Attenuation Map Synthesis with SPECT Images Using Generative Adversarial Networks

This is an implementation of the attenuation map synthesis algorithm described in the following paper:

Shi, Luyao, et al. "Generating Attenuation Map for SPECT-only systems using Generative Adversarial Networks." Journal of Nuclear Medicine 60 (supplement 1): 572.

Main script: SPECTGan_phantom.py

Sample results

Network was trained with a very small amount of images.

Trainable parameters of the discriminator: 462,721.

Trainable parameters of the generator: 3,833,713.

About Myself