Staingan: Stain Style Transfer for Digital Histological Images

By M. Shaban, Christoph Baur, Nassir Navab, Shadi Albarqouni. 2018

327 citations recorded by the source corpus.

Digitized Histological diagnosis is in increasing demand. However, color variations due to various factors are imposing obstacles to the diagnosis process. The problem of stain color variations is a well-defined problem with many proposed solutions. Most of these solutions are highly dependent on a reference template slide. We propose a deep-learning solution inspired by CycleGANs that is trained end-to-end, eliminating the need for an expert to pick a representative reference slide. Our approach showed superior results quantitatively and qualitatively against the state of the art methods (10% improvement visually using SSIM). We further validated our method on a clinical use-case, namely Breast Cancer tumor classification, showing a 12% increase in AUC. The code is made publicly available 1.1https://github.com/xtarx/StainGAN

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Style transfer-based image synthesis as an efficient regularization technique in deep learning (2019, background): “convert sketches to images) [24], image blending (mixing selected parts of two images to get a new one) [25], image inpainting (restoring missing pieces of an image) [26] and unconditional [27] and conditional synthesis [28].”
GANs for Medical Image Analysis (2018, methodology): “[47] alleviates the need for paired training data and co-registration by employing CycleGANs for the task of stain transfer.”
The impact of pre- and post-image processing techniques on deep learning frameworks: A comprehensive review for digital pathology image analysis (2020, methodology): “BenTaieb et al. [64] built a discriminative model with an intrinsic stain normalization component while Shaban et al. [65] designed a GAN for stain normalization without the need to pick a template image.”
Artificial intelligence in oncology (2020, methodology): “34 Some automated CNN-based tools, such as HistoQC and DeepFocus, 35,36 have been developed to standardize the quality of whole-slide imaging.”
Digital Pathology and Artificial Intelligence (2019, methodology): “In [9], the authors used the concept of Cycle-Consistent Adversarial Networks [16] to perform color normalization.”
Algorithm fairness in artificial intelligence for medicine and healthcare (2023, background): “For instance, deep-learning techniques using GANs can learn stain features as a form of ‘style transfer’, in which a GAN can be used to pre-process data at deployment time to match the training distribution 179,180 .”
Measuring Domain Shift for Deep Learning in Histopathology (2020, methodology): “[34] present two different ways of doing stain normalization (the latter a learning based method similar to [22]), and argue the importance of normalization rather than color augmentation.”
Deep neural network models for computational histopathology: A survey (2019, background): “…based regional Gleason pattern classification+ k-nearest-neighbor based Gleason grade prediction TCGA (397 cases) + two private sets (361 + 11 cases) Shaban et al. (2019a) (X) Mouth H&E tumour infiltrating lympho- cytes abundance score prediction for disease free survival CNN (MobileNet) based…”
Synthetic Data for Deep Learning (2019, background): “H&E staining, including transformation from unstained to stained lung histology images [43] and stain style transfer [538], multi-contract MRI (from contrast to contrast) [138], 3D cross-modality MRI [677], different styles of prostate histopathology [488], different datasets of chest X-rays [98], a”