Generative Network Utilizing Random Number Based on Memristor Array for Medical Image Synthesis

In this work, we demonstrate that realistic chest X-ray images can be generated using a generative adversarial network (GAN) model with true-random numbers (TRNs) generated from a memristor array. The high bandgap active materials of memristor enhanced the entropy of the generated TRN via tunneling conduction, enabling the successful generation of high-quality medical images