🤖 AI Summary
This study addresses two critical challenges in virtual histological staining: low pathological fidelity and the weak correlation between conventional image quality metrics (e.g., SSIM, PSNR) and clinical diagnostic utility. To this end, we propose CSSP2P GAN—a conditional generative adversarial network incorporating a customized adversarial loss to enhance structural integrity and biomarker expression fidelity in virtual immunohistochemical (IHC) staining. Moving beyond pixel-level evaluation, we establish a blinded expert pathology assessment framework to quantitatively measure diagnostic confidence. Experiments across multiple datasets demonstrate that CSSP2P GAN significantly outperforms baseline methods—including CycleGAN and Pix2Pix—with a 23.6% improvement in inter-expert diagnostic agreement under blinded evaluation. Furthermore, statistical analysis confirms negligible correlation between SSIM/PSNR scores and pathologists’ diagnostic judgments (r < 0.15), underscoring the necessity of clinically grounded evaluation paradigms for virtual staining.
📝 Abstract
In addition to evaluating tumor morphology using H&E staining, immunohistochemistry is used to assess the presence of specific proteins within the tissue. However, this is a costly and labor-intensive technique, for which virtual staining, as an image-to-image translation task, offers a promising alternative. Although recent, this is an emerging field of research with 64% of published studies just in 2024. Most studies use publicly available datasets of H&E-IHC pairs from consecutive tissue sections. Recognizing the training challenges, many authors develop complex virtual staining models based on conditional Generative Adversarial Networks, but ignore the impact of adversarial loss on the quality of virtual staining. Furthermore, overlooking the issues of model evaluation, they claim improved performance based on metrics such as SSIM and PSNR, which are not sufficiently robust to evaluate the quality of virtually stained images. In this paper, we developed CSSP2P GAN, which we demonstrate to achieve heightened pathological fidelity through a blind pathological expert evaluation. Furthermore, while iteratively developing our model, we study the impact of the adversarial loss and demonstrate its crucial role in the quality of virtually stained images. Finally, while comparing our model with reference works in the field, we underscore the limitations of the currently used evaluation metrics and demonstrate the superior performance of CSSP2P GAN.