A study about color normalization methods for histopathology images

Investigation on potential bias factors in histopathology datasets

Deep neural networks (DNNs) have demonstrated remarkable capabilities in medical applications, including digital pathology, where they excel at analyzing complex patterns in medical images to assist in accurate disease diagnosis and prognosis. However, concerns have arisen about potential biases in The Cancer Genome Atlas (TCGA) dataset, a comprehensive repository of digitized histopathology data and serves as both a training and validation source for deep learning models, suggesting that over-optimistic results of model performance may be due to reliance on biased features rather than histological characteristics. Surprisingly, recent studies have confirmed the existence of site-specific bias in the embedded features extracted for cancer-type discrimination, leading to high accuracy in acquisition site classification. This biased behavior motivated us to conduct an in-depth analysis to investigate potential causes behind this unexpected biased ability toward site-specific pattern recognition. The analysis was conducted on two cutting-edge DNN models: KimiaNet, a state-of-the-art DNN trained on TCGA images, and the self-trained EfficientNet. In this research study, the balanced accuracy metric is used to evaluate the performance of a model trained to classify data centers, which was originally designed to learn cancerous patterns, with the aim of investigating the potential factors contributing to the higher balanced accuracy in data center detection.

An international study of stain variability in histopathology using qualitative and quantitative analysis

Hematoxylin and eosin (H&E) staining accounts for over 80% of slides stained worldwide. Although routinely used, there are high levels of variation between labs due to different staining methods. Staining is a pivotal part of slide preparation, but quality control is largely subjective, with overall clinical assurance provided by external quality assessment (EQA) services, underpinned by expert assessment. Digital pathology offers the potential to provide objective quantification of stain, through color analysis, to augment EQA assessment. This large-scale study evaluated H&E staining in 247 international labs participating in the UK NEQAS CPT EQA programme. Tissue sections were circulated to each lab to stain using their routine H&E staining protocol. The slides were reviewed by independent expert UK NEQAS CPT assessors, and quantitative digital analysis was conducted, comprising of H&E color deconvolution and color difference determination (ΔE). Most labs (69%) achieved an EQA score indicating good or excellent staining, with high inter-observer concordance to support this (92.5% within one mark of each other). H&E color difference, ΔE, showed 60% of labs were within 2 ΔE of the mean, which is considered as only perceptible through close observation. There was little correlation found between H&E intensity and assessor score, however, the H&E intensity ratio indicated a trend with assessor score suggesting there may be an optimal stain relationship that should be investigated further. The presented hybrid analysis combines expert analysis with objective data. This has the potential to inform upon optimal tissue staining and allows us to consider quantitative standards of H&E staining in pathology practice.

Learnable color space conversion and fusion for stain normalization in pathology images

Variations in hue and contrast are common in H&E-stained pathology images due to differences in slide preparation across various institutions. Such stain variations, while not affecting pathologists much in diagnosing the biopsy, pose significant challenges for computer-assisted diagnostic systems, leading to potential underdiagnosis or misdiagnosis, especially when stain differentiation introduces substantial heterogeneity across datasets from different sources. Traditional stain normalization methods, aimed at mitigating these issues, often require labor-intensive selection of appropriate templates, limiting their practicality and automation. Innovatively, we propose a Learnable Stain Normalization layer, i.e. LStainNorm, designed as an easily integrable component for pathology image analysis. It minimizes the need for manual template selection by autonomously learning the optimal stain characteristics. Moreover, the learned optimal stain template provides the interpretability to enhance the understanding of the normalization process. Additionally, we demonstrate that fusing pathology images normalized in multiple color spaces can improve performance. Therefore, we extend LStainNorm with a novel self-attention mechanism to facilitate the fusion of features across different attributes and color spaces. Experimentally, LStainNorm outperforms the state-of-the-art methods including conventional ones and GANs on two classification datasets and three nuclei segmentation datasets by an average increase of 4.78% in accuracy, 3.53% in Dice coefficient, and 6.59% in IoU. Additionally, by enabling an end-to-end training and inference process, LStainNorm eliminates the need for intermediate steps between normalization and analysis, resulting in more efficient use of hardware resources and significantly faster inference time, i.e up to hundreds of times quicker than traditional methods. The code is publicly available at https://github.com/yjzscode/Optimal-Normalisation-in-Color-Spaces.

A High-Resolution Digital Pathological Image Staining Style Transfer Model Based on Gradient Guidance

Digital pathology images have long been regarded as the gold standard for cancer diagnosis in clinical medicine. A highly generalized digital pathological image diagnosis system can provide strong support for cancer diagnosis, help to improve the diagnostic efficiency and accuracy of doctors, and has important research value. The whole slide image of different centers can lead to very large staining differences due to different scanners and dyes, which pose a challenge to the generalization performance of the model application in multi-center data testing. In order to achieve the normalization of multi-center data, this paper proposes a style transfer algorithm based on an adversarial generative network for high-resolution images. The gradient-guided dye migration model proposed in this paper introduces a gradient-enhanced regularized term in the loss function design of the algorithm. A style transfer algorithm was applied to the source data, and the diagnostic performance of the multi-example learning model based on the domain data was significantly improved by validation in the pathological image datasets of two centers. The proposed method improved the AUC of the best classification model from 0.8856 to 0.9243, and another set of experiments improved the AUC from 0.8012 to 0.8313.

Image Analysis in Histopathology and Cytopathology: From Early Days to Current Perspectives

Both pathology and cytopathology still rely on recognizing microscopical morphologic features, and image analysis plays a crucial role, enabling the identification, categorization, and characterization of different tissue types, cell populations, and disease states within microscopic images. Historically, manual methods have been the primary approach, relying on expert knowledge and experience of pathologists to interpret microscopic tissue samples. Early image analysis methods were often constrained by computational power and the complexity of biological samples. The advent of computers and digital imaging technologies challenged the exclusivity of human eye vision and brain computational skills, transforming the diagnostic process in these fields. The increasing digitization of pathological images has led to the application of more objective and efficient computer-aided analysis techniques. Significant advancements were brought about by the integration of digital pathology, machine learning, and advanced imaging technologies. The continuous progress in machine learning and the increasing availability of digital pathology data offer exciting opportunities for the future. Furthermore, artificial intelligence has revolutionized this field, enabling predictive models that assist in diagnostic decision making. The future of pathology and cytopathology is predicted to be marked by advancements in computer-aided image analysis. The future of image analysis is promising, and the increasing availability of digital pathology data will invariably lead to enhanced diagnostic accuracy and improved prognostic predictions that shape personalized treatment strategies, ultimately leading to better patient outcomes.

Joint Image Processing with Learning-Driven Data Representation and Model Behavior for Non-Intrusive Anemia Diagnosis in Pediatric Patients

Anemia diagnosis is crucial for pediatric patients due to its impact on growth and development. Traditional methods, like blood tests, are effective but pose challenges, such as discomfort, infection risk, and frequent monitoring difficulties, underscoring the need for non-intrusive diagnostic methods. In light of this, this study proposes a novel method that combines image processing with learning-driven data representation and model behavior for non-intrusive anemia diagnosis in pediatric patients. The contributions of this study are threefold. First, it uses an image-processing pipeline to extract 181 features from 13 categories, with a feature-selection process identifying the most crucial data for learning. Second, a deep multilayered network based on long short-term memory (LSTM) is utilized to train a model for classifying images into anemic and non-anemic cases, where hyperparameters are optimized using Bayesian approaches. Third, the trained LSTM model is integrated as a layer into a learning model developed based on recurrent expansion rules, forming a part of a new deep network called a recurrent expansion network (RexNet). RexNet is designed to learn data representations akin to traditional deep-learning methods while also understanding the interaction between dependent and independent variables. The proposed approach is applied to three public datasets, namely conjunctival eye images, palmar images, and fingernail images of children aged up to 6 years. RexNet achieves an overall evaluation of 99.83 ± 0.02% across all classification metrics, demonstrating significant improvements in diagnostic results and generalization compared to LSTM networks and existing methods. This highlights RexNet's potential as a promising alternative to traditional blood-based methods for non-intrusive anemia diagnosis.

Improving performance in colorectal cancer histology decomposition using deep and ensemble machine learning

In routine colorectal cancer management, histologic samples stained with hematoxylin and eosin are commonly used. Nonetheless, their potential for defining objective biomarkers for patient stratification and treatment selection is still being explored. The current gold standard relies on expensive and time-consuming genetic tests. However, recent research highlights the potential of convolutional neural networks (CNNs) to facilitate the extraction of clinically relevant biomarkers from these readily available images. These CNN-based biomarkers can predict patient outcomes comparably to golden standards, with the added advantages of speed, automation, and minimal cost. The predictive potential of CNN-based biomarkers fundamentally relies on the ability of CNNs to accurately classify diverse tissue types from whole slide microscope images. Consequently, enhancing the accuracy of tissue class decomposition is critical to amplifying the prognostic potential of imaging-based biomarkers. This study introduces a hybrid deep transfer learning and ensemble machine learning model that improves upon previous approaches, including a transformer and neural architecture search baseline for this task. We employed a pairing of the EfficientNetV2 architecture with a random forest classification head. Our model achieved 96.74% accuracy (95% CI: 96.3%-97.1%) on the external test set and 99.89% on the internal test set. Recognizing the potential of these models in the task, we have made them publicly available.

Multi-domain stain normalization for digital pathology: A cycle-consistent adversarial network for whole slide images

The variation in histologic staining between different medical centers is one of the most profound challenges in the field of computer-aided diagnosis. The appearance disparity of pathological whole slide images causes algorithms to become less reliable, which in turn impedes the wide-spread applicability of downstream tasks like cancer diagnosis. Furthermore, different stainings lead to biases in the training which in case of domain shifts negatively affect the test performance. Therefore, in this paper we propose MultiStain-CycleGAN, a multi-domain approach to stain normalization based on CycleGAN. Our modifications to CycleGAN allow us to normalize images of different origins without retraining or using different models. We perform an extensive evaluation of our method using various metrics and compare it to commonly used methods that are multi-domain capable. First, we evaluate how well our method fools a domain classifier that tries to assign a medical center to an image. Then, we test our normalization on the tumor classification performance of a downstream classifier. Furthermore, we evaluate the image quality of the normalized images using the Structural similarity index and the ability to reduce the domain shift using the Fréchet inception distance. We show that our method proves to be multi-domain capable, provides a very high image quality among the compared methods, and can most reliably fool the domain classifier while keeping the tumor classifier performance high. By reducing the domain influence, biases in the data can be removed on the one hand and the origin of the whole slide image can be disguised on the other, thus enhancing patient data privacy.

Mitosis detection, fast and slow: Robust and efficient detection of mitotic figures

Counting of mitotic figures is a fundamental step in grading and prognostication of several cancers. However, manual mitosis counting is tedious and time-consuming. In addition, variation in the appearance of mitotic figures causes a high degree of discordance among pathologists. With advances in deep learning models, several automatic mitosis detection algorithms have been proposed but they are sensitive to domain shift often seen in histology images. We propose a robust and efficient two-stage mitosis detection framework, which comprises mitosis candidate segmentation (Detecting Fast) and candidate refinement (Detecting Slow) stages. The proposed candidate segmentation model, termed EUNet, is fast and accurate due to its architectural design. EUNet can precisely segment candidates at a lower resolution to considerably speed up candidate detection. Candidates are then refined using a deeper classifier network, EfficientNet-B7, in the second stage. We make sure both stages are robust against domain shift by incorporating domain generalization methods. We demonstrate state-of-the-art performance and generalizability of the proposed model on the three largest publicly available mitosis datasets, winning the two mitosis domain generalization challenge contests (MIDOG21 and MIDOG22). Finally, we showcase the utility of the proposed algorithm by processing the TCGA breast cancer cohort (1,124 whole-slide images) to generate and release a repository of more than 620K potential mitotic figures (not exhaustively validated).

Development and validation of a Radiopathomics model based on CT scans and whole slide images for discriminating between Stage I-II and Stage III gastric cancer

OBJECTIVE

This study aimed to develop and validate an artificial intelligence radiopathological model using preoperative CT scans and postoperative hematoxylin and eosin (HE) stained slides to predict the pathological staging of gastric cancer (stage I-II and stage III).

METHODS

This study included a total of 202 gastric cancer patients with confirmed pathological staging (training cohort: n = 141; validation cohort: n = 61). Pathological histological features were extracted from HE slides, and pathological models were constructed using logistic regression (LR), support vector machine (SVM), and NaiveBayes. The optimal pathological model was selected through receiver operating characteristic (ROC) curve analysis. Machine learnin algorithms were employed to construct radiomic models and radiopathological models using the optimal pathological model. Model performance was evaluated using ROC curve analysis, and clinical utility was estimated using decision curve analysis (DCA).

RESULTS

A total of 311 pathological histological features were extracted from the HE images, including 101 Term Frequency-Inverse Document Frequency (TF-IDF) features and 210 deep learning features. A pathological model was constructed using 19 selected pathological features through dimension reduction, with the SVM model demonstrating superior predictive performance (AUC, training cohort: 0.949; validation cohort: 0.777). Radiomic features were constructed using 6 selected features from 1834 radiomic features extracted from CT scans via SVM machine algorithm. Simultaneously, a radiopathomics model was built using 17 non-zero coefficient features obtained through dimension reduction from a total of 2145 features (combining both radiomics and pathomics features). The best discriminative ability was observed in the SVM_radiopathomics model (AUC, training cohort: 0.953; validation cohort: 0.851), and clinical decision curve analysis (DCA) demonstrated excellent clinical utility.

CONCLUSION

The radiopathomics model, combining pathological and radiomic features, exhibited superior performance in distinguishing between stage I-II and stage III gastric cancer. This study is based on the prediction of pathological staging using pathological tissue slides from surgical specimens after gastric cancer curative surgery and preoperative CT images, highlighting the feasibility of conducting research on pathological staging using pathological slides and CT images.

A multi-class brain tumor grading system based on histopathological images using a hybrid YOLO and RESNET networks

Gliomas are primary brain tumors caused by glial cells. These cancers' classification and grading are crucial for prognosis and treatment planning. Deep learning (DL) can potentially improve the digital pathology investigation of brain tumors. In this paper, we developed a technique for visualizing a predictive tumor grading model on histopathology pictures to help guide doctors by emphasizing characteristics and heterogeneity in forecasts. The proposed technique is a hybrid model based on YOLOv5 and ResNet50. The function of YOLOv5 is to localize and classify the tumor in large histopathological whole slide images (WSIs). The suggested technique incorporates ResNet into the feature extraction of the YOLOv5 framework, and the detection results show that our hybrid network is effective for identifying brain tumors from histopathological images. Next, we estimate the glioma grades using the extreme gradient boosting classifier. The high-dimensional characteristics and nonlinear interactions present in histopathology images are well-handled by this classifier. DL techniques have been used in previous computer-aided diagnosis systems for brain tumor diagnosis. However, by combining the YOLOv5 and ResNet50 architectures into a hybrid model specifically designed for accurate tumor localization and predictive grading within histopathological WSIs, our study presents a new approach that advances the field. By utilizing the advantages of both models, this creative integration goes beyond traditional techniques to produce improved tumor localization accuracy and thorough feature extraction. Additionally, our method ensures stable training dynamics and strong model performance by integrating ResNet50 into the YOLOv5 framework, addressing concerns about gradient explosion. The proposed technique is tested using the cancer genome atlas dataset. During the experiments, our model outperforms the other standard ways on the same dataset. Our results indicate that the proposed hybrid model substantially impacts tumor subtype discrimination between low-grade glioma (LGG) II and LGG III. With 97.2% of accuracy, 97.8% of precision, 98.6% of sensitivity, and the Dice similarity coefficient of 97%, the proposed model performs well in classifying four grades. These results outperform current approaches for identifying LGG from high-grade glioma and provide competitive performance in classifying four categories of glioma in the literature.

A Robust Method for the Unsupervised Scoring of Immunohistochemical Staining

Immunohistochemistry is a powerful technique that is widely used in biomedical research and clinics; it allows one to determine the expression levels of some proteins of interest in tissue samples using color intensity due to the expression of biomarkers with specific antibodies. As such, immunohistochemical images are complex and their features are difficult to quantify. Recently, we proposed a novel method, including a first separation stage based on non-negative matrix factorization (NMF), that achieved good results. However, this method was highly dependent on the parameters that control sparseness and non-negativity, as well as on algorithm initialization. Furthermore, the previously proposed method required a reference image as a starting point for the NMF algorithm. In the present work, we propose a new, simpler and more robust method for the automated, unsupervised scoring of immunohistochemical images based on bright field. Our work is focused on images from tumor tissues marked with blue (nuclei) and brown (protein of interest) stains. The new proposed method represents a simpler approach that, on the one hand, avoids the use of NMF in the separation stage and, on the other hand, circumvents the need for a control image. This new approach determines the subspace spanned by the two colors of interest using principal component analysis (PCA) with dimension reduction. This subspace is a two-dimensional space, allowing for color vector determination by considering the point density peaks. A new scoring stage is also developed in our method that, again, avoids reference images, making the procedure more robust and less dependent on parameters. Semi-quantitative image scoring experiments using five categories exhibit promising and consistent results when compared to manual scoring carried out by experts.

Shedding light on the black box of a neural network used to detect prostate cancer in whole slide images by occlusion-based explainability

Diagnostic histopathology faces increasing demands due to aging populations and expanding healthcare programs. Semi-automated diagnostic systems employing deep learning methods are one approach to alleviate this pressure. The learning models for histopathology are inherently complex and opaque from the user's perspective. Hence different methods have been developed to interpret their behavior. However, relatively limited attention has been devoted to the connection between interpretation methods and the knowledge of experienced pathologists. The main contribution of this paper is a method for comparing morphological patterns used by expert pathologists to detect cancer with the patterns identified as important for inference of learning models. Given the patch-based nature of processing large-scale histopathological imaging, we have been able to show statistically that the VGG16 model could utilize all the structures that are observable by the pathologist, given the patch size and scan resolution. The results show that the neural network approach to recognizing prostatic cancer is similar to that of a pathologist at medium optical resolution. The saliency maps identified several prevailing histomorphological features characterizing carcinoma, e.g., single-layered epithelium, small lumina, and hyperchromatic nuclei with halo. A convincing finding was the recognition of their mimickers in non-neoplastic tissue. The method can also identify differences, i.e., standard patterns not used by the learning models and new patterns not yet used by pathologists. Saliency maps provide added value for automated digital pathology to analyze and fine-tune deep learning systems and improve trust in computer-based decisions.

Artifact Detection and Restoration in Histology Images With Stain-Style and Structural Preservation

The artifacts in histology images may encumber the accurate interpretation of medical information and cause misdiagnosis. Accordingly, prepending manual quality control of artifacts considerably decreases the degree of automation. To close this gap, we propose a methodical pre-processing framework to detect and restore artifacts, which minimizes their impact on downstream AI diagnostic tasks. First, the artifact recognition network AR-Classifier first differentiates common artifacts from normal tissues, e.g., tissue folds, marking dye, tattoo pigment, spot, and out-of-focus, and also catalogs artifact patches by their restorability. Then, the succeeding artifact restoration network AR-CycleGAN performs de-artifact processing where stain styles and tissue structures can be maximally retained. We construct a benchmark for performance evaluation, curated from both clinically collected WSIs and public datasets of colorectal and breast cancer. The functional structures are compared with state-of-the-art methods, and also comprehensively evaluated by multiple metrics across multiple tasks, including artifact classification, artifact restoration, downstream diagnostic tasks of tumor classification and nuclei segmentation. The proposed system allows full automation of deep learning based histology image analysis without human intervention. Moreover, the structure-independent characteristic enables its processing with various artifact subtypes. The source code and data in this research are available at https://github.com/yunboer/AR-classifier-and-AR-CycleGAN.

Evaluating the effectiveness of stain normalization techniques in automated grading of invasive ductal carcinoma histopathological images

Debates persist regarding the impact of Stain Normalization (SN) on recent breast cancer histopathological studies. While some studies propose no influence on classification outcomes, others argue for improvement. This study aims to assess the efficacy of SN in breast cancer histopathological classification, specifically focusing on Invasive Ductal Carcinoma (IDC) grading using Convolutional Neural Networks (CNNs). The null hypothesis asserts that SN has no effect on the accuracy of CNN-based IDC grading, while the alternative hypothesis suggests the contrary. We evaluated six SN techniques, with five templates selected as target images for the conventional SN techniques. We also utilized seven ImageNet pre-trained CNNs for IDC grading. The performance of models trained with and without SN was compared to discern the influence of SN on classification outcomes. The analysis unveiled a p-value of 0.11, indicating no statistically significant difference in Balanced Accuracy Scores between models trained with StainGAN-normalized images, achieving a score of 0.9196 (the best-performing SN technique), and models trained with non-normalized images, which scored 0.9308. As a result, we did not reject the null hypothesis, indicating that we found no evidence to support a significant discrepancy in effectiveness between stain-normalized and non-normalized datasets for IDC grading tasks. This study demonstrates that SN has a limited impact on IDC grading, challenging the assumption of performance enhancement through SN.

Accounting for intensity variation in image analysis of large-scale multiplexed clinical trial datasets

Multiplex immunofluorescence (mIF) imaging can provide comprehensive quantitative and spatial information for multiple immune markers for tumour immunoprofiling. However, application at scale to clinical trial samples sourced from multiple institutions is challenging due to pre-analytical heterogeneity. This study reports an analytical approach to the largest multi-parameter immunoprofiling study of clinical trial samples to date. We analysed 12,592 tissue microarray (TMA) spots from 3,545 colorectal cancers sourced from more than 240 institutions in two clinical trials (QUASAR 2 and SCOT) stained for CD4, CD8, CD20, CD68, FoxP3, pan-cytokeratin, and DAPI by mIF. TMA slides were multi-spectrally imaged and analysed by cell-based and pixel-based marker analysis. We developed an adaptive thresholding method to account for inter- and intra-slide intensity variation in TMA analysis. Applying this method effectively ameliorated inter- and intra-slide intensity variation improving the image analysis results compared with methods using a single global threshold. Correlation of CD8 data derived by our mIF analysis approach with single-plex chromogenic immunohistochemistry CD8 data derived from subsequent sections indicates the validity of our method (Spearman's rank correlation coefficients ρ between 0.63 and 0.66, p ≪ 0.01) as compared with the current gold standard analysis approach. Evaluation of correlation between cell-based and pixel-based analysis results confirms equivalency (ρ > 0.8, p ≪ 0.01, except for CD20 in the epithelial region) of both analytical approaches. These data suggest that our adaptive thresholding approach can enable analysis of mIF-stained clinical trial TMA datasets by digital pathology at scale for precision immunoprofiling.

Preparing Data for Artificial Intelligence in Pathology with Clinical-Grade Performance

The pathology is decisive for disease diagnosis but relies heavily on experienced pathologists. In recent years, there has been growing interest in the use of artificial intelligence in pathology (AIP) to enhance diagnostic accuracy and efficiency. However, the impressive performance of deep learning-based AIP in laboratory settings often proves challenging to replicate in clinical practice. As the data preparation is important for AIP, the paper has reviewed AIP-related studies in the PubMed database published from January 2017 to February 2022, and 118 studies were included. An in-depth analysis of data preparation methods is conducted, encompassing the acquisition of pathological tissue slides, data cleaning, screening, and subsequent digitization. Expert review, image annotation, dataset division for model training and validation are also discussed. Furthermore, we delve into the reasons behind the challenges in reproducing the high performance of AIP in clinical settings and present effective strategies to enhance AIP's clinical performance. The robustness of AIP depends on a randomized collection of representative disease slides, incorporating rigorous quality control and screening, correction of digital discrepancies, reasonable annotation, and sufficient data volume. Digital pathology is fundamental in clinical-grade AIP, and the techniques of data standardization and weakly supervised learning methods based on whole slide image (WSI) are effective ways to overcome obstacles of performance reproduction. The key to performance reproducibility lies in having representative data, an adequate amount of labeling, and ensuring consistency across multiple centers. Digital pathology for clinical diagnosis, data standardization and the technique of WSI-based weakly supervised learning will hopefully build clinical-grade AIP.

A comprehensive survey of intestine histopathological image analysis using machine vision approaches

Colorectal Cancer (CRC) is currently one of the most common and deadly cancers. CRC is the third most common malignancy and the fourth leading cause of cancer death worldwide. It ranks as the second most frequent cause of cancer-related deaths in the United States and other developed countries. Histopathological images contain sufficient phenotypic information, they play an indispensable role in the diagnosis and treatment of CRC. In order to improve the objectivity and diagnostic efficiency for image analysis of intestinal histopathology, Computer-aided Diagnosis (CAD) methods based on machine learning (ML) are widely applied in image analysis of intestinal histopathology. In this investigation, we conduct a comprehensive study on recent ML-based methods for image analysis of intestinal histopathology. First, we discuss commonly used datasets from basic research studies with knowledge of intestinal histopathology relevant to medicine. Second, we introduce traditional ML methods commonly used in intestinal histopathology, as well as deep learning (DL) methods. Then, we provide a comprehensive review of the recent developments in ML methods for segmentation, classification, detection, and recognition, among others, for histopathological images of the intestine. Finally, the existing methods have been studied, and the application prospects of these methods in this field are given.

Histopathological Images Analysis and Predictive Modeling Implemented in Digital Pathology-Current Affairs and Perspectives

In modern clinical practice, digital pathology has an essential role, being a technological necessity for the activity in the pathological anatomy laboratories. The development of information technology has majorly facilitated the management of digital images and their sharing for clinical use; the methods to analyze digital histopathological images, based on artificial intelligence techniques and specific models, quantify the required information with significantly higher consistency and precision compared to that provided by optical microscopy. In parallel, the unprecedented advances in machine learning facilitate, through the synergy of artificial intelligence and digital pathology, the possibility of diagnosis based on image analysis, previously limited only to certain specialties. Therefore, the integration of digital images into the study of pathology, combined with advanced algorithms and computer-assisted diagnostic techniques, extends the boundaries of the pathologist's vision beyond the microscopic image and allows the specialist to use and integrate his knowledge and experience adequately. We conducted a search in PubMed on the topic of digital pathology and its applications, to quantify the current state of knowledge. We found that computer-aided image analysis has a superior potential to identify, extract and quantify features in more detail compared to the human pathologist's evaluating possibilities; it performs tasks that exceed its manual capacity, and can produce new diagnostic algorithms and prediction models applicable in translational research that are able to identify new characteristics of diseases based on changes at the cellular and molecular level.

CNN stability training improves robustness to scanner and IHC-based image variability for epithelium segmentation in cervical histology

Background

In digital pathology, image properties such as color, brightness, contrast and blurriness may vary based on the scanner and sample preparation. Convolutional Neural Networks (CNNs) are sensitive to these variations and may underperform on images from a different domain than the one used for training. Robustness to these image property variations is required to enable the use of deep learning in clinical practice and large scale clinical research.

Aims

CNN Stability Training (CST) is proposed and evaluated as a method to increase CNN robustness to scanner and Immunohistochemistry (IHC)-based image variability.

Methods

CST was applied to segment epithelium in immunohistological cervical Whole Slide Images (WSIs). CST randomly distorts input tiles and factors the difference between the CNN prediction for the original and distorted inputs within the loss function. CNNs were trained using 114 p16-stained WSIs from the same scanner, and evaluated on 6 WSI test sets, each with 23 to 24 WSIs of the same tissue but different scanner/IHC combinations. Relative robustness (rAUC) was measured as the difference between the AUC on the training domain test set (i.e., baseline test set) and the remaining test sets.

Results

Across all test sets, The AUC of CST models outperformed "No CST" models (AUC: 0.940-0.989 vs. 0.905-0.986, p < 1e - 8), and obtained an improved robustness (rAUC: [-0.038, -0.003] vs. [-0.081, -0.002]). At a WSI level, CST models showed an increase in performance in 124 of the 142 WSIs. CST models also outperformed models trained with random on-the-fly data augmentation (DA) in all test sets ([0.002, 0.021], p < 1e-6).

Conclusion

CST offers a path to improve CNN performance without the need for more data and allows customizing distortions to specific use cases. A python implementation of CST is publicly available at https://github.com/TIGACenter/CST_v1.

Color restoration based on digital pathology image

OBJECTIVE

Protective color restoration of faded digital pathology images based on color transfer algorithm.

METHODS

Twenty fresh tissue samples of invasive breast cancer from the pathology department of Qingdao Central Hospital in 2021 were screened. After HE staining, HE stained sections were irradiated with sunlight to simulate natural fading, and every 7 days was a fading cycle, and a total of 8 cycles were experienced. At the end of each cycle, the sections were digitally scanned to retain clear images, and the color changes of the sections during the fading process were recorded. The color transfer algorithm was applied to restore the color of the faded images; Adobe Lightroom Classic software presented the histogram of the image color distribution; UNet++ cell recognition segmentation model was used to identify the color restored images; Natural Image Quality Evaluator (NIQE), Information Entropy (Entropy), and Average Gradient (AG) were applied to evaluate the quality of the restored images.

RESULTS

The restored image color met the diagnostic needs of pathologists. Compared with the faded images, the NIQE value decreased (P<0.05), Entropy value increased (P<0.01), and AG value increased (P<0.01). The cell recognition rate of the restored image was significantly improved.

CONCLUSION

The color transfer algorithm can effectively repair faded pathology images, restore the color contrast between nucleus and cytoplasm, improve the image quality, meet the diagnostic needs and improve the cell recognition rate of the deep learning model.

Comparison of deep learning models for digital H&E staining from unpaired label-free multispectral microscopy images

BACKGROUND AND OBJECTIVE

This paper presents the quantitative comparison of three generative models of digital staining, also known as virtual staining, in H&E modality (i.e., Hematoxylin and Eosin) that are applied to 5 types of breast tissue. Moreover, a qualitative evaluation of the results achieved with the best model was carried out. This process is based on images of samples without staining captured by a multispectral microscope with previous dimensional reduction to three channels in the RGB range.

METHODS

The models compared are based on conditional GAN (pix2pix) which uses images aligned with/without staining, and two models that do not require image alignment, Cycle GAN (cycleGAN) and contrastive learning-based model (CUT). These models are compared based on the structural similarity and chromatic discrepancy between samples with chemical staining and their corresponding ones with digital staining. The correspondence between images is achieved after the chemical staining images are subjected to digital unstaining by means of a model obtained to guarantee the cyclic consistency of the generative models.

RESULTS

The comparison of the three models corroborates the visual evaluation of the results showing the superiority of cycleGAN both for its larger structural similarity with respect to chemical staining (mean value of SSIM ∼ 0.95) and lower chromatic discrepancy (10%). To this end, quantization and calculation of EMD (Earth Mover's Distance) between clusters is used. In addition, quality evaluation through subjective psychophysical tests with three experts was carried out to evaluate quality of the results with the best model (cycleGAN).

CONCLUSIONS

The results can be satisfactorily evaluated by metrics that use as reference image a chemically stained sample and the digital staining images of the reference sample with prior digital unstaining. These metrics demonstrate that generative staining models that guarantee cyclic consistency provide the closest results to chemical H&E staining that also is consistent with the result of qualitative evaluation by experts.

Enhanced Pathology Image Quality with Restore-Generative Adversarial Network

Whole slide imaging is becoming a routine procedure in clinical diagnosis. Advanced image analysis techniques have been developed to assist pathologists in disease diagnosis, staging, subtype classification, and risk stratification. Recently, deep learning algorithms have achieved state-of-the-art performances in various imaging analysis tasks, including tumor region segmentation, nuclei detection, and disease classification. However, widespread clinical use of these algorithms is hampered by their performances often degrading due to image quality issues commonly seen in real-world pathology imaging data such as low resolution, blurring regions, and staining variation. Restore-Generative Adversarial Network (GAN), a deep learning model, was developed to improve the imaging qualities by restoring blurred regions, enhancing low resolution, and normalizing staining colors. The results demonstrate that Restore-GAN can significantly improve image quality, which leads to improved model robustness and performance for existing deep learning algorithms in pathology image analysis. Restore-GAN has the potential to be used to facilitate the applications of deep learning models in digital pathology analyses.

Deep Learning Prediction of TERT Promoter Mutation Status in Thyroid Cancer Using Histologic Images

Background and objectives: Telomerase reverse transcriptase (TERT) promoter mutation, found in a subset of patients with thyroid cancer, is strongly associated with aggressive biologic behavior. Predicting TERT promoter mutation is thus necessary for the prognostic stratification of thyroid cancer patients. Materials and Methods: In this study, we evaluate TERT promoter mutation status in thyroid cancer through the deep learning approach using histologic images. Our analysis included 13 consecutive surgically resected thyroid cancers with TERT promoter mutations (either C228T or C250T) and 12 randomly selected surgically resected thyroid cancers with a wild-type TERT promoter. Our deep learning model was created using a two-step cascade approach. First, tumor areas were identified using convolutional neural networks (CNNs), and then TERT promoter mutations within tumor areas were predicted using the CNN–recurrent neural network (CRNN) model. Results: Using the hue–saturation–value (HSV)-strong color transformation scheme, the overall experiment results show 99.9% sensitivity and 60% specificity (improvements of approximately 25% and 37%, respectively, compared to image normalization as a baseline model) in predicting TERT mutations. Conclusions: Highly sensitive screening for TERT promoter mutations is possible using histologic image analysis based on deep learning. This approach will help improve the classification of thyroid cancer patients according to the biologic behavior of tumors.

Impact of Stain Normalization on Pathologist Assessment of Prostate Cancer: A Comparative Study

In clinical routine, the quality of whole-slide images plays a key role in the pathologist's diagnosis, and suboptimal staining may be a limiting factor. The stain normalization process helps to solve this problem through the standardization of color appearance of a source image with respect to a target image with optimal chromatic features. The analysis is focused on the evaluation of the following parameters assessed by two experts on original and normalized slides: (i) perceived color quality, (ii) diagnosis for the patient, (iii) diagnostic confidence and (iv) time required for diagnosis. Results show a statistically significant increase in color quality in the normalized images for both experts (p < 0.0001). Regarding prostate cancer assessment, the average times for diagnosis are significantly lower for normalized images than original ones (first expert: 69.9 s vs. 77.9 s with p < 0.0001; second expert: 37.4 s vs. 52.7 s with p < 0.0001), and at the same time, a statistically significant increase in diagnostic confidence is proven. The improvement of poor-quality images and greater clarity of diagnostically important details in normalized slides demonstrate the potential of stain normalization in the routine practice of prostate cancer assessment.

Stain color translation of multi-domain OSCC histopathology images using attention gated cGAN

Oral Squamous Cell Carcinoma (OSCC) is the most prevalent type of oral cancer across the globe. Histopathology examination is the gold standard for OSCC examination, where stained histopathology slides help in studying and analyzing the cell structures under a microscope to determine the stages and grading of OSCC. One of the staining methods popularly known as H&E staining is used to produce differential coloration, highlight key tissue features, and improve contrast, which makes cell analysis easier. However, the stained H&E histopathology images exhibit inter and intra-variation due to staining techniques, incubation times, and staining reagents. These variations negatively impact computer-aided diagnosis (CAD) and Machine learning algorithm's accuracy and development. A pre-processing procedure called stain normalization must be employed to reduce stain variance's negative impacts. Numerous state-of-the-art stain normalization methods are introduced. However, a robust multi-domain stain normalization approach is still required because, in a real-world situation, the OSCC histopathology images will include more than two color variations involving several domains. In this paper, a multi-domain stain translation method is proposed. The proposed method is an attention gated generator based on a Conditional Generative Adversarial Network (cGAN) with a novel objective function to enforce color distribution and the perpetual resemblance between the source and target domains. Instead of using WSI scanner images like previous techniques, the proposed method is experimented on OSCC histopathology images obtained by several conventional microscopes coupled with cameras. The proposed method receives the L* channel from the L*a*b* color space in inference mode and generates the G(a*b*) channel, which are color-adapted. The proposed technique uses mappings learned during training phases to translate the source domain to the target domain; mapping are learned using the whole color distribution of the target domain instead of one reference image. The suggested technique outperforms the four state-of-the-art methods in multi-domain OSCC histopathological translation, the claim is supported by results obtained after assessment in both quantitative and qualitative ways.

Automated staining analysis in digital cytopathology and applications

The progress of digital pathology in recent years has been an opportunity for the development of automated image analysis algorithms for quantitative measurements and computer aided diagnosis. With those new methods comes the need for high staining quality and reproducibility, as image analysis tools are typically more sensible to slight stain variations than trained pathologists. This article presents a method for the automated analysis of cytology slides stains specifically adapted to the challenges encountered in digital cytopathology. In particular, the variety of cell types in cytology slides, the 3D distribution of the cellular material, the presence of superposed cells and the need for independent analysis of sub-cellular compartments are addressed. The proposed method is applied to the quantification of staining variations for quality control, resulting from changes in the staining protocol such as reagent immersion time or a reagent change. Another demonstrated application is the selection of staining protocol parameters that maximize the visible details in nucleus. Finally the analysis pipeline is also used to compare different stain normalization algorithms on digital cytology slides. Code available at: https://gitlab.com/vitadx/articles/automated_staining_analysis.

Deep learning-based framework for slide-based histopathological image analysis

Digital pathology coupled with advanced machine learning (e.g., deep learning) has been changing the paradigm of whole-slide histopathological images (WSIs) analysis. Major applications in digital pathology using machine learning include automatic cancer classification, survival analysis, and subtyping from pathological images. While most pathological image analyses are based on patch-wise processing due to the extremely large size of histopathology images, there are several applications that predict a single clinical outcome or perform pathological diagnosis per slide (e.g., cancer classification, survival analysis). However, current slide-based analyses are task-dependent, and a general framework of slide-based analysis in WSI has been seldom investigated. We propose a novel slide-based histopathology analysis framework that creates a WSI representation map, called HipoMap, that can be applied to any slide-based problems, coupled with convolutional neural networks. HipoMap converts a WSI of various shapes and sizes to structured image-type representation. Our proposed HipoMap outperformed existing methods in intensive experiments with various settings and datasets. HipoMap showed the Area Under the Curve (AUC) of 0.96±0.026 (5% improved) in the experiments for lung cancer classification, and c-index of 0.787±0.013 (3.5% improved) and coefficient of determination ([Formula: see text]) of 0.978±0.032 (24% improved) in survival analysis and survival prediction with TCGA lung cancer data respectively, as a general framework of slide-based analysis with a flexible capability. The results showed significant improvement comparing to the current state-of-the-art methods on each task. We further discussed experimental results of HipoMap as pathological viewpoints and verified the performance using publicly available TCGA datasets. A Python package is available at https://pypi.org/project/hipomap , and the package can be easily installed using Python PIP. The open-source codes in Python are available at: https://github.com/datax-lab/HipoMap .

Stain normalization in digital pathology: Clinical multi-center evaluation of image quality

In digital pathology, the final appearance of digitized images is affected by several factors, resulting in stain color and intensity variation. Stain normalization is an innovative solution to overcome stain variability. However, the validation of color normalization tools has been assessed only from a quantitative perspective, through the computation of similarity metrics between the original and normalized images. To the best of our knowledge, no works investigate the impact of normalization on the pathologist's evaluation. The objective of this paper is to propose a multi-tissue (i.e., breast, colon, liver, lung, and prostate) and multi-center qualitative analysis of a stain normalization tool with the involvement of pathologists with different years of experience. Two qualitative studies were carried out for this purpose: (i) a first study focused on the analysis of the perceived image quality and absence of significant image artifacts after the normalization process; (ii) a second study focused on the clinical score of the normalized image with respect to the original one. The results of the first study prove the high quality of the normalized image with a low impact artifact generation, while the second study demonstrates the superiority of the normalized image with respect to the original one in clinical practice. The normalization process can help both to reduce variability due to tissue staining procedures and facilitate the pathologist in the histological examination. The experimental results obtained in this work are encouraging and can justify the use of a stain normalization tool in clinical routine.

Interpretable deep learning of myelin histopathology in age-related cognitive impairment

Age-related cognitive impairment is multifactorial, with numerous underlying and frequently co-morbid pathological correlates. Amyloid beta (Aβ) plays a major role in Alzheimer's type age-related cognitive impairment, in addition to other etiopathologies such as Aβ-independent hyperphosphorylated tau, cerebrovascular disease, and myelin damage, which also warrant further investigation. Classical methods, even in the setting of the gold standard of postmortem brain assessment, involve semi-quantitative ordinal staging systems that often correlate poorly with clinical outcomes, due to imperfect cognitive measurements and preconceived notions regarding the neuropathologic features that should be chosen for study. Improved approaches are needed to identify histopathological changes correlated with cognition in an unbiased way. We used a weakly supervised multiple instance learning algorithm on whole slide images of human brain autopsy tissue sections from a group of elderly donors to predict the presence or absence of cognitive impairment (n = 367 with cognitive impairment, n = 349 without). Attention analysis allowed us to pinpoint the underlying subregional architecture and cellular features that the models used for the prediction in both brain regions studied, the medial temporal lobe and frontal cortex. Despite noisy labels of cognition, our trained models were able to predict the presence of cognitive impairment with a modest accuracy that was significantly greater than chance. Attention-based interpretation studies of the features most associated with cognitive impairment in the top performing models suggest that they identified myelin pallor in the white matter. Our results demonstrate a scalable platform with interpretable deep learning to identify unexpected aspects of pathology in cognitive impairment that can be translated to the study of other neurobiological disorders.

Multiclass classification of breast cancer histopathology images using multilevel features of deep convolutional neural network

Breast cancer is a common malignancy and a leading cause of cancer-related deaths in women worldwide. Its early diagnosis can significantly reduce the morbidity and mortality rates in women. To this end, histopathological diagnosis is usually followed as the gold standard approach. However, this process is tedious, labor-intensive, and may be subject to inter-reader variability. Accordingly, an automatic diagnostic system can assist to improve the quality of diagnosis. This paper presents a deep learning approach to automatically classify hematoxylin-eosin-stained breast cancer microscopy images into normal tissue, benign lesion, in situ carcinoma, and invasive carcinoma using our collected dataset. Our proposed model exploited six intermediate layers of the Xception (Extreme Inception) network to retrieve robust and abstract features from input images. First, we optimized the proposed model on the original (unnormalized) dataset using 5-fold cross-validation. Then, we investigated its performance on four normalized datasets resulting from Reinhard, Ruifrok, Macenko, and Vahadane stain normalization. For original images, our proposed framework yielded an accuracy of 98% along with a kappa score of 0.969. Also, it achieved an average AUC-ROC score of 0.998 as well as a mean AUC-PR value of 0.995. Specifically, for in situ carcinoma and invasive carcinoma, it offered sensitivity of 96% and 99%, respectively. For normalized images, the proposed architecture performed better for Makenko normalization compared to the other three techniques. In this case, the proposed model achieved an accuracy of 97.79% together with a kappa score of 0.965. Also, it attained an average AUC-ROC score of 0.997 and a mean AUC-PR value of 0.991. Especially, for in situ carcinoma and invasive carcinoma, it offered sensitivity of 96% and 99%, respectively. These results demonstrate that our proposed model outperformed the baseline AlexNet as well as state-of-the-art VGG16, VGG19, Inception-v3, and Xception models with their default settings. Furthermore, it can be inferred that although stain normalization techniques offered competitive performance, they could not surpass the results of the original dataset.

Development and Evaluation of a Novel Deep-Learning-Based Framework for the Classification of Renal Histopathology Images

Kidney cancer has several types, with renal cell carcinoma (RCC) being the most prevalent and severe type, accounting for more than 85% of adult patients. The manual analysis of whole slide images (WSI) of renal tissues is the primary tool for RCC diagnosis and prognosis. However, the manual identification of RCC is time-consuming and prone to inter-subject variability. In this paper, we aim to distinguish between benign tissue and malignant RCC tumors and identify the tumor subtypes to support medical therapy management. We propose a novel multiscale weakly-supervised deep learning approach for RCC subtyping. Our system starts by applying the RGB-histogram specification stain normalization on the whole slide images to eliminate the effect of the color variations on the system performance. Then, we follow the multiple instance learning approach by dividing the input data into multiple overlapping patches to maintain the tissue connectivity. Finally, we train three multiscale convolutional neural networks (CNNs) and apply decision fusion to their predicted results to obtain the final classification decision. Our dataset comprises four classes of renal tissues: non-RCC renal parenchyma, non-RCC fat tissues, clear cell RCC (ccRCC), and clear cell papillary RCC (ccpRCC). The developed system demonstrates a high classification accuracy and sensitivity on the RCC biopsy samples at the slide level. Following a leave-one-subject-out cross-validation approach, the developed RCC subtype classification system achieves an overall classification accuracy of 93.0% ± 4.9%, a sensitivity of 91.3% ± 10.7%, and a high classification specificity of 95.6% ± 5.2%, in distinguishing ccRCC from ccpRCC or non-RCC tissues. Furthermore, our method outperformed the state-of-the-art Resnet-50 model.

H&E Multi-Laboratory Staining Variance Exploration with Machine Learning

In diagnostic histopathology, hematoxylin and eosin (H&E) staining is a critical process that highlights salient histological features. Staining results vary between laboratories regardless of the histopathological task, although the method does not change. This variance can impair the accuracy of algorithms and histopathologists’ time-to-insight. Investigating this variance can help calibrate stain normalization tasks to reverse this negative potential. With machine learning, this study evaluated the staining variance between different laboratories on three tissue types. We received H&E-stained slides from 66 different laboratories. Each slide contained kidney, skin, and colon tissue samples stained by the method routinely used in each laboratory. The samples were digitized and summarized as red, green, and blue channel histograms. Dimensions were reduced using principal component analysis. The data projected by principal components were inserted into the k-means clustering algorithm and the k-nearest neighbors classifier with the laboratories as the target. The k-means silhouette index indicated that K = 2 clusters had the best separability in all tissue types. The supervised classification result showed laboratory effects and tissue-type bias. Both supervised and unsupervised approaches suggested that tissue type also affected inter-laboratory variance. We suggest tissue type to also be considered upon choosing the staining and color-normalization approach.

Automated knowledge-assisted mitosis cells detection framework in breast histopathology images

Based on the Nottingham Histopathology Grading (NHG) system, mitosis cells detection is one of the important criteria to determine the grade of breast carcinoma. Mitosis cells detection is a challenging task due to the heterogeneous microenvironment of breast histopathology images. Recognition of complex and inconsistent objects in the medical images could be achieved by incorporating domain knowledge in the field of interest. In this study, the strategies of the histopathologist and domain knowledge approach were used to guide the development of the image processing framework for automated mitosis cells detection in breast histopathology images. The detection framework starts with color normalization and hyperchromatic nucleus segmentation. Then, a knowledge-assisted false positive reduction method is proposed to eliminate the false positive (i.e., non-mitosis cells). This stage aims to minimize the percentage of false positive and thus increase the F1-score. Next, features extraction was performed. The mitosis candidates were classified using a Support Vector Machine (SVM) classifier. For evaluation purposes, the knowledge-assisted detection framework was tested using two datasets: a custom dataset and a publicly available dataset (i.e., MITOS dataset). The proposed knowledge-assisted false positive reduction method was found promising by eliminating at least 87.1% of false positive in both the dataset producing promising results in the F1-score. Experimental results demonstrate that the knowledge-assisted detection framework can achieve promising results in F1-score (custom dataset: 89.1%; MITOS dataset: 88.9%) and outperforms the recent works.

Breast histopathological image analysis using image processing techniques for diagnostic puposes: A methodological review

Breast cancer in women is the second most common cancer worldwide. Early detection of breast cancer can reduce the risk of human life. Non-invasive techniques such as mammograms and ultrasound imaging are popularly used to detect the tumour. However, histopathological analysis is necessary to determine the malignancy of the tumour as it analyses the image at the cellular level. Manual analysis of these slides is time consuming, tedious, subjective and are susceptible to human errors. Also, at times the interpretation of these images are inconsistent between laboratories. Hence, a Computer-Aided Diagnostic system that can act as a decision support system is need of the hour. Moreover, recent developments in computational power and memory capacity led to the application of computer tools and medical image processing techniques to process and analyze breast cancer histopathological images. This review paper summarizes various traditional and deep learning based methods developed to analyze breast cancer histopathological images. Initially, the characteristics of breast cancer histopathological images are discussed. A detailed discussion on the various potential regions of interest is presented which is crucial for the development of Computer-Aided Diagnostic systems. We summarize the recent trends and choices made during the selection of medical image processing techniques. Finally, a detailed discussion on the various challenges involved in the analysis of BCHI is presented along with the future scope.

The utility of color normalization for AI-based diagnosis of hematoxylin and eosin-stained pathology images

The color variation of hematoxylin and eosin (H&E)-stained tissues has presented a challenge for applications of artificial intelligence (AI) in digital pathology. Many color normalization algorithms have been developed in recent years in order to reduce the color variation between H&E images. However, previous efforts in benchmarking these algorithms have produced conflicting results and none have sufficiently assessed the efficacy of the various color normalization methods for improving diagnostic performance of AI systems. In this study, we systematically investigated eight color normalization algorithms for AI-based classification of H&E-stained histopathology slides, in the context of using images both from one center and from multiple centers. Our results show that color normalization does not consistently improve classification performance when both training and testing data are from a single center. However, using four multi-center datasets of two cancer types (ovarian and pleural) and objective functions, we show that color normalization can significantly improve the classification accuracy of images from external datasets (ovarian cancer: 0.25 AUC increase, p = 1.6 e-05; pleural cancer: 0.21 AUC increase, p = 1.4 e-10). Furthermore, we introduce a novel augmentation strategy by mixing color-normalized images using three easily accessible algorithms that consistently improves the diagnosis of test images from external centers, even when the individual normalization methods had varied results. We anticipate our study to be a starting point for reliable use of color normalization to improve AI-based, digital pathology-empowered diagnosis of cancers sourced from multiple centers. © 2021 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Deep Learning of Histopathology Images at the Single Cell Level

The tumor immune microenvironment (TIME) encompasses many heterogeneous cell types that engage in extensive crosstalk among the cancer, immune, and stromal components. The spatial organization of these different cell types in TIME could be used as biomarkers for predicting drug responses, prognosis and metastasis. Recently, deep learning approaches have been widely used for digital histopathology images for cancer diagnoses and prognoses. Furthermore, some recent approaches have attempted to integrate spatial and molecular omics data to better characterize the TIME. In this review we focus on machine learning-based digital histopathology image analysis methods for characterizing tumor ecosystem. In this review, we will consider three different scales of histopathological analyses that machine learning can operate within: whole slide image (WSI)-level, region of interest (ROI)-level, and cell-level. We will systematically review the various machine learning methods in these three scales with a focus on cell-level analysis. We will provide a perspective of workflow on generating cell-level training data sets using immunohistochemistry markers to "weakly-label" the cell types. We will describe some common steps in the workflow of preparing the data, as well as some limitations of this approach. Finally, we will discuss future opportunities of integrating molecular omics data with digital histopathology images for characterizing tumor ecosystem.

HistoClean: Open-source software for histological image pre-processing and augmentation to improve development of robust convolutional neural networks

The growth of digital pathology over the past decade has opened new research pathways and insights in cancer prediction and prognosis. In particular, there has been a surge in deep learning and computer vision techniques to analyse digital images. Common practice in this area is to use image pre-processing and augmentation to prevent bias and overfitting, creating a more robust deep learning model. This generally requires consultation of documentation for multiple coding libraries, as well as trial and error to ensure that the techniques used on the images are appropriate. Herein we introduce HistoClean; a user-friendly, graphical user interface that brings together multiple image processing modules into one easy to use toolkit. HistoClean is an application that aims to help bridge the knowledge gap between pathologists, biomedical scientists and computer scientists by providing transparent image augmentation and pre-processing techniques which can be applied without prior coding knowledge. In this study, we utilise HistoClean to pre-process images for a simple convolutional neural network used to detect stromal maturity, improving the accuracy of the model at a tile, region of interest, and patient level. This study demonstrates how HistoClean can be used to improve a standard deep learning workflow via classical image augmentation and pre-processing techniques, even with a relatively simple convolutional neural network architecture. HistoClean is free and open-source and can be downloaded from the Github repository here: https://github.com/HistoCleanQUB/HistoClean.

Artificial Intelligence for solid tumor diagnosis in digital pathology

Tumor diagnosis relies on the visual examination of histological slides by pathologists through a microscope eyepiece. Digital pathology, the digitalization of histological slides at high magnification with slides scanners, has raised the opportunity to extract quantitative information thanks to image analysis. In the last decade, medical image analysis has made exceptional progress due to the development of artificial intelligence (AI) algorithms. AI has been successfully used in the field of medical imaging and more recently in digital pathology. The feasibility and usefulness of AI assisted pathology tasks have been demonstrated in the very last years and we can expect those developments to be applied on routine histopathology in the future. In this review, we will describe and illustrate this technique and present the most recent applications in the field of tumor histopathology.

Artificial Intelligence-Based Screening for Mycobacteria in Whole-Slide Images of Tissue Samples

OBJECTIVES

This study aimed to develop and validate a deep learning algorithm to screen digitized acid fast-stained (AFS) slides for mycobacteria within tissue sections.

METHODS

A total of 441 whole-slide images (WSIs) of AFS tissue material were used to develop a deep learning algorithm. Regions of interest with possible acid-fast bacilli (AFBs) were displayed in a web-based gallery format alongside corresponding WSIs for pathologist review. Artificial intelligence (AI)-assisted analysis of another 138 AFS slides was compared to manual light microscopy and WSI evaluation without AI support.

RESULTS

Algorithm performance showed an area under the curve of 0.960 at the image patch level. More AI-assisted reviews identified AFBs than manual microscopy or WSI examination (P < .001). Sensitivity, negative predictive value, and accuracy were highest for AI-assisted reviews. AI-assisted reviews also had the highest rate of matching the original sign-out diagnosis, were less time-consuming, and were much easier for pathologists to perform (P < .001).

CONCLUSIONS

This study reports the successful development and clinical validation of an AI-based digital pathology system to screen for AFBs in anatomic pathology material. AI assistance proved to be more sensitive and accurate, took pathologists less time to screen cases, and was easier to use than either manual microscopy or viewing WSIs.

A Review of Computer-Aided Expert Systems for Breast Cancer Diagnosis

A computer-aided diagnosis (CAD) expert system is a powerful tool to efficiently assist a pathologist in achieving an early diagnosis of breast cancer. This process identifies the presence of cancer in breast tissue samples and the distinct type of cancer stages. In a standard CAD system, the main process involves image pre-processing, segmentation, feature extraction, feature selection, classification, and performance evaluation. In this review paper, we reviewed the existing state-of-the-art machine learning approaches applied at each stage involving conventional methods and deep learning methods, the comparisons within methods, and we provide technical details with advantages and disadvantages. The aims are to investigate the impact of CAD systems using histopathology images, investigate deep learning methods that outperform conventional methods, and provide a summary for future researchers to analyse and improve the existing techniques used. Lastly, we will discuss the research gaps of existing machine learning approaches for implementation and propose future direction guidelines for upcoming researchers.

Histopathological Classification of Canine Cutaneous Round Cell Tumors Using Deep Learning: A Multi-Center Study

Canine cutaneous round cell tumors (RCT) represent one of the routine diagnostic challenges for veterinary pathologists. Computer-aided approaches are developed to overcome these restrictions and to increase accuracy and consistency of diagnosis. These systems are also of high benefit reducing errors when a large number of cases are screened daily. In this study we describe ARCTA (Automated Round Cell Tumors Assessment), a fully automated algorithm for cutaneous RCT classification and mast cell tumors grading in canine histopathological images. ARCTA employs a deep learning strategy and was developed on 416 RCT images and 213 mast cell tumors images. In the test set, our algorithm exhibited an excellent classification performance in both RCT classification (accuracy: 91.66%) and mast cell tumors grading (accuracy: 100%). Misdiagnoses were encountered for histiocytomas in the train set and for melanomas in the test set. For mast cell tumors the reduction of a grade was observed in the train set, but not in the test set. To the best of our knowledge, the proposed model is the first fully automated algorithm in histological images specifically developed for veterinary medicine. Being very fast (average computational time 2.63 s), this algorithm paves the way for an automated and effective evaluation of canine tumors.

Retinex model based stain normalization technique for whole slide image analysis

Medical imaging provides the means for diagnosing many of the medical phenomena currently studied in clinical medicine and pathology. The variations of color and intensity in stained histological slides affect the quantitative analysis of the histopathological images. Moreover, stain normalization utilizing color for the classification of pixels into different stain components is challenging. The staining also suffers from variability, which complicates the automatization of tissue area segmentation with different staining and the analysis of whole slide images. We have developed a Retinex model based stain normalization technique in terms of area segmentation from stained tissue images to quantify the individual stain components of the histochemical stains for the ideal removal of variability. The performance was experimentally compared to reference methods and tested on organotypic carcinoma model based on myoma tissue and our method consistently has the smallest standard deviation, skewness value, and coefficient of variation in normalized median intensity measurements. Our method also achieved better quality performance in terms of Quaternion Structure Similarity Index Metric (QSSIM), Structural Similarity Index Metric (SSIM), and Pearson Correlation Coefficient (PCC) by improving robustness against variability and reproducibility. The proposed method could potentially be used in the development of novel research as well as diagnostic tools with the potential improvement of accuracy and consistency in computer aided diagnosis in biobank applications.

An adaptive digital stain separation method for deep learning-based automatic cell profile counts

BACKGROUND

Quantifying cells in a defined region of biological tissue is critical for many clinical and preclinical studies, especially in the fields of pathology, toxicology, cancer and behavior. As part of a program to develop accurate, precise and more efficient automatic approaches for quantifying morphometric changes in biological tissue, we have shown that both deep learning-based and hand-crafted algorithms can estimate the total number of histologically stained cells at their maximal profile of focus in Extended Depth of Field (EDF) images. Deep learning-based approaches show accuracy comparable to manual counts on EDF images but significant enhancement in reproducibility, throughput efficiency and reduced error from human factors. However, a majority of the automated counts are designed for single-immunostained tissue sections.

NEW METHOD

To expand the automatic counting methods to more complex dual-staining protocols, we developed an adaptive method to separate stain color channels on images from tissue sections stained by a primary immunostain with secondary counterstain.

COMPARISON WITH EXISTING METHODS

The proposed method overcomes the limitations of the state-of-the-art stain-separation methods, like the requirement of pure stain color basis as a prerequisite or stain color basis learning on each image.

RESULTS

Experimental results are presented for automatic counts using deep learning-based and hand-crafted algorithms for sections immunostained for neurons (Neu-N) or microglial cells (Iba-1) with cresyl violet counterstain.

CONCLUSION

Our findings show more accurate counts by deep learning methods compared to the handcrafted method. Thus, stain-separated images can function as input for automatic deep learning-based quantification methods designed for single-stained tissue sections.

Hidden Variables in Deep Learning Digital Pathology and Their Potential to Cause Batch Effects: Prediction Model Study

BACKGROUND

An increasing number of studies within digital pathology show the potential of artificial intelligence (AI) to diagnose cancer using histological whole slide images, which requires large and diverse data sets. While diversification may result in more generalizable AI-based systems, it can also introduce hidden variables. If neural networks are able to distinguish/learn hidden variables, these variables can introduce batch effects that compromise the accuracy of classification systems.

OBJECTIVE

The objective of the study was to analyze the learnability of an exemplary selection of hidden variables (patient age, slide preparation date, slide origin, and scanner type) that are commonly found in whole slide image data sets in digital pathology and could create batch effects.

METHODS

We trained four separate convolutional neural networks (CNNs) to learn four variables using a data set of digitized whole slide melanoma images from five different institutes. For robustness, each CNN training and evaluation run was repeated multiple times, and a variable was only considered learnable if the lower bound of the 95% confidence interval of its mean balanced accuracy was above 50.0%.

RESULTS

A mean balanced accuracy above 50.0% was achieved for all four tasks, even when considering the lower bound of the 95% confidence interval. Performance between tasks showed wide variation, ranging from 56.1% (slide preparation date) to 100% (slide origin).

CONCLUSIONS

Because all of the analyzed hidden variables are learnable, they have the potential to create batch effects in dermatopathology data sets, which negatively affect AI-based classification systems. Practitioners should be aware of these and similar pitfalls when developing and evaluating such systems and address these and potentially other batch effect variables in their data sets through sufficient data set stratification.

Effect of Biochar on the Production of L-Histidine From Glucose Through Escherichia coli Metabolism

The organic compounds from biochar play a role of hormone analogs, stimulating the expression of metabolites by controlling related gene and protein. In this experiment, we reported the L-histidine biosysthesis was promoted by biochar treatment in E. coli unlike genetic engineering of the traditional method. The related results indicated the most optimal concentration was found to be 3%, and 7% is the lethal dose. E. coli was inhibited in the high-concentration treatment. On the other hand, docking technology was usually used as drug screening, basing on Lock-and-key model of protein in order to better understand mechanisms. So the organic compounds of biochar from GC-MS analysis that acted as ligands were connected to HisG protein controlling L-histidine biosysthesis in E. coli. The result showed that the three organic molecules interacted with HisG protein by hydrogen bond. So we considered that these three compounds play regulatory roles in L-histidine biosysthesis, and the hisG gene expression fully supports this conclusion.

Style transfer strategy for developing a generalizable deep learning application in digital pathology

BACKGROUND AND OBJECTIVES

Despite recent advances in artificial intelligence for medical images, the development of a robust deep learning model for identifying malignancy on pathology slides has been limited by problems related to substantial inter- and intra-institutional heterogeneity attributable to tissue preparation. The paucity of available data aggravates this limitation for relatively rare cancers. Here, using ovarian cancer pathology images, we explored the effect of image-to-image style transfer approaches on diagnostic performance.

METHODS

We leveraged a relatively large public image set for 142 patients with ovarian cancer from The Cancer Image Archive (TCIA) to fine-tune the renowned deep learning model Inception V3 for identifying malignancy on tissue slides. As an external validation, the performance of the developed classifier was tested using a relatively small institutional pathology image set for 32 patients. To reduce deterioration of the performance associated with the inter-institutional heterogeneity of pathology slides, we translated the style of the small image set of the local institution into the large image set style of the TCIA using cycle-consistent generative adversarial networks.

RESULTS

Without style transfer, the performance of the classifier was as follows: area under the receiver operating characteristic curve (AUROC) = 0.737 and area under the precision recall curve (AUPRC) = 0.710. After style transfer, AUROC and AUPRC improved to 0.916 and 0.898, respectively.

CONCLUSIONS

This study provides a case of the successful application of style transfer technology to generalize a deep learning model into small image sets in the field of digital pathology. Researchers at local institutions can select this collaborative system to make their small image sets acceptable to the deep learning model.

Self-Attentive Adversarial Stain Normalization

Hematoxylin and Eosin (H&E) stained Whole Slide Images (WSIs) are utilized for biopsy visualization-based diagnostic and prognostic assessment of diseases. Variation in the H&E staining process across different lab sites can lead to significant variations in biopsy image appearance. These variations introduce an undesirable bias when the slides are examined by pathologists or used for training deep learning models. Traditionally proposed stain normalization and color augmentation strategies can handle the human level bias. But deep learning models can easily disentangle the linear transformation used in these approaches, resulting in undesirable bias and lack of generalization. To handle these limitations, we propose a Self-Attentive Adversarial Stain Normalization (SAASN) approach for the normalization of multiple stain appearances to a common domain. This unsupervised generative adversarial approach includes self-attention mechanism for synthesizing images with finer detail while preserving the structural consistency of the biopsy features during translation. SAASN demonstrates consistent and superior performance compared to other popular stain normalization techniques on H&E stained duodenal biopsy image data.

Experimental Assessment of Color Deconvolution and Color Normalization for Automated Classification of Histology Images Stained with Hematoxylin and Eosin

Histological evaluation plays a major role in cancer diagnosis and treatment. The appearance of H&E-stained images can vary significantly as a consequence of differences in several factors, such as reagents, staining conditions, preparation procedure and image acquisition system. Such potential sources of noise can all have negative effects on computer-assisted classification. To minimize such artefacts and their potentially negative effects several color pre-processing methods have been proposed in the literature-for instance, color augmentation, color constancy, color deconvolution and color transfer. Still, little work has been done to investigate the efficacy of these methods on a quantitative basis. In this paper, we evaluated the effects of color constancy, deconvolution and transfer on automated classification of H&E-stained images representing different types of cancers-specifically breast, prostate, colorectal cancer and malignant lymphoma. Our results indicate that in most cases color pre-processing does not improve the classification accuracy, especially when coupled with color-based image descriptors. Some pre-processing methods, however, can be beneficial when used with some texture-based methods like Gabor filters and Local Binary Patterns.