algorithms to identify methylation, primary or recurrence among other qualities

Features

• IVY GAP dataset web scraping tool
• Various training methods supported (VGG16, VGG19, ENET, UNET{coming soon!}, ConvNet)
• Pre-processing pipeline included for image modification and dataset augmentation
• Sample datasets and training routines

Tech Stack

Client: CUDA enabled Nvidia GPU (can be modified to accomodate other setups)

contact

For support, email ramu@mcmaster.ca For more info visit https://sites.google.com/view/udbhav-ram

Acknowledgements

• Ivy Glioblastoma Atlas Project