Synthetic ERG waveform Generation

In our latest publication: Kulyabin, M., Zhdanov, A., Lee, I.O. et al. Synthetic electroretinogram signal generation using a conditional generative adversarial network. Doc Ophthalmol (2025). https://doi.org/10.1007/s10633-025-10019-0

We have used an adversarial network that generates sybtgeitic of 'fake' ERG waveforms natural or real ERG waveforms with a Discriminator. The increase in sample size enables teh augmentation of the origibnal dataset to improve AI modelling for ASD classification in this case.

Synthetic ERG signal generation may help improve modelling of rare disorders where availability of datasets may be limited. Generative Adversarila Networks have been ued in allied health fiekls such as cardiology but this is teh first implementation with ERGs