Histopathological Dataset Research Articles

Data consistency and classification model transferability across biomedical Raman spectroscopy systems

AbstractSurgical guidance applications using Raman spectroscopy are being developed at a rapid pace in oncology to ensure safe and complete tumor resection during surgery. Clinical translation of these approaches relies on the acquisition of large spectral and histopathological data sets to train classification models. Data calibration must ensure compatibility across Raman systems and predictive model transferability to allow multi‐centric studies to be conducted. This paper addresses issues relating to Raman measurement standardization by first comparing Raman spectral measurements made on an optical phantom and acquired with nine distinct point probe systems and one wide‐field imaging instrument. Data standardization method led to normalized root‐mean‐square deviations between instruments of 2%. A classification model discriminating between white and gray matter was trained with one point probe system. When used to classify independent data sets acquired with the other systems, model predictions led to >95% accuracy, preliminarily demonstrating model transferability across different biomedical Raman spectroscopy instruments.

Gastric histopathology image segmentation using a hierarchical conditional random field

For the Convolutional Neural Networks (CNNs) applied in the intelligent diagnosis of gastric cancer, existing methods mostly focus on individual characteristics or network frameworks without a policy to depict the integral information. Mainly, conditional random field (CRF), an efficient and stable algorithm for analyzing images containing complicated contents, can characterize spatial relation in images. In this paper, a novel hierarchical conditional random field (HCRF) based gastric histopathology image segmentation (GHIS) method is proposed, which can automatically localize abnormal (cancer) regions in gastric histopathology images obtained by an optical microscope to assist histopathologists in medical work. This HCRF model is built up with higher order potentials, including pixel-level and patch-level potentials, and graph-based post-processing is applied to further improve its segmentation performance. Especially, a CNN is trained to build up the pixel-level potentials and another three CNNs are fine-tuned to build up the patch-level potentials for sufficient spatial segmentation information. In the experiment, a hematoxylin and eosin (H&E) stained gastric histopathological dataset with 560 abnormal images are divided into training, validation and test sets with a ratio of 1:1:2. Finally, segmentation accuracy, recall and specificity of 78.91%, 65.59%, and 81.33% are achieved on the test set. Our HCRF model demonstrates high segmentation performance and shows its effectiveness and future potential in the GHIS field.

Efficient Regularization Framework for Histopathological Image Classification Using Convolutional Neural Networks.

Deep learning methods are characterized by their capacity to learn data representation compared to the traditional machine learning algorithms. However, these methods are prone to overfitting on small volumes of data. The objective of this research is to overcome this limitation by improving the generalization in the proposed deep learning framework based on various techniques: data augmentation, small models, optimizer selection, and ensemble learning. For ensembling, the authors used selected models from different checkpoints and both voting and unweighted average methods for combination. The experimental study on the lymphomas histopathological dataset highlights the efficiency of the MobileNet2 network combined with the stochastic gradient descent (SGD) optimizer in terms of generalization. The best results have been achieved by the combination of the best three checkpoint models (98.67% of accuracy). These findings provide important insights into the efficiency of the checkpoint ensemble learning method for histopathological image classification.

A New Deep Learning Model Selection Method for Colorectal Cancer Classification

Deep learning is one of the most commonly used techniques in computer-aided diagnosis systems. Their exploitation for histopathological image analysis is important because of the complex morphology of whole slide images. However, the main limitation of these methods is the restricted number of available medical images, which can lead to an overfitting problem. Many studies have suggested the use of static ensemble learning methods to address this issue. This article aims to propose a new dynamic ensemble deep learning method. First, it generates a set of models based on the transfer learning strategy from deep neural networks. Then, the relevant subset of models is selected by the particle swarm optimization algorithm and combined by voting or averaging methods. The proposed approach was tested on a histopathological dataset for colorectal cancer classification, based on seven types of CNNs. The method has achieved accurate results (94.52%) by the Resnet121 model and the voting strategy, which provides important insights into the efficiency of dynamic ensembling in deep learning.

Hierarchical conditional random field model for multi‐object segmentation in gastric histopathology images

In this Letter, a hierarchical conditional random field (HCRF) model-based gastric histopathology image segmentation (GHIS) method is proposed, which can localise abnormal (cancer) regions in gastric histopathology images to assist histopathologists in medical work. First, to obtain pixel-level segmentation information, the authors retrain a convolutional neural network (CNN) to build up their pixel-level potentials. Then, to obtain abundant spatial segmentation information in patch level, they fine tune another three CNNs to build up their patch-level potentials. Thirdly, based on the pixel- and patch-level potentials, their HCRF model is structured. Finally, a graph-based post-processing is applied to further improve their segmentation performance. In the experiment, a segmentation accuracy of 78.91 % is achieved on a haematoxylin and eosin stained gastric histopathological dataset with 560 images, showing the effectiveness and future potential of the proposed GHIS method.

Implementing Deep Learning Algorithms in Anatomic Pathology Using Open-source Deep Learning Libraries.

The application of artificial intelligence technologies to anatomic pathology has the potential to transform the practice of pathology, but, despite this, many pathologists are unfamiliar with how these models are created, trained, and evaluated. In addition, many pathologists may feel that they do not possess the necessary skills to allow them to embark on research into this field. This article aims to act as an introductory tutorial to illustrate how to create, train, and evaluate simple artificial learning models (neural networks) on histopathology data sets in the programming language Python using the popular freely available, open-source libraries Keras, TensorFlow, PyTorch, and Detecto. Furthermore, it aims to introduce pathologists to commonly used terms and concepts used in artificial intelligence.

How Deeply to Fine-Tune a Convolutional Neural Network: A Case Study Using a Histopathology Dataset

Accurate classification of medical images is of great importance for correct disease diagnosis. The automation of medical image classification is of great necessity because it can provide a second opinion or even a better classification in case of a shortage of experienced medical staff. Convolutional neural networks (CNN) were introduced to improve the image classification domain by eliminating the need to manually select which features to use to classify images. Training CNN from scratch requires very large annotated datasets that are scarce in the medical field. Transfer learning of CNN weights from another large non-medical dataset can help overcome the problem of medical image scarcity. Transfer learning consists of fine-tuning CNN layers to suit the new dataset. The main questions when using transfer learning are how deeply to fine-tune the network and what difference in generalization that will make. In this paper, all of the experiments were done on two histopathology datasets using three state-of-the-art architectures to systematically study the effect of block-wise fine-tuning of CNN. Results show that fine-tuning the entire network is not always the best option; especially for shallow networks, alternatively fine-tuning the top blocks can save both time and computational power and produce more robust classifiers.

The effect of batch size on the generalizability of the convolutional neural networks on a histopathology dataset

Abstract Many hyperparameters have to be tuned to have a robust convolutional neural network that will be able to accurately classify images. One of the most important hyperparameters is the batch size, which is the number of images used to train a single forward and backward pass. In this study, the effect of batch size on the performance of convolutional neural networks and the impact of learning rates will be studied for image classification, specifically for medical images. To train the network faster, a VGG16 network with ImageNet weights was used in this experiment. Our results concluded that a higher batch size does not usually achieve high accuracy, and the learning rate and the optimizer used will have a significant impact as well. Lowering the learning rate and decreasing the batch size will allow the network to train better, especially in the case of fine-tuning.

A Novel Attribute-Based Symmetric Multiple Instance Learning for Histopathological Image Analysis.

Histopathological image analysis is a challenging task due to a diverse histology feature set as well as due to the presence of large non-informative regions in whole slide images. In this paper, we propose a multiple-instance learning (MIL) method for image-level classification as well as for annotating relevant regions in the image. In MIL, a common assumption is that negative bags contain only negative instances while positive bags contain one or more positive instances. This asymmetric assumption may be inappropriate for some application scenarios where negative bags also contain representative negative instances. We introduce a novel symmetric MIL framework associating each instance in a bag with an attribute which can be either negative, positive, or irrelevant. We extend the notion of relevance by introducing control over the number of relevant instances. We develop a probabilistic graphical model that incorporates the aforementioned paradigm and a corresponding computationally efficient inference for learning the model parameters and obtaining an instance level attribute-learning classifier. The effectiveness of the proposed method is evaluated on available histopathology datasets with promising results.

Automated invasive ductal carcinoma detection based using deep transfer learning with whole-slide images

Advances in artificial intelligence technologies have made it possible to obtain more accurate and reliable results using digital images. Due to the advances in digital histopathological images obtained using whole slide image (WSI) scanners, automated analysis of digital images by computer support systems has become interesting. In particular, deep learning architectures, are one of the preferred approaches in the analysis of digital histopathology images. The deeper networks trained on large amounts of image data are adapted for different tasks using transfer learning technique. In this study, automated detection of invasive ductal carcinoma (IDC), which is the most common subtype of breast cancers, is proposed using deep transfer learning technique. We have used deep learning pre-trained models, ResNet-50 and DenseNet-161 for the IDC detection task. The public histopathology dataset containing 277,524 image patches were used in our experimental studies. As a result of training on the last layers of pre-trained deep networks, DenseNet-161 model has yielded F-sore of 92.38% and balanced accuracy value of 91.57%. Similarly, we have obtained F-score of 94.11% and balanced accuracy value of 90.96% using ResNet-50 architecture. In addition, our developed model is validated using the publicly available BreakHis breast cancer dataset and obtained promising results in classifying magnification independent histopathology images into benign and malignant classes. Our developed system obtained the highest classification performance as compared to the state-of-art techniques and is ready to be tested with more diverse huge databases.

Deep Transfer Learning for Histopathological Diagnosis of Cervical Cancer Using Convolutional Neural Networks with Visualization Schemes

This study aimed to propose a deep transfer learning framework for histopathological image analysis by using convolutional neural networks (CNNs) with visualization schemes, and to evaluate its usage for automated and interpretable diagnosis of cervical cancer. First, in order to examine the potential of the transfer learning for classifying cervix histopathological images, we pre-trained three state-of-the-art CNN architectures on large-size natural image datasets and then fine-tuned them on small-size histopathological datasets. Second, we investigated the impact of three learning strategies on classification accuracy. Third, we visualized both the multiple-layer convolutional kernels of CNNs and the regions of interest so as to increase the clinical interpretability of the networks. Our method was evaluated on a database of 4993 cervical histological images (2503 benign and 2490 malignant). The experimental results demonstrated that our method achieved 95.88% sensitivity, 98.93% specificity, 97.42% accuracy, 94.81% Youden's index and 99.71% area under the receiver operating characteristic curve. Our method can reduce the cognitive burden on pathologists for cervical disease classification and improve their diagnostic efficiency and accuracy. It may be potentially used in clinical routine for histopathological diagnosis of cervical cancer.

HWDCNN: Multi-class recognition in breast histopathology with Haar wavelet decomposed image based convolution neural network

Among the predominant cancers, breast cancer is one of the main causes of cancer deaths impacting women worldwide. However, breast cancer classification is challenging due to numerous morphological and textural variations that appeared in intra-class images. Also, the direct processing of high resolution histological images is uneconomical in terms of GPU memory. In the present study, we have proposed a new approach for breast histopathological image classification that uses a deep convolution neural network (CNN) with wavelet decomposed images. The original microscopic image patches of 2048 × 1536 × 3 pixels are decomposed into 512 × 384 × 3 using 2-level Haar wavelet and subsequently used in proposed CNN model. The image decomposition step considerably reduces convolution time in deep CNNs and computational resources, without any performance downgrade. The CNN model extracts the deep features from Haar wavelet decomposed images and incorporates multi-scale discriminant features for precise prognostication of class labels. This paper also solves the demand for massive histopathology dataset by means of transfer learning and data augmentation techniques. The efficacy of proposed approach is corroborated on two publicly available breast histology datasets-(a) one provided as a part of international conference on image analysis and recognition (ICIAR 2018) grand challenge and (b) on BreakHis data. On the ICIAR 2018 validation data, our model showed an accuracy of 98.2% for both 4-class and 2-class recognition. Further, on hidden test data of the ICIAR 2018, we achieved an accuracy of 91%, outperforming existing state-of-the-art results significantly. Besides, on BreakHis dataset, the model achieved competing performance with 96.85% multi-class accuracy.

Inflammatory leucocyte infiltrates are associated with recovery in biopsy-proven acute interstitial nephritis: a 20-year registry-based case series.

BackgroundAcute interstitial nephritis (AIN) is a renal injury causing renal function deterioration and requiring renal replacement therapy (RRT) in a substantial number of cases. Therapy is based on withdrawal of suspicious causative drugs or the underlying diseases and/or steroid application if renal function is not restored after cessation of the underlying condition. Hard clinical evidence for augmenting steroid therapy is not available.MethodsWe reviewed the course and diagnosis for >20 years among all 1126 biopsied samples of our tertiary renal centre.Results49 (4.4%) were diagnosed with primary AIN, corresponding to an annual incidence of 1/100 000 population; 17 out of 49 biopsy-proven AIN patients required short-term or long-term (n = 5) RRT. According to a combined outcome criterion of coming off dialysis and/or reaching serum creatinine <200 µmol/L, 19 patients reached recovery whereas 20 did not. Among 39 patients with a comprehensive clinical and histopathological data set, presence of cortical scars, AIN histological activity (acute leucocyte infiltrates) and proteinuria were baseline parameters discriminating significantly between groups with or without recovery. No associations with the presence of specific drugs were found. Therapeutic use of steroids was associated with a lower probability of recovery (P = 0.008), presumably due to inclusion bias.ConclusionsFollowing our basic finding of the importance of histopathological parameters of acuity associated with recovery, we argue for the inauguration of grading measures to characterize this issue quantitatively and make it usable for future controlled investigations. Finally, we provide a suggestion for a therapeutic algorithm in the management of AIN.

Content-based histopathological image retrieval using multi-scale and multichannel decoder based LTP

The paper proposes a content based histopathological image retrieval scheme based on multi-scale, multichannel decoded local ternary pattern features and VLAD coding. The scheme analyzes an image into multiple scales wherein a decoder based local ternary pattern is utilized to extract upper and lower texture images at each level. A dense patch based histogram representation is obtained at each scale which is followed by VLAD coding and power-law normalization. Simulations performed on a standard 20-class histopathological dataset. The proposed scheme shows considerable improvement over state-of-the-art techniques.

Curation and analysis of clinical pathology parameters and histopathologic findings from eTOXsys, a large database project (eTOX) for toxicologic studies

Large data sharing projects amongst the pharmaceutical industry have the potential to generate new insights using data on a scale that has not been previously available. A retrospective analysis of the preclinical toxicology data collected as part of the eTOX project was conducted with the aim to provide background rates and treatment-related value analysis on both clinical pathology and histopathology datasets. Incorporated into this analysis was an extensive data consolidation task to standardise all data. Reference intervals for common clinical pathology parameters in rat and dog were generated, alongside background histopathology incidence rates in the liver, heart and kidney. Systematically applied decision thresholds allowed consistent relabelling of data points considered anomalous, and maximum fold change estimates. Relabelling of anomalous data points was conducted for the histopathology data using a Bayesian model to identify dose-dependent increases in pathologies. The results of this study allow: newly generated data to be analysed using the same methodology, rates and distributions to be used when building predictive dose-response models, and the possibility to correlate clinical pathology findings with concurrent histopathology findings. In the first half of this paper we discuss data curation, in the second half we report on the analytical methods and results.

Contour-Seed Pairs Learning-Based Framework for Simultaneously Detecting and Segmenting Various Overlapping Cells/Nuclei in Microscopy Images.

In this paper, we propose a novel contour-seed pairs learning-based framework for robust and automated cell/nucleus segmentation. Automated granular object segmentation in microscopy images has significant clinical importance for pathology grading of the cell carcinoma and gene expression. The focus of the past literature is dominated by either segmenting a certain type of cells/nuclei or simply splitting the clustered objects without contours inference of them. Our method addresses these issues by formulating the detection and segmentation tasks in terms of a unified regression problem, where a cascade sparse regression chain model is trained and then applied to return object locations and entire boundaries of clustered objects. In particular, we first learn a set of online convolutional features in each layer. Then, in the proposed cascade sparse regression chain, with the input from the learned features, we iteratively update the locations and clustered object boundaries until convergence. In this way, the boundary evidences of each individual object can be easily delineated and be further fed to a complete contour inference procedure optimized by the minimum description length principle. For any probe image, our method enables to analyze free-lying and overlapping cells with complex shapes. Experimental results show that the proposed method is very generic and performs well on contour inferences of various cell/nucleus types. Compared with the current segmentation techniques, our approach achieves state-of-the-art performances on four challenging datasets, i.e., the kidney renal cell carcinoma histopathology dataset, Drosophila Kc167 cellular dataset, differential interference contrast red blood cell dataset, and cervical cytology dataset.

Improved Random Forest for Classification.

We propose an improved random forest classifier that performs classification with minimum number of trees. The proposed method iteratively removes some unimportant features. Based on the number of important and unimportant features, we formulate a novel theoretical upper limit on the number of trees to be added to the forest to ensure improvement in classification accuracy. Our algorithm converges with a reduced but important set of features. We prove that further addition of trees or further reduction of features does not improve classification performance. The efficacy of the proposed approach is demonstrated through experiments on benchmark datasets. We further use the proposed classifier to detect mitotic nuclei in the histopathological datasets of breast tissues. We also apply our method on the industrial dataset of dual phase steel microstructures to classify different phases. Results of our method on different datasets show significant reduction in average classification error compared to a number of competing methods.

Synergies between texture features: an abstract feature based framework for meningioma subtypes classification

Histopathology is the gold standard for accurate diagnosis of cancer, tumors and similar diseases. Real-world pathological images, due to non-homogeneous nature and unorganized spatial intensity variations, are complex to analyze and classify. The major challenge in classifying pathological images is the complexity due to high intra-class variability and low inter-class variation in texture. Accuracy of histopathological image classification is highly dependent on the relevancy of the selected features to the problem. This paper is an effort in the same direction and presents an abstract feature based framework called abstract feature framework (AFF) to select optimal set of the most relevant features to classify pathological images. An abstract feature is created by identifying interlinked run-length texture features and grouping them. AFF is comprised of a new data structure called Abstract Feature Tree (AFT) and an algorithm for manipulating it. AFT is a tree structure in which nodes are abstract features. The Linkage Learning Algorithm for manipulating AFT is the brain of this framework and inspired by genetic algorithm. It creates better abstract features by first identifying interlinked abstract features and then combining them. This process is repeated until no improvement is found. On termination, the final list of abstract features is used for classifying pathological images. The proposed framework was tested on real-world histopathological meningioma dataset. Results obtained proved that the proposed framework outperformed the best-known rank-based feature selection techniques by using, on average, approximately three times less features to achieve 22% higher classification accuracy.

An Abstract to See the Adequacy of Essential Criteria Filled While Submitting Specimen Request Forms in Histopathology

Adequacy of clinical information on the specimen request forms in histopathology datasets provide guidelines for items to be included on the histopathology request forms. These include full name, date of birth (DOB), NHS number or hospital number, specimen type, and date of procedure. According to the histopathology handbook at our trust, referring consultant and location are …

Estimation of automatic detection of erythemato-squamous diseases through AdaBoost and its hybrid classifiers

This paper focuses on efficient techniques based on AdaBoost and its hybrid (AdaBoost-SVM) classifiers for an automatic detection of the erythemato-squamous diseases (ESD). The classification of ESD requires an enormous amount of computational effort, as all the features of the diseases have more than 90 % commonality. Some of the approaches reported in literature are fuzzy logic, artificial neural networks, neuro-fuzzy models and support vector machine. Recently, AdaBoost and its hybrid algorithms are widely used to enhance the accuracy of the learning algorithms. In this paper, we propose AdaBoost and its hybrid algorithms for diagnosis of erythemato-squamous diseases for the first time. User-friendly-interface is designed to assist dermatologists to estimate the ESD with the help of 34 features includes patient's histopathological and clinical data sets. Efficient interface design is based on the following steps: (1) feature selection for all the 34 parameters involved in the erythemato-squamous disease is done using association rule based Apriori algorithm, (2) association rules are used to find $$\upalpha \%$$?% of transactions (diseases) to meet $$\upbeta \%$$β% of the features of the diseases, (3) support value of all the subset of $$\upalpha $$? transaction from $$\upalpha \%$$?% of transactions is found and the dominant subset is found using Apriori algorithm, and (4) AdaBoost and its hybrid classifiers are used to classify the diseases. In this paper, the real, modest, gentle and hybrid AdaBoost algorithms are implemented and tested with the dataset of 366 patients. The data preprocessing reduces the dimensionality of the dataset and improves the time complexity of the estimation of the diseases. Proposed methodologies are proved as a potential solution for diagnosing ESD with significant improvement in computational time and accuracy compared to other models discussed in the recent literature.