Histology Image Classification Research Articles

Survey of Recent Deep Neural Networks with Strong Annotated Supervision in Histopathology

Deep learning (DL) and convolutional neural networks (CNNs) have achieved state-of-the-art performance in many medical image analysis tasks. Histopathological images contain valuable information that can be used to diagnose diseases and create treatment plans. Therefore, the application of DL for the classification of histological images is a rapidly expanding field of research. The popularity of CNNs has led to a rapid growth in the number of works related to CNNs in histopathology. This paper aims to provide a clear overview for better navigation. In this paper, recent DL-based classification studies in histopathology using strongly annotated data have been reviewed. All the works have been categorized from two points of view. First, the studies have been categorized into three groups according to the training approach and model construction: 1. fine-tuning of pre-trained networks for one-stage classification, 2. training networks from scratch for one-stage classification, and 3. multi-stage classification. Second, the papers summarized in this study cover a wide range of applications (e.g., breast, lung, colon, brain, kidney). To help navigate through the studies, the classification of reviewed works into tissue classification, tissue grading, and biomarker identification was used.

Knowledge Distillation in Histology Landscape by Multi-Layer Features Supervision.

Automatic tissue classification is a fundamental task in computational pathology for profiling tumor micro-environments. Deep learning has advanced tissue classification performance at the cost of significant computational power. Shallow networks have also been end-to-end trained using direct supervision however their performance degrades because of the lack of capturing robust tissue heterogeneity. Knowledge distillation has recently been employed to improve the performance of the shallow networks used as student networks by using additional supervision from deep neural networks used as teacher networks. In the current work, we propose a novel knowledge distillation algorithm to improve the performance of shallow networks for tissue phenotyping in histology images. For this purpose, we propose multi-layer feature distillation such that a single layer in the student network gets supervision from multiple teacher layers. In the proposed algorithm, the size of the feature map of two layers is matched by using a learnable multi-layer perceptron. The distance between the feature maps of the two layers is then minimized during the training of the student network. The overall objective function is computed by summation of the loss over multiple layers combination weighted with a learnable attention-based parameter. The proposed algorithm is named as Knowledge Distillation for Tissue Phenotyping (KDTP). Experiments are performed on five different publicly available histology image classification datasets using several teacher-student network combinations within the KDTP algorithm. Our results demonstrate a significant performance increase in the student networks by using the proposed KDTP algorithm compared to direct supervision-based training methods.

One model is all you need: Multi-task learning enables simultaneous histology image segmentation and classification.

The recent surge in performance for image analysis of digitised pathology slides can largely be attributed to the advances in deep learning. Deep models can be used to initially localise various structures in the tissue and hence facilitate the extraction of interpretable features for biomarker discovery. However, these models are typically trained for a single task and therefore scale poorly as we wish to adapt the model for an increasing number of different tasks. Also, supervised deep learning models are very data hungry and therefore rely on large amounts of training data to perform well. In this paper, we present a multi-task learning approach for segmentation and classification of nuclei, glands, lumina and different tissue regions that leverages data from multiple independent data sources. While ensuring that our tasks are aligned by the same tissue type and resolution, we enable meaningful simultaneous prediction with a single network. As a result of feature sharing, we also show that the learned representation can be used to improve the performance of additional tasks via transfer learning, including nuclear classification and signet ring cell detection. As part of this work, we train our developed Cerberus model on a huge amount of data, consisting of over 600 thousand objects for segmentation and 440 thousand patches for classification. We use our approach to process 599 colorectal whole-slide images from TCGA, where we localise 377 million, 900 thousand and 2.1 million nuclei, glands and lumina respectively. We make this resource available to remove a major barrier in the development of explainable models for computational pathology.

Convolutional Neural Networks for Classifying Cervical Cancer Types Using Histological Images.

Cervical cancer is the most common cancer among women worldwide. The diagnosis and classification of cancer are extremely important, as it influences the optimal treatment and length of survival. The objective was to develop and validate a diagnosis system based on convolutional neural networks (CNN) that identifies cervical malignancies and provides diagnostic interpretability. A total of 8496 labeled histology images were extracted from 229 cervical specimens (cervical squamous cell carcinoma, SCC, n = 37; cervical adenocarcinoma, AC, n = 8; nonmalignant cervical tissues, n = 184). AlexNet, VGG-19, Xception, and ResNet-50 with five-fold cross-validation were constructed to distinguish cervical cancer images from nonmalignant images. The performance of CNNs was quantified in terms of accuracy, precision, recall, and the area under the receiver operating curve (AUC). Six pathologists were recruited to make a comparison with the performance of CNNs. Guided Backpropagation and Gradient-weighted Class Activation Mapping (Grad-CAM) were deployed to highlight the area of high malignant probability. The Xception model had excellent performance in identifying cervical SCC and AC in test sets. For cervical SCC, AUC was 0.98 (internal validation) and 0.974 (external validation). For cervical AC, AUC was 0.966 (internal validation) and 0.958 (external validation). The performance of CNNs falls between experienced and inexperienced pathologists. Grad-CAM and Guided Gard-CAM ensured diagnoses interpretability by highlighting morphological features of malignant changes. CNN is efficient for histological image classification tasks of distinguishing cervical malignancies from benign tissues and could highlight the specific areas of concern. All these findings suggest that CNNs could serve as a diagnostic tool to aid pathologic diagnosis.

A Scalable Graph-Based Framework for Multi-Organ Histology Image Classification

Graph-based approaches are successful for histology image classification tasks but still face many challenges, such as: 1) the lack of nuclei-level labels and the significant variations between histology images make it extremely difficult to extract discriminative high-level nuclei features like nuclei type, texture and micro-environment; 2) graph-based approaches cannot handle large-scale cell graph nodes typically contained in histology images; and 3) graph neural networks (GNNs) struggle to learn the long-range dependency of cell graphs. To address the above challenges, we propose a scalable graph-based framework for multi-organ histology image classification. We develop a two-step masked nuclei patches supervised training approach to extract discriminative high-level nuclei features for histology images without nuclei-level labels. Additionally, we introduce a nuclei sampling strategy to make our graph-based framework scalable for large-scale cell graphs. Furthermore, we propose <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><b>H</b></u> ier <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><b>A</b></u> rchical <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><b>T</b></u> ransformer Graph Neural <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><b>Net</b></u> work (HAT-Net+) for cell graph classi- fications. HAT-Net+ adopts Transformer to model the long-range dependency of cell graphs and a parameter-free approach to adaptively fuse different hierarchical graph representations of each layer. We achieved the state-of-the-art results on four public histology image classification datasets: CRC dataset (100%), Extended CRC dataset (98%), UZH dataset (96.9%) and BACH dataset (88%). Unlike other methods, our approach can be used in various histology image classification tasks, even for images without nuclei-level labels, indicating its potential in cancer diagnosis. The code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/suyouooooo/HAT-Net</uri> .

Automatic classification of histology images into normal/cancer class with pre-trained CNN

Deep-Learning-Scheme (DLS) based medical data assessment has been widely employed in recent years due to its improved accuracy. Our goal is to study the performance of the pre-trained DLS on RGB-scale breast-histology images. The implemented idea holds these phases; (i) Data collection, pre-processing and resizing, (ii) Training the DLS with chosen test-pictures, (iii) Testing and validating the performance of the DLS with 5-fold cross-validation. This investigation considered the breast-histology pictures for the study and binary classification is employed to achieve Normal/Cancer class grouping of images. The proposed work compared the classification performance of AlexNet, VGG16 and VGG19.The experimental outcome of this study authenticates that the AlexNet with the Random-Forest (RF) classifier helps to get a higher classification accuracy (>87%) compared to VGG16 and VGG19.

Automatic Detection System to Identify Invasive Ductal Carcinoma by Predicting Bloom Richardson Grading from Histopathological Images

After skin cancer, the most common type of cancer is breast cancer among the world population. Breast cancer is the leading cause of cancer-induced mortality among women. Breast cancer is frequently diagnosed by using biopsies in which tissue is removed from the breast and studied under a microscope. The results of these biopsies are based on the qualifications and experience of the pathologist who diagnoses the abnormal cell under the microscope. With the emergence of advancements in the fields of image processing and artificial intelligence, there is an area of interest in developing a deep learning model to improve and enhance the quality and accuracy of breast cancer diagnosis. This study proposed a deep learning model that automatically analyses the multiclass classification of hematoxylin and eosin-stained histological images of invasive ductal carcinoma (IDC) by discriminating the IDC into grades such as G-1, G-2, and G-3. The methodology is focused on a deep learning model to detect grades of invasive ductal carcinoma by adopting the Sequential Convolutional Neural Network Two-Dimensional (CNN2D). We used DataBiox, a public dataset taken from an internet source consisting of 922 images. We evaluate the result of multiclass classification by dividing 80% and 20% of the dataset into training and testing data, respectively. As a result of the training and testing of the pre-trained CNN model, sequential CNN yields the accuracy of the model of 92.81%. Our model accurately classifies a multi-class classification of histological images of grades of breast cancer, specifically IDC. It is ready to be tested with a more diverse and massive database in the future.

Noise-reducing attention cross fusion learning transformer for histological image classification of osteosarcoma

The degree of malignancy of osteosarcoma and its tendency to metastasize/spread mainly depend on the pathological grade (determined by observing the morphology of the tumor under a microscope). The purpose of this study is to use artificial intelligence to classify osteosarcoma histological images and to assess tumor survival and necrosis, which will help doctors reduce their workload, improve the accuracy of osteosarcoma cancer detection, and make a better prognosis for patients. The study proposes a typical transformer image classification framework by integrating noise reduction convolutional autoencoder and feature cross fusion learning (NRCA-FCFL) to classify osteosarcoma histological images. Noise reduction convolutional autoencoder could well denoise histological images of osteosarcoma, resulting in more pure images for osteosarcoma classification. Moreover, we introduce feature cross fusion learning, which integrates two scale image patches, to sufficiently explore their interactions by using additional classification tokens. As a result, a refined fusion feature is generated, which is fed to the residual neural network for label predictions. We conduct extensive experiments to evaluate the performance of the proposed approach. The experimental results demonstrate that our method outperforms the traditional and deep learning approaches on various evaluation metrics, with an accuracy of 99.17% to support osteosarcoma diagnosis.

TSFD-Net: Tissue specific feature distillation network for nuclei segmentation and classification

Nuclei segmentation and classification of hematoxylin and eosin-stained histology images is a challenging task due to a variety of issues, such as color inconsistency that results from the non-uniform manual staining operations, clustering of nuclei, and blurry and overlapping nuclei boundaries. Existing approaches involve segmenting nuclei by drawing their polygon representations or by measuring the distances between nuclei centroids. In contrast, we leverage the fact that morphological features (appearance, shape, and texture) of nuclei in a tissue vary greatly depending upon the tissue type. We exploit this information by extracting tissue specific (TS) features from raw histopathology images using the proposed tissue specific feature distillation (TSFD) backbone. The bi-directional feature pyramid network (BiFPN) within TSFD-Net generates a robust hierarchical feature pyramid utilizing TS features where the interlinked decoders jointly optimize and fuse these features to generate final predictions. We also propose a novel combinational loss function for joint optimization and faster convergence of our proposed network. Extensive ablation studies are performed to validate the effectiveness of each component of TSFD-Net. The proposed network outperforms state-of-the-art networks such as StarDist, Micro-Net, Mask-RCNN, Hover-Net, and CPP-Net on the PanNuke dataset, which contains 19 different tissue types and 5 clinically important tumor classes, achieving 50.4% and 63.77% mean and binary panoptic quality, respectively. The code is available at: https://github.com/Mr-TalhaIlyas/TSFD.

Deep Interpretable Classification and Weakly-Supervised Segmentation of Histology Images via Max-Min Uncertainty.

Weakly-supervised learning (WSL) has recently triggered substantial interest as it mitigates the lack of pixel-wise annotations. Given global image labels, WSL methods yield pixel-level predictions (segmentations), which enable to interpret class predictions. Despite their recent success, mostly with natural images, such methods can face important challenges when the foreground and background regions have similar visual cues, yielding high false-positive rates in segmentations, as is the case in challenging histology images. WSL training is commonly driven by standard classification losses, which implicitly maximize model confidence, and locate the discriminative regions linked to classification decisions. Therefore, they lack mechanisms for modeling explicitly non-discriminative regions and reducing false-positive rates. We propose novel regularization terms, which enable the model to seek both non-discriminative and discriminative regions, while discouraging unbalanced segmentations. We introduce high uncertainty as a criterion to localize non-discriminative regions that do not affect classifier decision, and describe it with original Kullback-Leibler (KL) divergence losses evaluating the deviation of posterior predictions from the uniform distribution. Our KL terms encourage high uncertainty of the model when the latter inputs the latent non-discriminative regions. Our loss integrates: (i) a cross-entropy seeking a foreground, where model confidence about class prediction is high; (ii) a KL regularizer seeking a background, where model uncertainty is high; and (iii) log-barrier terms discouraging unbalanced segmentations. Comprehensive experiments and ablation studies over the public GlaS colon cancer data and a Camelyon16 patch-based benchmark for breast cancer show substantial improvements over state-of-the-art WSL methods, and confirm the effect of our new regularizers (our code is publicly available at https://github.com/sbelharbi/deep-wsl-histo-min-max-uncertainty).

Improving efficiency in convolutional neural networks with 3D image filters

Background and objectiveThe effective performance of deep networks has provided the solution to various state-of-the-art problems. Convolutional Neural Network (CNN) is accepted as an accurate, effective, and reliable practice in image-based applications. However, there is a need to use pre-trained models in case of insufficient data in CNN. This study aims to present an alternative solution to this problem with the proposed 3D image-based filter generation approach with simpler CNNs for the classification of small datasets. MethodsIn this study, a novel 3D image filters-based CNN (Hist3DCNN) is proposed. The proposed filter generation approach is based on 3D object images taken from different perspectives. The efficiency of Hist3DCNN is shown on a novel histological dataset that contains blood, connective, epithelium, muscle, and nerve tissue images. Various case studies are carried out with generated filters assigned as the initial value to AlexNet and the designed Hist3DCNN model that is simpler than AlexNet. ResultsBased on results, the classification accuracy of AlexNet with proposed filters used in convolution layers were 84.65% and 85.34%. The accuracy was increased to 85.47% by Hist3DCNN on the histological image classification. Moreover, four different benchmark datasets were tested to demonstrate the robustness of Hist3DCNN on various datasets. ConclusionsThis study provides a new aspect to literature due to 3D image-based filter generation approach to initialize convolution filters. Experimental results validate that Hist3DCNN can be used as a filter value initialization method with simple CNN models that contain less learnable parameters for the classification task of small datasets.

Deep Learning-Enabled Technologies for Bioimage Analysis.

Deep learning (DL) is a subfield of machine learning (ML), which has recently demonstrated its potency to significantly improve the quantification and classification workflows in biomedical and clinical applications. Among the end applications profoundly benefitting from DL, cellular morphology quantification is one of the pioneers. Here, we first briefly explain fundamental concepts in DL and then we review some of the emerging DL-enabled applications in cell morphology quantification in the fields of embryology, point-of-care ovulation testing, as a predictive tool for fetal heart pregnancy, cancer diagnostics via classification of cancer histology images, autosomal polycystic kidney disease, and chronic kidney diseases.

MCUa: Multi-Level Context and Uncertainty Aware Dynamic Deep Ensemble for Breast Cancer Histology Image Classification.

Breast histology image classification is a crucial step in the early diagnosis of breast cancer. In breast pathological diagnosis, Convolutional Neural Networks (CNNs) have demonstrated great success using digitized histology slides. However, tissue classification is still challenging due to the high visual variability of the large-sized digitized samples and the lack of contextual information. In this paper, we propose a novel CNN, called Multi-level Context and Uncertainty aware (MCUa) dynamic deep learning ensemble model. MCUa model consists of several multi-level context-aware models to learn the spatial dependency between image patches in a layer-wise fashion. It exploits the high sensitivity to the multi-level contextual information using an uncertainty quantification component to accomplish a novel dynamic ensemble model. MCUa model has achieved a high accuracy of 98.11% on a breast cancer histology image dataset. Experimental results show the superior effectiveness of the proposed solution compared to the state-of-the-art histology classification models.

Methodical approaches to annotation and labeling of histological images in order to automatically detect the layers of the stomach wall and the depth of invasion of gastric cancer

Development of original methodological approaches to annotation and labeling of histological images in relation to the problem of automatic segmentation of the layers of the stomach wall. Three image collections were used in the study: NCT-CRC-HE-100K, CRC-VAL-HE-7K, and part of the PATH-DT-MSU collection. The used part of the original PATH-DT-MSU collection contains 20 histological images obtained using a high performance digital scanning microscope. A labeling system is proposed that includes 5 classes (tissue types): areas of gastric adenocarcinoma (TUM), unchanged areas of the lamina propria (LP), unchanged areas of the muscular lamina of the mucosa (MM), a class of underlying tissues (AT), including areas of the submucosa, own muscular layer of the stomach and subserous sections, image background (BG). The advantage of this marking technique is to ensure high efficiency of recognition of the muscularis lamina (MM) - a natural «line» separating the lamina propria of the mucous membrane and all other underlying layers of the stomach wall. The disadvantages of the technique include a small number of classes, which leads to insufficient detailing of automatic segmentation. In the course of the study, an original technique for labeling and annotating images was developed, including 5 classes (types of tissues). This technique is effective at the initial stages of teaching mathematical algorithms for the classification and segmentation of histological images. Further stages in the development of a real diagnostic algorithm to automatically determine the depth of invasion of gastric cancer require the correction and development of the presented method of marking and annotation.

Detection of Breast Cancer with Lightweight Deep Neural Networks for Histology Image Classification.

Many researchers have developed computer-assisted diagnostic (CAD) methods to diagnose breast cancer using histopathology microscopic images. These techniques help to improve the accuracy of biopsy diagnosis with hematoxylin and eosin-stained images. On the other hand, most CAD systems usually rely on inefficient and time-consuming manual feature extraction methods. Using a deep learning (DL) model with convolutional layers, we present a method to extract the most useful pictorial information for breast cancer classification. Breast biopsy images stained with hematoxylin and eosin can be categorized into four groups namely benign lesions, normal tissue, carcinoma in situ, and invasive carcinoma. To correctly classify different types of breast cancer, it is important to classify histopathological images accurately. The MobileNet architecture model is used to obtain high accuracy with less resource utilization. The proposed model is fast, inexpensive, and safe due to which it is suitable for the detection of breast cancer at an early stage. This lightweight deep neural network can be accelerated using field-programmable gate arrays for the detection of breast cancer. DL has been implemented to successfully classify breast cancer. The model uses categorical cross-entropy to learn to give the correct class a high probability and other classes a low probability. It is used in the classification stage of the convolutional neural network (CNN) after the clustering stage, thereby improving the performance of the proposed system. To measure training and validation accuracy, the model was trained on Google Colab for 280 epochs with a powerful GPU with 2496 CUDA cores, 12 GB GDDR5 VRAM, and 12.6 GB RAM. Our results demonstrate that deep CNN with a chi-square test has improved the accuracy of histopathological image classification of breast cancer by greater than 11% compared with other state-of-the-art methods.

DBLCNN: Dependency-based lightweight convolutional neural network for multi-classification of breast histopathology images

Breast histopathology analysis is the gold standard for diagnosing breast cancer. Convolutional neural network-based methods for breast histology image classification have emerged in recent years to make the analysis process simple and fast. Due to the limitation of hardware devices, these classification methods still face the problem of difficult balance recognition performance and computational efficiency. In this paper, we propose the Dependency-based lightweight convolutional neural network (DBLCNN) for the multi-classification task of breast histopathology images. Firstly, we design a new network in which dependencies (magnification and binary classification probability) were used to guide subspecies features for better recognition. Secondly, we redesign the backbone MobileNet to greatly reduce the model parameters and computation while ensuring excellent recognition performance. At the same time, transfer learning based on ImageNet is applied to the DBLCNN network. Extensive experiments on the BreakHis dataset have shown that the DBLCNN network has state-of-the-art effects in terms of recognition performance and computational utilization.

Fuzzy ensemble of deep learning models using choquet fuzzy integral, coalition game and information theory for breast cancer histology classification

Millions of women, worldwide, suffer from breast cancer and a large number of them succumb to death. In recent years, computer-aided diagnosis (CAD) systems are being developed for the detection of Breast Cancer. A number of fusion techniques have been proposed in this domain, but none of them take into consideration the decisions taken by a subset of classifiers during fusion. Our method, which uses Choquet Integral, considers subsets of classifiers and is thus stronger than the existing methods and beat all of these existing fusion methods in terms of accuracy. This however poses a significant challenge in terms of complexity, since the calculation of the fuzzy measures is a complicated and complex task, which we have dealt with using a novel heuristic method by employing Coalition Game, Information Theory, and by defining a novel mathematical function. In the present work, we have fused VGG16, VGG19, Xception, Inception V3, and InceptionResnet V2 for the classification of breast cancer histology images using a Choquet integral, Coalition game theory, and Information theory. The dataset used for evaluating the proposed model is the ICIAR 2018 Grand Challenge on Breast Cancer Histology (popularly known as BACH) images, which consist of 2-class and 4-class problems. To the best of our knowledge, our experimental results outperform almost all the state-of-the-art methods. For the two-class problem, the best test accuracy among the five deep learning models was achieved by Xception and it was 95% while the Fusion method has a test accuracy of 96%. For the four-class problem, Xception and InceptionResnet V2 have achieved the best test accuracy and both have a test accuracy of 91% while the Fusion method has a test accuracy of 95%. Again, in the case of the two-class problem the best precision and recall by the deep learning models are 0.95 and 0.95 respectively, while the precision and recall for after fusion are 0.96 and 0.96 respectively which is an increase of .01. In the case of the four-class problem, the best precision and recall by the deep learning models are 0.91 and 0.91 respectively, while the precision and recall after fusion are 0.95 and 0.95 respectively which is a very significant increase of .04. The source code for this project can be accessed at https://github.com/subhankar01/fuzzyBACH

Blind color deconvolution, normalization, and classification of histological images using general super Gaussian priors and Bayesian inference

Background and Objective:Color variations in digital histopathology severely impact the performance of computer-aided diagnosis systems. They are due to differences in the staining process and acquisition system, among other reasons. Blind color deconvolution techniques separate multi-stained images into single stained bands which, once normalized, can be used to eliminate these negative color variations and improve the performance of machine learning tasks. Methods:In this work, we decompose the observed RGB image in its hematoxylin and eosin components. We apply Bayesian modeling and inference based on the use of Super Gaussian sparse priors for each stain together with prior closeness to a given reference color-vector matrix. The hematoxylin and eosin components are then used for image normalization and classification of histological images. The proposed framework is tested on stain separation, image normalization, and cancer classification problems. The results are measured using the peak signal to noise ratio, normalized median intensity and the area under ROC curve on five different databases. Results:The obtained results show the superiority of our approach to current state-of-the-art blind color deconvolution techniques. In particular, the fidelity to the tissue improves 1,27 dB in mean PSNR. The normalized median intensity shows a good normalization quality of the proposed approach on the tested datasets. Finally, in cancer classification experiments the area under the ROC curve improves from 0.9491 to 0.9656 and from 0.9279 to 0.9541 on Camelyon-16 and Camelyon-17, respectively, when the original and processed images are used. Furthermore, these figures of merits are better than those obtained by the methods compared with. Conclusions:The proposed framework for blind color deconvolution, normalization and classification of images guarantees fidelity to the tissue structure and can be used both for normalization and classification. In addition, color deconvolution enables the use of the optical density space for classification, which improves the classification performance.

Artificial intelligence in computational pathology – challenges and future directions

The field of digital histopathology has seen incredible growth in recent years. Digital pathology is becoming a relevant tool in healthcare, industrial and research sectors to reduce the saturation of pathology departments and improve the productivity of pathologists by increasing diagnostic accuracy and reducing turnaround times. Artificial Intelligence (AI) algorithms may be used for the identification of relevant regions, extraction of features from a histological image and overall classification of images into specific classes. The combination of digital histopathology imaging and AI therefore presents a significant opportunity for the support of the pathologists' tasks and opens up a whole new world of computational analysis. In this paper, we have analysed the present, the challenges and the future of the computational pathology discussing the different existing strategies to overcome its main limitations and ensure the computational pathology acceptance. The lack of labelled data, which is the possibly largest challenge for all medical AI applications, is even more pronounced in computational pathology because of the multi-gigapixel nature of the images and high data heterogeneity. We consider the future of the computational pathology is the combination of weak label strategies with active learning and crowdsourcing scenarios since it would remove some of the workload from clinical experts and manual annotation obtaining clinically satisfactory performance with minimal annotation effort. In addition, we believe areas such as explainable AI, data fusion and secure role-based data sharing will be receiving increasing research attention in computational pathology in the close future.

Resolution-based distillation for efficient histology image classification

Developing deep learning models to analyze histology images has been computationally challenging, as the massive size of the images causes excessive strain on all parts of the computing pipeline. This paper proposes a novel deep learning-based methodology for improving the computational efficiency of histology image classification. The proposed approach is robust when used with images that have reduced input resolution, and it can be trained effectively with limited labeled data. Moreover, our approach operates at either the tissue- or slide-level, removing the need for laborious patch-level labeling. Our method uses knowledge distillation to transfer knowledge from a teacher model pre-trained at high resolution to a student model trained on the same images at a considerably lower resolution. Also, to address the lack of large-scale labeled histology image datasets, we perform the knowledge distillation in a self-supervised fashion. We evaluate our approach on three distinct histology image datasets associated with celiac disease, lung adenocarcinoma, and renal cell carcinoma. Our results on these datasets demonstrate that a combination of knowledge distillation and self-supervision allows the student model to approach and, in some cases, surpass the teacher model's classification accuracy while being much more computationally efficient. Additionally, we observe an increase in student classification performance as the size of the unlabeled dataset increases, indicating that there is potential for this method to scale further with additional unlabeled data. Our model outperforms the high-resolution teacher model for celiac disease in accuracy, F1-score, precision, and recall while requiring 4 times fewer computations. For lung adenocarcinoma, our results at 1.25× magnification are within 1.5% of the results for the teacher model at 10× magnification, with a reduction in computational cost by a factor of 64. Our model on renal cell carcinoma at 1.25× magnification performs within 1% of the teacher model at 5× magnification while requiring 16 times fewer computations. Furthermore, our celiac disease outcomes benefit from additional performance scaling with the use of more unlabeled data. In the case of 0.625× magnification, using unlabeled data improves accuracy by 4% over the tissue-level baseline. Therefore, our approach can improve the feasibility of deep learning solutions for digital pathology on standard computational hardware and infrastructures.