Stacked Sparse Autoencoder (SSAE) for Nuclei Detection on Breast Cancer Histopathology Images

Nucleus segmentation across imaging experiments: the 2018 Data Science Bowl

Segmenting the nuclei of cells in microscopy images is often the first step in the quantitative analysis of imaging data for biological and biomedical applications. Many bioimage analysis tools can segment nuclei in images but need to be selected and configured for every experiment. The 2018 Data Science Bowl attracted 3,891 teams worldwide to make the first attempt to build a segmentation method that could be applied to any two-dimensional light microscopy image of stained nuclei across experiments, with no human interaction. Top participants in the challenge succeeded in this task, developing deep-learning-based models that identified cell nuclei across many image types and experimental conditions without the need to manually adjust segmentation parameters. This represents an important step toward configuration-free bioimage analysis software tools.

Ophthalmic diagnosis using deep learning with fundus images - A critical review

An overview of the applications of deep learning for ophthalmic diagnosis using retinal fundus images is presented. We describe various retinal image datasets that can be used for deep learning purposes. Applications of deep learning for segmentation of optic disk, optic cup, blood vessels as well as detection of lesions are reviewed. Recent deep learning models for classification of diseases such as age-related macular degeneration, glaucoma, and diabetic retinopathy are also discussed. Important critical insights and future research directions are given.

Improving Myoelectric Pattern Recognition Robustness to Electrode Shift by Autoencoder

It is evident that the electrode shift will result in a degradation of myoelectric pattern recognition classification accuracy, which is inevitable during the prosthetic socket donning and doffing. To cope with this limitation, we propose an unsupervised feature extraction method called sparse autoencoder (SAE) to extract the robust spatial structure and correlation of high density (HD) electromyography (EMG). The algorithm is evaluated on nine intact-limbed subjects and one amputee. The experimental results show that SAE achieves lower classification error without shift, and significantly decrease the sensitivity to electrode shift with ±1 cm compared with the timedomain and autoregressive features (TDAR). Furthermore, SAE is not sensitive to the shift direction that is perpendicular to the muscle fibers. The promising results of this study make great contribution to promoting the applications of pattern recognition based myoelectric control system in real-world condition.

Transfer Learning for Molecular Cancer Classification Using Deep Neural Networks

The emergence of deep learning has impacted numerous machine learning based applications and research. The reason for its success lies in two main advantages: 1) it provides the ability to learn very complex non-linear relationships between features and 2) it allows one to leverage information from unlabeled data that does not belong to the problem being handled. This paper presents a transfer learning procedure for cancer classification, which uses feature selection and normalization techniques in conjunction with s sparse auto-encoders on gene expression data. While classifying any two tumor types, data of other tumor types were used in unsupervised manner to improve the feature representation. The performance of our algorithm was tested on 36 two-class benchmark datasets from the GEMLeR repository. On performing statistical tests, it is clearly ascertained that our algorithm statistically outperforms several generally used cancer classification approaches. The deep learning based molecular disease classification can be used to guide decisions made on the diagnosis and treatment of diseases, and therefore may have important applications in precision medicine.

Assessing the Impact of Color Normalization in Convolutional Neural Network-Based Nuclei Segmentation Frameworks

Image analysis tools for cancer, such as automatic nuclei segmentation, are impacted by the inherent variation contained in pathology image data. Convolutional neural networks (CNN), demonstrate success in generalizing to variable data, illustrating great potential as a solution to the problem of data variability. In some CNN-based segmentation works for digital pathology, authors apply color normalization (CN) to reduce color variability of data as a preprocessing step prior to prediction, while others do not. Both approaches achieve reasonable performance and yet, the reasoning for utilizing this step has not been justified. It is therefore important to evaluate the necessity and impact of CN for deep learning frameworks, and its effect on downstream processes. In this paper, we evaluate the effect of popular CN methods on CNN-based nuclei segmentation frameworks.

A hierarchical integration deep flexible neural forest framework for cancer subtype classification by integrating multi-omics data

Background

Cancer subtype classification attains the great importance for accurate diagnosis and personalized treatment of cancer. Latest developments in high-throughput sequencing technologies have rapidly produced multi-omics data of the same cancer sample. Many computational methods have been proposed to classify cancer subtypes, however most of them generate the model by only employing gene expression data. It has been shown that integration of multi-omics data contributes to cancer subtype classification.

Results

A new hierarchical integration deep flexible neural forest framework is proposed to integrate multi-omics data for cancer subtype classification named as HI-DFNForest. Stacked autoencoder (SAE) is used to learn high-level representations in each omics data, then the complex representations are learned by integrating all learned representations into a layer of autoencoder. Final learned data representations (from the stacked autoencoder) are used to classify patients into different cancer subtypes using deep flexible neural forest (DFNForest) model.Cancer subtype classification is verified on BRCA, GBM and OV data sets from TCGA by integrating gene expression, miRNA expression and DNA methylation data. These results demonstrated that integrating multiple omics data improves the accuracy of cancer subtype classification than only using gene expression data and the proposed framework has achieved better performance compared with other conventional methods.

Conclusion

The new hierarchical integration deep flexible neural forest framework(HI-DFNForest) is an effective method to integrate multi-omics data to classify cancer subtypes.

A Novel Digital Score for Abundance of Tumour Infiltrating Lymphocytes Predicts Disease Free Survival in Oral Squamous Cell Carcinoma

Oral squamous cell carcinoma (OSCC) is the most common type of head and neck (H&N) cancers with an increasing worldwide incidence and a worsening prognosis. The abundance of tumour infiltrating lymphocytes (TILs) has been shown to be a key prognostic indicator in a range of cancers with emerging evidence of its role in OSCC progression and treatment response. However, the current methods of TIL analysis are subjective and open to variability in interpretation. An automated method for quantification of TIL abundance has the potential to facilitate better stratification and prognostication of oral cancer patients. We propose a novel method for objective quantification of TIL abundance in OSCC histology images. The proposed TIL abundance (TILAb) score is calculated by first segmenting the whole slide images (WSIs) into underlying tissue types (tumour, lymphocytes, etc.) and then quantifying the co-localization of lymphocytes and tumour areas in a novel fashion. We investigate the prognostic significance of TILAb score on digitized WSIs of Hematoxylin and Eosin (H&E) stained slides of OSCC patients. Our deep learning based tissue segmentation achieves high accuracy of 96.31%, which paves the way for reliable downstream analysis. We show that the TILAb score is a strong prognostic indicator (p = 0.0006) of disease free survival (DFS) on our OSCC test cohort. The automated TILAb score has a significantly higher prognostic value than the manual TIL score (p = 0.0024). In summary, the proposed TILAb score is a digital biomarker which is based on more accurate classification of tumour and lymphocytic regions, is motivated by the biological definition of TILs as tumour infiltrating lymphocytes, with the added advantages of objective and reproducible quantification.

Towards pixel-to-pixel deep nucleus detection in microscopy images

BACKGROUND

Nucleus is a fundamental task in microscopy image analysis and supports many other quantitative studies such as object counting, segmentation, tracking, etc. Deep neural networks are emerging as a powerful tool for biomedical image computing; in particular, convolutional neural networks have been widely applied to nucleus/cell detection in microscopy images. However, almost all models are tailored for specific datasets and their applicability to other microscopy image data remains unknown. Some existing studies casually learn and evaluate deep neural networks on multiple microscopy datasets, but there are still several critical, open questions to be addressed.

RESULTS

We analyze the applicability of deep models specifically for nucleus detection across a wide variety of microscopy image data. More specifically, we present a fully convolutional network-based regression model and extensively evaluate it on large-scale digital pathology and microscopy image datasets, which consist of 23 organs (or cancer diseases) and come from multiple institutions. We demonstrate that for a specific target dataset, training with images from the same types of organs might be usually necessary for nucleus detection. Although the images can be visually similar due to the same staining technique and imaging protocol, deep models learned with images from different organs might not deliver desirable results and would require model fine-tuning to be on a par with those trained with target data. We also observe that training with a mixture of target and other/non-target data does not always mean a higher accuracy of nucleus detection, and it might require proper data manipulation during model training to achieve good performance.

CONCLUSIONS

We conduct a systematic case study on deep models for nucleus detection in a wide variety of microscopy images, aiming to address several important but previously understudied questions. We present and extensively evaluate an end-to-end, pixel-to-pixel fully convolutional regression network and report a few significant findings, some of which might have not been reported in previous studies. The model performance analysis and observations would be helpful to nucleus detection in microscopy images.

Deep Multi-View Feature Learning for EEG-Based Epileptic Seizure Detection

Epilepsy is a neurological illness caused by abnormal discharge of brain neurons, where epileptic seizure can lead to life-threatening emergencies. By analyzing the encephalogram (EEG) signals of patients with epilepsy, their conditions can be monitored and seizure can be detected and intervened in time. As the identification of effective features in EEG signals is important for accurate seizure detection, this paper proposes a multi-view deep feature extraction method in attempt to achieve this goal. The method first uses fast Fourier transform (FFT) and wavelet packet decomposition (WPD) to construct the initial multi-view features. Convolutional neural network (CNN) is then used to automatically learn deep features from the initial multi-view features, which reduces the dimensionality and obtain the features with better seizure identification ability. Furthermore, the multi-view Takagi-Sugeno-Kang fuzzy system (MV-TSK-FS), an interpretable rule-based classifier, is used to construct a classification model with strong generalizability based on the deep multi-view features obtained. Experimental studies show that the classification accuracy of the proposed multi-view deep feature extraction method is at least 1% higher than that of common feature extraction methods such as principal component analysis (PCA), FFT and WPD. The classification accuracy is also at least 4% higher than the average accuracy achieved with single-view deep features.

Prior Information Guided Regularized Deep Learning for Cell Nucleus Detection

Cell nuclei detection is a challenging research topic because of limitations in cellular image quality and diversity of nuclear morphology, i.e., varying nuclei shapes, sizes, and overlaps between multiple cell nuclei. This has been a topic of enduring interest with promising recent success shown by deep learning methods. These methods train convolutional neural networks (CNNs) with a training set of input images and known, labeled nuclei locations. Many such methods are supplemented by spatial or morphological processing. Using a set of canonical cell nuclei shapes, prepared with the help of a domain expert, we develop a new approach that we call shape priors (SPs) with CNNs (SPs-CNN). We further extend the network to introduce an SP layer and then allowing it to become trainable (i.e., optimizable). We call this network as tunable SP-CNN (TSP-CNN). In summary, we present new network structures that can incorporate "expected behavior" of nucleus shapes via two components: learnable layers that perform the nucleus detection and a fixed processing part that guides the learning with prior information. Analytically, we formulate two new regularization terms that are targeted at: 1) learning the shapes and 2) reducing false positives while simultaneously encouraging detection inside the cell nucleus boundary. Experimental results on two challenging datasets reveal that the proposed SP-CNN and TSP-CNN can outperform the state-of-the-art alternatives.

RMDL: Recalibrated multi-instance deep learning for whole slide gastric image classification

The whole slide histopathology images (WSIs) play a critical role in gastric cancer diagnosis. However, due to the large scale of WSIs and various sizes of the abnormal area, how to select informative regions and analyze them are quite challenging during the automatic diagnosis process. The multi-instance learning based on the most discriminative instances can be of great benefit for whole slide gastric image diagnosis. In this paper, we design a recalibrated multi-instance deep learning method (RMDL) to address this challenging problem. We first select the discriminative instances, and then utilize these instances to diagnose diseases based on the proposed RMDL approach. The designed RMDL network is capable of capturing instance-wise dependencies and recalibrating instance features according to the importance coefficient learned from the fused features. Furthermore, we build a large whole-slide gastric histopathology image dataset with detailed pixel-level annotations. Experimental results on the constructed gastric dataset demonstrate the significant improvement on the accuracy of our proposed framework compared with other state-of-the-art multi-instance learning methods. Moreover, our method is general and can be extended to other diagnosis tasks of different cancer types based on WSIs.

Learning to detect lymphocytes in immunohistochemistry with deep learning

The immune system is of critical importance in the development of cancer. The evasion of destruction by the immune system is one of the emerging hallmarks of cancer. We have built a dataset of 171,166 manually annotated CD3⁺ and CD8⁺ cells, which we used to train deep learning algorithms for automatic detection of lymphocytes in histopathology images to better quantify immune response. Moreover, we investigate the effectiveness of four deep learning based methods when different subcompartments of the whole-slide image are considered: normal tissue areas, areas with immune cell clusters, and areas containing artifacts. We have compared the proposed methods in breast, colon and prostate cancer tissue slides collected from nine different medical centers. Finally, we report the results of an observer study on lymphocyte quantification, which involved four pathologists from different medical centers, and compare their performance with the automatic detection. The results give insights on the applicability of the proposed methods for clinical use. U-Net obtained the highest performance with an F1-score of 0.78 and the highest agreement with manual evaluation (κ=0.72), whereas the average pathologists agreement with reference standard was κ=0.64. The test set and the automatic evaluation procedure are publicly available at lyon19.grand-challenge.org.

Autoencoder-Based Detection of Dynamic Allostery Triggered by Ligand Binding Based on Molecular Dynamics

Dynamic allostery on proteins, triggered by regulator binding or chemical modifications, transmits information from the binding site to distant regions, dramatically altering protein function. It is accompanied by subtle changes in side-chain conformations of the protein, indicating that the changes in dynamics, and not rigid or large conformational changes, are essential to understand regulation of protein function. Although a lot of experimental and theoretical studies have been dedicated to investigate this issue, the regulation mechanism of protein function is still being debated. Here, we propose an autoencoder-based method that can detect dynamic allostery. The method is based on the comparison of time fluctuations of protein structures, in the form of distance matrices, obtained from molecular dynamics simulations in ligand-bound and -unbound forms. Our method detected that the changes in dynamics by ligand binding in the PDZ2 domain led to the reorganization of correlative fluctuation motions among residue pairs, which revealed a different view of the correlated motions from the PCA and DCCM. In addition, other correlative motions were also found as a result of the dynamic perturbation from the ligand binding, which may lead to dynamic allostery. This autoencoder-based method would be usefully applied to the signal transduction and mutagenesis systems involved in protein functions and severe diseases.

Fast ScanNet: Fast and Dense Analysis of Multi-Gigapixel Whole-Slide Images for Cancer Metastasis Detection

Lymph node metastasis is one of the most important indicators in breast cancer diagnosis, that is traditionally observed under the microscope by pathologists. In recent years, with the dramatic advance of high-throughput scanning and deep learning technology, automatic analysis of histology from whole-slide images has received a wealth of interest in the field of medical image computing, which aims to alleviate pathologists' workload and simultaneously reduce misdiagnosis rate. However, the automatic detection of lymph node metastases from whole-slide images remains a key challenge because such images are typically very large, where they can often be multiple gigabytes in size. Also, the presence of hard mimics may result in a large number of false positives. In this paper, we propose a novel method with anchor layers for model conversion, which not only leverages the efficiency of fully convolutional architectures to meet the speed requirement in clinical practice but also densely scans the whole-slide image to achieve accurate predictions on both micro- and macro-metastases. Incorporating the strategies of asynchronous sample prefetching and hard negative mining, the network can be effectively trained. The efficacy of our method is corroborated on the benchmark dataset of 2016 Camelyon Grand Challenge. Our method achieved significant improvements in comparison with the state-of-the-art methods on tumor localization accuracy with a much faster speed and even surpassed human performance on both challenge tasks.

Osteoporosis Recognition in Rats under Low-Power Lens Based on Convexity Optimization Feature Fusion

Considering the poor medical conditions in some regions of China, this paper attempts to develop a simple and easy way to extract and process the bone features of blurry medical images and improve the diagnosis accuracy of osteoporosis as much as possible. After reviewing the previous studies on osteoporosis, especially those focusing on texture analysis, a convexity optimization model was proposed based on intra-class dispersion, which combines texture features and shape features. Experimental results show that the proposed model boasts a larger application scope than Lasso, a popular feature selection method that only supports generalized linear models. The research findings ensure the accuracy of osteoporosis diagnosis and enjoy good potentials for clinical application.

A deep learning algorithm for one-step contour aware nuclei segmentation of histopathology images

This paper addresses the task of nuclei segmentation in high-resolution histopathology images. We propose an automatic end-to-end deep neural network algorithm for segmentation of individual nuclei. A nucleus-boundary model is introduced to predict nuclei and their boundaries simultaneously using a fully convolutional neural network. Given a color-normalized image, the model directly outputs an estimated nuclei map and a boundary map. A simple, fast, and parameter-free post-processing procedure is performed on the estimated nuclei map to produce the final segmented nuclei. An overlapped patch extraction and assembling method is also designed for seamless prediction of nuclei in large whole-slide images. We also show the effectiveness of data augmentation methods for nuclei segmentation task. Our experiments showed our method outperforms prior state-of-the-art methods. Moreover, it is efficient that one 1000×1000 image can be segmented in less than 5 s. This makes it possible to precisely segment the whole-slide image in acceptable time. The source code is available at https://github.com/easycui/nuclei_segmentation . Graphical Abstract The neural network for nuclei segmentation.

Automated identification of malignancy in whole-slide pathological images: identification of eyelid malignant melanoma in gigapixel pathological slides using deep learning

BACKGROUND/AIMS

To develop a deep learning system (DLS) that can automatically detect malignant melanoma (MM) in the eyelid from histopathological sections with colossal information density.

METHODS

Setting: Double institutional study.

STUDY POPULATION

We retrospectively reviewed 225 230 pathological patches (small section cut from pathologist-labelled area from an H&E image), cut from 155 H&E-stained whole-slide images (WSI).

OBSERVATION PROCEDURES

Labelled gigapixel pathological WSIs were used to train and test a model designed to assign patch-level classification. Using malignant probability from a convolutional neural network, the patches were embedded back into each WSI to generate a visualisation heatmap and leveraged a random forest model to establish a WSI-level diagnosis.

MAIN OUTCOME MEASURE(S)

For classification, the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity and specificity were used to evaluate the efficacy of the DLS in detecting MM.

RESULTS

For patch diagnosis, the model achieved an AUC of 0.989 (95% CI 0.989 to 0.991), with an accuracy, sensitivity and specificity of 94.9%, 94.7% and 95.3%, respectively. We displayed the lesion area on the WSIs as graded by malignant potential. For WSI, the obtained sensitivity, specificity and accuracy were 100%, 96.5% and 98.2%, respectively, with an AUC of 0.998 (95% CI 0.994 to 1.000).

CONCLUSION

Our DLS, which uses artificial intelligence, can automatically detect MM in histopathological slides and highlight the lesion area on WSIs using a probabilistic heatmap. In addition, our approach has the potential to be applied to the histopathological sections of other tumour types.

Variety Identification of Orchids Using Fourier Transform Infrared Spectroscopy Combined with Stacked Sparse Auto-Encoder

The feasibility of using the fourier transform infrared (FTIR) spectroscopic technique with a stacked sparse auto-encoder (SSAE) to identify orchid varieties was studied. Spectral data of 13 orchids varieties covering the spectral range of 4000-550 cm^-1 were acquired to establish discriminant models and to select optimal spectral variables. K nearest neighbors (KNN), support vector machine (SVM), and SSAE models were built using full spectra. The SSAE model performed better than the KNN and SVM models and obtained a classification accuracy 99.4% in the calibration set and 97.9% in the prediction set. Then, three algorithms, principal component analysis loading (PCA-loading), competitive adaptive reweighted sampling (CARS), and stacked sparse auto-encoder guided backward (SSAE-GB), were used to select 39, 300, and 38 optimal wavenumbers, respectively. The KNN and SVM models were built based on optimal wavenumbers. Most of the optimal wavenumbers-based models performed slightly better than the all wavenumbers-based models. The performance of the SSAE-GB was better than the other two from the perspective of the accuracy of the discriminant models and the number of optimal wavenumbers. The results of this study showed that the FTIR spectroscopic technique combined with the SSAE algorithm could be adopted in the identification of the orchid varieties.

Breast Cancer Diagnosis Using Feature Ensemble Learning Based on Stacked Sparse Autoencoders and Softmax Regression

Nowadays, the most frequent cancer in women is breast cancer (malignant tumor). If breast cancer is detected at the beginning stage, it can often be cured. Many researchers proposed numerous methods for early prediction of this Cancer. In this paper, we proposed feature ensemble learning based on Sparse Autoencoders and Softmax Regression for classification of Breast Cancer into benign (non-cancerous) and malignant (cancerous). We used Breast Cancer Wisconsin (Diagnostic) medical data sets from the UCI machine learning repository. The proposed method is assessed using various performance indices like true classification accuracy, specificity, sensitivity, recall, precision, f measure, and MCC. Simulation and result proved that the proposed approach gives better results in terms of different parameters. The prediction results obtained by the proposed approach were very promising (98.60% true accuracy). In addition, the proposed method outperforms the Stacked Sparse Autoencoders and Softmax Regression based (SSAE-SM) model and other State-of-the-art classifiers in terms of various performance indices. Experimental simulations, empirical results, and statistical analyses are also showing that the proposed model is an efficient and beneficial model for classification of Breast Cancer. It is also comparable with the existing machine learning and soft computing approaches present in the related literature.

Studying the Manifold Structure of Alzheimer's Disease: A Deep Learning Approach Using Convolutional Autoencoders

Many classical machine learning techniques have been used to explore Alzheimer's disease (AD), evolving from image decomposition techniques such as principal component analysis toward higher complexity, non-linear decomposition algorithms. With the arrival of the deep learning paradigm, it has become possible to extract high-level abstract features directly from MRI images that internally describe the distribution of data in low-dimensional manifolds. In this work, we try a new exploratory data analysis of AD based on deep convolutional autoencoders. We aim at finding links between cognitive symptoms and the underlying neurodegeneration process by fusing the information of neuropsychological test outcomes, diagnoses, and other clinical data with the imaging features extracted solely via a data-driven decomposition of MRI. The distribution of the extracted features in different combinations is then analyzed and visualized using regression and classification analysis, and the influence of each coordinate of the autoencoder manifold over the brain is estimated. The imaging-derived markers could then predict clinical variables with correlations above 0.6 in the case of neuropsychological evaluation variables such as the MMSE or the ADAS11 scores, achieving a classification accuracy over 80% for the diagnosis of AD.

Robust Histopathology Image Analysis: to Label or to Synthesize?

Detection, segmentation and classification of nuclei are fundamental analysis operations in digital pathology. Existing state-of-the-art approaches demand extensive amount of supervised training data from pathologists and may still perform poorly in images from unseen tissue types. We propose an unsupervised approach for histopathology image segmentation that synthesizes heterogeneous sets of training image patches, of every tissue type. Although our synthetic patches are not always of high quality, we harness the motley crew of generated samples through a generally applicable importance sampling method. This proposed approach, for the first time, re-weighs the training loss over synthetic data so that the ideal (unbiased) generalization loss over the true data distribution is minimized. This enables us to use a random polygon generator to synthesize approximate cellular structures (i.e., nuclear masks) for which no real examples are given in many tissue types, and hence, GAN-based methods are not suited. In addition, we propose a hybrid synthesis pipeline that utilizes textures in real histopathology patches and GAN models, to tackle heterogeneity in tissue textures. Compared with existing state-of-the-art supervised models, our approach generalizes significantly better on cancer types without training data. Even in cancer types with training data, our approach achieves the same performance without supervision cost. We release code and segmentation results on over 5000 Whole Slide Images (WSI) in The Cancer Genome Atlas (TCGA) repository, a dataset that would be orders of magnitude larger than what is available today.

A dense multi-path decoder for tissue segmentation in histopathology images

BACKGROUND AND OBJECTIVE

Segmenting different tissue components in histopathological images is of great importance for analyzing tissues and tumor environments. In recent years, an encoder-decoder family of convolutional neural networks has increasingly adopted to develop automated segmentation tools. While an encoder has been the main focus of most investigations, the role of a decoder so far has not been well studied and understood. Herein, we proposed an improved design of a decoder for the segmentation of epithelium and stroma components in histopathology images.

METHODS

The proposed decoder is built upon a multi-path layout and dense shortcut connections between layers to maximize the learning and inference capability. Equipped with the proposed decoder, neural networks are built using three types of encoders (VGG, ResNet and preactived ResNet). To assess the proposed method, breast and prostate tissue datasets are utilized, including 108 and 52 hematoxylin and eosin (H&E) breast tissues images and 224 H&E prostate tissue images.

RESULTS

Combining the pre-activated ResNet encoder and the proposed decoder, we achieved a pixel wise accuracy (ACC) of 0.9122, a rand index (RAND) score of 0.8398, an area under receiver operating characteristic curve (AUC) of 0.9716, Dice coefficient for stroma (DICE_STR) of 0.9092 and Dice coefficient for epithelium (DICE_EPI) of 0.9150 on the breast tissue dataset. The same network obtained 0.9074 ACC, 0.8320 Rand index, 0.9719 AUC, 0.9021 DICE_EPI and 0.9121 DICE_STR on the prostate dataset.

CONCLUSIONS

In general, the experimental results confirmed that the proposed network is superior to the networks combined with the conventional decoder. Therefore, the proposed decoder could aid in improving tissue analysis in histopathology images.

Unsupervised Feature Extraction via Deep Learning for Histopathological Classification of Colon Tissue Images

Histopathological examination is today's gold standard for cancer diagnosis. However, this task is time consuming and prone to errors as it requires a detailed visual inspection and interpretation of a pathologist. Digital pathology aims at alleviating these problems by providing computerized methods that quantitatively analyze digitized histopathological tissue images. The performance of these methods mainly relies on the features that they use, and thus, their success strictly depends on the ability of these features by successfully quantifying the histopathology domain. With this motivation, this paper presents a new unsupervised feature extractor for effective representation and classification of histopathological tissue images. This feature extractor has three main contributions: First, it proposes to identify salient subregions in an image, based on domain-specific prior knowledge, and to quantify the image by employing only the characteristics of these subregions instead of considering the characteristics of all image locations. Second, it introduces a new deep learning-based technique that quantizes the salient subregions by extracting a set of features directly learned on image data and uses the distribution of these quantizations for image representation and classification. To this end, the proposed deep learning-based technique constructs a deep belief network of the restricted Boltzmann machines (RBMs), defines the activation values of the hidden unit nodes in the final RBM as the features, and learns the quantizations by clustering these features in an unsupervised way. Third, this extractor is the first example for successfully using the restricted Boltzmann machines in the domain of histopathological image analysis. Our experiments on microscopic colon tissue images reveal that the proposed feature extractor is effective to obtain more accurate classification results compared to its counterparts.

Double Deep Autoencoder for Heterogeneous Distributed Clustering

Given the issues relating to big data and privacy-preserving challenges, distributed data mining (DDM) has received much attention recently. Here, we focus on the clustering problem of distributed environments. Several distributed clustering algorithms have been proposed to solve this problem, however, previous studies have mainly considered homogeneous data. In this paper, we develop a double deep autoencoder structure for clustering in distributed and heterogeneous datasets. Three datasets are used to demonstrate the proposed algorithms, and show their usefulness according to the consistent accuracy index.

Myocardium Detection by Deep SSAE Feature and Within-Class Neighborhood Preserved Support Vector Classifier and Regressor

Automatic detection of left ventricle myocardium is essential to subsequent cardiac image registration and tissue segmentation. However, it is considered challenging mainly because of the complex and varying shape of the myocardium and surrounding tissues across slices and phases. In this study, a hybrid model is proposed to detect myocardium in cardiac magnetic resonance (MR) images combining region proposal and deep feature classification and regression. The model firstly generates candidate regions using new structural similarity-enhanced supervoxel over-segmentation plus hierarchical clustering. Then it adopts a deep stacked sparse autoencoder (SSAE) network to learn the discriminative deep feature to represent the regions. Finally, the features are fed to train a novel nonlinear within-class neighborhood preserved soft margin support vector (C-SVC) classifier and multiple-output support vector (ε-SVR) regressor for refining the location of myocardium. To improve the stability and generalization, the model also takes hard negative sample mining strategy to fine-tune the SSAE and the classifier. The proposed model with impacts of different components were extensively evaluated and compared to related methods on public cardiac data set. Experimental results verified the effectiveness of proposed integrated components, and demonstrated that it was robust in myocardium localization and outperformed the state-of-the-art methods in terms of typical metrics. This study would be beneficial in some cardiac image processing such as region-of-interest cropping and left ventricle volume measurement.

Comparative assessment of CNN architectures for classification of breast FNAC images

Fine needle aspiration cytology (FNAC) entails using a narrow gauge (25-22 G) needle to collect a sample of a lesion for microscopic examination. It allows a minimally invasive, rapid diagnosis of tissue but does not preserve its histological architecture. FNAC is commonly used for diagnosis of breast cancer, with traditional practice being based on the subjective visual assessment of the breast cytopathology cell samples under a microscope to evaluate the state of various cytological features. Therefore, there are many challenges in maintaining consistency and reproducibility of findings. However, the advent of digital imaging and computational aid in diagnosis can improve the diagnostic accuracy and reduce the effective workload of pathologists. This paper presents a comparison of various deep convolutional neural network (CNN) based fine-tuned transfer learned classification approach for the diagnosis of the cell samples. The proposed approach has been tested using VGG16, VGG19, ResNet-50 and GoogLeNet-V3 (aka Inception V3) architectures of CNN on an image dataset of 212 images (99 benign and 113 malignant), later augmented and cleansed to 2120 images (990 benign and 1130 malignant), where the network was trained using images of 80% cell samples and tested on the rest. This paper presents a comparative assessment of the models giving a new dimension to FNAC study where GoogLeNet-V3 (fine-tuned) achieved an accuracy of 96.25% which is highly satisfactory.

Reconstruction of Hidden Representation for Robust Feature Extraction

This article aims to develop a new and robust approach to feature representation. Motivated by the success of Auto-Encoders, we first theoretically analyze and summarize the general properties of all algorithms that are based on traditional Auto-Encoders: (1) The reconstruction error of the input cannot be lower than a lower bound, which can be viewed as a guiding principle for reconstructing the input. Additionally, when the input is corrupted with noises, the reconstruction error of the corrupted input also cannot be lower than a lower bound. (2) The reconstruction of a hidden representation achieving its ideal situation is the necessary condition for the reconstruction of the input to reach the ideal state. (3) Minimizing the Frobenius norm of the Jacobian matrix of the hidden representation has a deficiency and may result in a much worse local optimum value. We believe that minimizing the reconstruction error of the hidden representation is more robust than minimizing the Frobenius norm of the Jacobian matrix of the hidden representation. Based on the above analysis, we propose a new model termed Double Denoising Auto-Encoders (DDAEs), which uses corruption and reconstruction on both the input and the hidden representation. We demonstrate that the proposed model is highly flexible and extensible and has a potentially better capability to learn invariant and robust feature representations. We also show that our model is more robust than Denoising Auto-Encoders (DAEs) for dealing with noises or inessential features. Furthermore, we detail how to train DDAEs with two different pretraining methods by optimizing the objective function in a combined and separate manner, respectively. Comparative experiments illustrate that the proposed model is significantly better for representation learning than the state-of-the-art models.

Recent Advances of Deep Learning in Bioinformatics and Computational Biology

Extracting inherent valuable knowledge from omics big data remains as a daunting problem in bioinformatics and computational biology. Deep learning, as an emerging branch from machine learning, has exhibited unprecedented performance in quite a few applications from academia and industry. We highlight the difference and similarity in widely utilized models in deep learning studies, through discussing their basic structures, and reviewing diverse applications and disadvantages. We anticipate the work can serve as a meaningful perspective for further development of its theory, algorithm and application in bioinformatic and computational biology.

Real-time gastric polyp detection using convolutional neural networks

Computer-aided polyp detection in gastric gastroscopy has been the subject of research over the past few decades. However, despite significant advances, automatic polyp detection in real time is still an unsolved problem. In this paper, we report on a convolutional neural network (CNN) for polyp detection that is constructed based on Single Shot MultiBox Detector (SSD) architecture and which we call SSD for Gastric Polyps (SSD-GPNet). To take full advantages of feature maps' information from the feature pyramid and to acquire higher accuracy, we re-use information that is abandoned by Max-Pooling layers. In other words, we reuse the lost data from the pooling layers and concatenate that data as extra feature maps to contribute to classification and detection. Meanwhile, in the feature pyramid, we concatenate feature maps of the lower layers and feature maps that are deconvolved from upper layers to make explicit relationships between layers and to effectively increase the number of channels. The results show that our enhanced SSD for gastric polyp detection can realize real-time polyp detection with 50 frames per second (FPS) and can improve the mean average precision (mAP) from 88.5% to 90.4%, with only a little loss in time-performance. And the further experiment shows that SSD-GPNet has excellent performance in improving polyp detection recalls over 10% (p = 0.00053), especially in small polyp detection. This can help endoscopic physicians more easily find missed polyps and decrease the gastric polyp miss rate. It may be applicable in daily clinical practice to reduce the burden on physicians.

Convolutional neural network initialized active contour model with adaptive ellipse fitting for nuclear segmentation on breast histopathological images

Automated detection and segmentation of nuclei from high-resolution histopathological images is a challenging problem owing to the size and complexity of digitized histopathologic images. In the context of breast cancer, the modified Bloom-Richardson Grading system is highly correlated with the morphological and topological nuclear features are highly correlated with Modified Bloom-Richardson grading. Therefore, to develop a computer-aided prognosis system, automated detection and segmentation of nuclei are critical prerequisite steps. We present a method for automated detection and segmentation of breast cancer nuclei named a convolutional neural network initialized active contour model with adaptive ellipse fitting (CoNNACaeF). The CoNNACaeF model is able to detect and segment nuclei simultaneously, which consist of three different modules: convolutional neural network (CNN) for accurate nuclei detection, (2) region-based active contour (RAC) model for subsequent nuclear segmentation based on the initial CNN-based detection of nuclear patches, and (3) adaptive ellipse fitting for overlapping solution of clumped nuclear regions. The performance of the CoNNACaeF model is evaluated on three different breast histological data sets, comprising a total of 257 H&E-stained images. The model is shown to have improved detection accuracy of F-measure 80.18%, 85.71%, and 80.36% and average area under precision-recall curves (AveP) 77%, 82%, and 74% on a total of 3 million nuclei from 204 whole slide images from three different datasets. Additionally, CoNNACaeF yielded an F-measure at 74.01% and 85.36%, respectively, for two different breast cancer datasets. The CoNNACaeF model also outperformed the three other state-of-the-art nuclear detection and segmentation approaches, which are blue ratio initialized local region active contour, iterative radial voting initialized local region active contour, and maximally stable extremal region initialized local region active contour models.

Sparse Autoencoder for Unsupervised Nucleus Detection and Representation in Histopathology Images

We propose a sparse Convolutional Autoencoder (CAE) for simultaneous nucleus detection and feature extraction in histopathology tissue images. Our CAE detects and encodes nuclei in image patches in tissue images into sparse feature maps that encode both the location and appearance of nuclei. A primary contribution of our work is the development of an unsupervised detection network by using the characteristics of histopathology image patches. The pretrained nucleus detection and feature extraction modules in our CAE can be fine-tuned for supervised learning in an end-to-end fashion. We evaluate our method on four datasets and achieve state-of-the-art results. In addition, we are able to achieve comparable performance with only 5% of the fully- supervised annotation cost.

Diagnosis of mesothelioma with deep learning

Malignant mesothelioma (MM) is a rare but aggressive cancer. The definitive diagnosis of MM is critical for effective treatment and has important medicolegal significance. However, the definitive diagnosis of MM is challenging due to its composite epithelial/mesenchymal pattern. The aim of the current study was to develop a deep learning method to automatically diagnose MM. A retrospective analysis of 324 participants with or without MM was performed. Significant features were selected using a genetic algorithm (GA) or a ReliefF algorithm performed in MATLAB software. Subsequently, the current study constructed and trained several models based on a backpropagation (BP) algorithm, extreme learning machine algorithm and stacked sparse autoencoder (SSAE) to diagnose MM. A confusion matrix, F-measure and a receiver operating characteristic (ROC) curve were used to evaluate the performance of each model. A total of 34 potential variables were analyzed, while the GA and ReliefF algorithms selected 19 and 5 effective features, respectively. The selected features were used as the inputs of the three models. SSAE and GA+SSAE demonstrated the highest performance in terms of classification accuracy, specificity, F-measure and the area under the ROC curve. Overall, the GA+SSAE model was the preferred model since it required a shorter CPU time and fewer variables. Therefore, the SSAE with GA feature selection was selected as the most accurate model for the diagnosis of MM. The deep learning methods developed based on the GA+SSAE model may assist physicians with the diagnosis of MM.

Deep Learning: Current and Emerging Applications in Medicine and Technology

Machine learning is enabling researchers to analyze and understand increasingly complex physical and biological phenomena in traditional fields such as biology, medicine, and engineering and emerging fields like synthetic biology, automated chemical synthesis, and biomanufacturing. These fields require new paradigms toward understanding increasingly complex data and converting such data into medical products and services for patients. The move toward deep learning and complex modeling is an attempt to bridge the gap between acquiring massive quantities of complex data, and converting such data into practical insights. Here, we provide an overview of the field of machine learning, its current applications and needs in traditional and emerging fields, and discuss an illustrative attempt at using deep learning to understand swarm behavior of molecular shuttles.

Automatic Vertebrae Localization and Identification by Combining Deep SSAE Contextual Features and Structured Regression Forest

Automatic vertebrae localization and identification in medical computed tomography (CT) scans is of great value for computer-aided spine diseases diagnosis. In order to overcome the disadvantages of the approaches employing hand-crafted, low-level features and based on field-of-view priori assumption of spine structure, an automatic method is proposed to localize and identify vertebrae by combining deep stacked sparse autoencoder (SSAE) contextual features and structured regression forest (SRF). The method employs SSAE to learn image deep contextual features instead of hand-crafted ones by building larger-range input samples to improve their contextual discrimination ability. In the localization and identification stage, it incorporates the SRF model to achieve whole spine localization, then screens those vertebrae within the image, thus relieves the assumption that the part of spine in the field of image is visible. In the end, the output distribution of SRF and spine CT scans properties are assembled to develop a two-stage progressive refining strategy, where the mean-shift kernel density estimation and Otsu method instead of Markov random field (MRF) are adopted to reduce model complexity and refine vertebrae localization results. Extensive evaluation was performed on a challenging data set of 98 spine CT scans. Compared with the hidden Markov model and the method based on convolutional neural network (CNN), the proposed approach could effectively and automatically locate and identify spinal targets in CT scans, and achieve higher localization accuracy, low model complexity, and no need for any assumptions about visual field in CT scans.

Identification of topological features in renal tumor microenvironment associated with patient survival

Motivation

As a highly heterogeneous disease, the progression of tumor is not only achieved by unlimited growth of the tumor cells, but also supported, stimulated, and nurtured by the microenvironment around it. However, traditional qualitative and/or semi-quantitative parameters obtained by pathologist’s visual examination have very limited capability to capture this interaction between tumor and its microenvironment. With the advent of digital pathology, computerized image analysis may provide a better tumor characterization and give new insights into this problem.

Results

We propose a novel bioimage informatics pipeline for automatically characterizing the topological organization of different cell patterns in the tumor microenvironment. We apply this pipeline to the only publicly available large histopathology image dataset for a cohort of 190 patients with papillary renal cell carcinoma obtained from The Cancer Genome Atlas project. Experimental results show that the proposed topological features can successfully stratify early- and middle-stage patients with distinct survival, and show superior performance to traditional clinical features and cellular morphological and intensity features. The proposed features not only provide new insights into the topological organizations of cancers, but also can be integrated with genomic data in future studies to develop new integrative biomarkers.

Availability and implementation

https://github.com/chengjun583/KIRP-topological-features

Supplementary information

Supplementary data are available atBioinformatics online.