CycleGAN Models for MRI Image Translation
Cassandra Czobit, Reza Samavi
TL;DR
The study addresses the challenge of augmenting small MRI/DTI datasets by translating neuroimages between MRI field strengths using CycleGAN, comparing it to DCGAN. It demonstrates that CycleGAN achieves meaningful intra-modality translations with PSNR reaching 25.69–27.22 dB and MAE/MSE in the low-to-mid thousands range, while DCGAN yields poor image quality and limited diversity. This work suggests CycleGAN as a viable data-augmentation tool for field-strength translation, potentially improving robustness and privacy-preserving synthetic data. Future work should scale up datasets and incorporate classifier-based evaluation to validate image realism and domain alignment more rigorously.
Abstract
Image-to-image translation has gained popularity in the medical field to transform images from one domain to another. Medical image synthesis via domain transformation is advantageous in its ability to augment an image dataset where images for a given class is limited. From the learning perspective, this process contributes to data-oriented robustness of the model by inherently broadening the model's exposure to more diverse visual data and enabling it to learn more generalized features. In the case of generating additional neuroimages, it is advantageous to obtain unidentifiable medical data and augment smaller annotated datasets. This study proposes the development of a CycleGAN model for translating neuroimages from one field strength to another (e.g., 3 Tesla to 1.5). This model was compared to a model based on DCGAN architecture. CycleGAN was able to generate the synthetic and reconstructed images with reasonable accuracy. The mapping function from the source (3 Tesla) to target domain (1.5 Tesla) performed optimally with an average PSNR value of 25.69 $\pm$ 2.49 dB and an MAE value of 2106.27 $\pm$ 1218.37.