GAN-enhanced non-contrast CT radiomics for gastric cancer diagnosis across four centers

Synthetic contrast-enhanced CT from a GAN improved gastric cancer histologic grade prediction AUC by 5.9% (internal) and 0.7% (external) over non-contrast CT in a 1,757-patient multicenter study. Lauren classification prediction improved 6.8% and 7.3%.

• The proposed model significantly outperformed Pix2Pix, CycleGAN, and BBDM on image quality metrics (MAE, MSE, PSNR, SSIM) in internal and external test sets.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.