Insiders Hub

Medical Image Generation

Bhavya Shah

Bhavya Shah

Unknown

0 0
  • 0 Collaborators

The project is aimed at the generation of synthetic medical images using Generative Adversarial Networks (GANs) for the development of medical imaging databases which in the real world is scarce, expensive and bonded with legal concerns regarding patient privacy. ...learn more

Project status: Concept

Artificial Intelligence

Intel Technologies
Intel Integrated Graphics

Code Samples [1]

Overview / Usage

The aim of developing autonomous models for classification of diseases or disorders can be achieved by producing more data for the data-hungry neural networks. This project aims at the generation of synthetic medical images in several medical divisions like skin melanoma, benign lesions, CT – PET scan images, retinal fundi, brain segmentation, etc.

Methodology / Approach

I implemented a simple deep convolutional GAN for generating malignant skin melanoma images by training on the ISIC 2017 database. Further, I aim to generate high-resolution skin lesion images using Progressive Generative Adversarial Networks.

Technologies Used

Python - OpenCV, numpy, Tensorflow, scikitlearn, matplotlib, pandas, glob etc..

Hardware: Tesla K80 GPU based PC

Repository

https://drive.google.com/open?id=1ekYxjEQKAu9HM_jcVhBEGb2JAt5Hry-N

Comments (0)