Automated breast cancer multi-classification from histopathological images plays a key role in computer-aided breast cancer diagnosis or prognosis. Breast cancer multi-classification is to identify subordinate classes of breast cancer (Ductal carcinoma, Fibroadenoma, Lobular carcinoma, etc.). However, breast cancer multi-classification from histopathological images faces two main challenges from: (1) the great difficulties in breast cancer multi-classification methods contrasting with the classification of binary classes (benign and malignant), and (2) the subtle differences in multiple classes due to the broad variability of high-resolution image appearances, high coherency of cancerous cells, and extensive inhomogeneity of color distribution. Therefore, automated breast cancer multi-classification from histopathological images is of great clinical significance yet has never been explored. Existing works in literature only focus on the binary classification but do not support further breast cancer quantitative assessment. In this study, we propose a breast cancer multi-classification method using a newly proposed deep learning model. The structured deep learning model has achieved remarkable performance (average 93.2% accuracy) on a large-scale dataset, which demonstrates the strength of our method in providing an efficient tool for breast cancer multi-classification in clinical settings.

Breast Cancer Multi-classification from Histopathological Images with Structured Deep Learning Model.

Applications of deep learning in fundus images: A review.

A Fixed-Point Continuation Method for L_1-Regularization with Application to Compressed Sensing

The automatic and precision classification for breast cancer histopathological image has a great significance in clinical application. However, the existing analysis approaches are difficult to addressing the breast cancer classification problem because the feature subtle differences of inter-class histopathological image and the classification accuracy still hard to meet the clinical application. Recent advancements in data-driven sharing processing and multi-level hierarchical feature learning have made available considerable chance to dope out a solution to this problem. To address the challenging problem, we propose a novel breast cancer histopathological image classification method based on deep convolutional neural networks, named as BiCNN model, to address the two-class breast cancer classification on the pathological image. This deep learning model considers class and sub-class labels of breast cancer as prior knowledge, which can restrain the distance of features of different breast cancer pathological images. In addition, an advanced data augmented method is proposed to fit tolerance whole slide image recognition, which can full reserve image edge feature of cancerization region. The transfer learning and fine-tuning method are adopted as an optimal training strategy to improve breast cancer histopathological image classification accuracy. The experiment results show that the proposed method leads to a higher classification accuracy (up to 97%) and displays good robustness and generalization, which provides efficient tools for breast cancer clinical diagnosis.

Deep learning model based breast cancer histopathological image classification

Breast cancer continues to be among the leading causes of death for women and much effort has been expended in the form of screening programs for prevention. Given the exponential growth in the number of mammograms collected by these programs, computer-assisted diagnosis has become a necessity. Computer-assisted detection techniques developed to date to improve diagnosis without multiple systematic readings have not resulted in a significant improvement in performance measures. In this context, the use of automatic image processing techniques resulting from deep learning represents a promising avenue for assisting in the diagnosis of breast cancer. In this paper, we present a deep learning approach based on a Convolutional Neural Network (CNN) model for multi-class breast cancer classification. The proposed approach aims to classify the breast tumors in non-just benign or malignant but we predict the subclass of the tumors like Fibroadenoma, Lobular carcinoma, etc. Experimental results on histopathological images using the BreakHis dataset show that the DenseNet CNN model achieved high processing performances with 95.4% of accuracy in the multi-class breast cancer classification task when compared with state-of-the-art models.

/pdf/multi-class-breast-cancer-classification-using-deep-learning-2vz3nwgoui.pdf

Multi-class breast cancer classification using deep learning convolutional neural network

It is crucial to understand the specificity of HIV-1 protease for designing HIV-1 protease inhibitors In this paper, a new feature selection method combined with neural network structure optimization is proposed to analyze the specificity of HIV-1 protease and find the important positions in an octapeptide that determined its cleavability Two kinds of newly proposed features based on Amino Acid Index database plus traditional orthogonal encoding features are used in this paper, taking both physiochemical and sequence information into consideration Results of feature selection prove that , , , and are the most important positions Two feature fusion methods are used in this paper: combination fusion and decision fusion aiming to get comprehensive feature representation and improve prediction performance Decision fusion of subsets that getting after feature selection obtains excellent prediction performance, which proves feature selection combined with decision fusion is an effective and useful method for the task of HIV-1 protease cleavage site prediction The results and analysis in this paper can provide useful instruction and help designing HIV-1 protease inhibitor in the future

/pdf/feature-selection-combined-with-neural-network-structure-2k2n84yzhi.pdf

Feature Selection Combined with Neural Network Structure Optimization for HIV-1 Protease Cleavage Site Prediction

Purpose: To develop and evaluate a computerized semiautomatic segmentation method for accurate extraction of three-dimensional lesions from dynamic contrast-enhanced magnetic resonance images (DCE-MRIs) of the breast. Methods: The authors propose a new background distribution-based active contour model using level set (BDACMLS) to segment lesions in breast DCE-MRIs. The method starts with manual selection of a region of interest (ROI) that contains the entire lesion in a single slice where the lesion is enhanced. Then the lesion volume from the volume data of interest, which is captured automatically, is separated. The core idea of BDACMLS is a new signed pressure function which is based solely on the intensity distribution combined with pathophysiological basis. To compare the algorithm results, two experienced radiologists delineated all lesions jointly to obtain the ground truth. In addition, results generated by other different methods based on level set (LS) are also compared with the authors’ method. Finally, the performance of the proposed method is evaluated by several region-based metrics such as the overlap ratio. Results: Forty-two studies with 46 lesions that contain 29 benign and 17 malignant lesions are evaluated. The dataset includes various typical pathologies of the breast such as invasive ductal carcinoma, ductal carcinomain situ,more » scar carcinoma, phyllodes tumor, breast cysts, fibroadenoma, etc. The overlap ratio for BDACMLS with respect to manual segmentation is 79.55% ± 12.60% (mean ± s.d.). Conclusions: A new active contour model method has been developed and shown to successfully segment breast DCE-MRI three-dimensional lesions. The results from this model correspond more closely to manual segmentation, solve the weak-edge-passed problem, and improve the robustness in segmenting different lesions.« less

A new background distribution-based active contour model for three-dimensional lesion segmentation in breast DCE-MRI

Accurate staging of hepatic cirrhosis is important in investigating the cause and slowing down the effects of cirrhosis. Computer-aided diagnosis (CAD) can provide doctors with an alternative second opinion and assist them to make a specific treatment with accurate cirrhosis stage. MRI has many advantages, including high resolution for soft tissue, no radiation, and multiparameters imaging modalities. So in this paper, multisequences MRIs, including T1-weighted, T2-weighted, arterial, portal venous, and equilibrium phase, are applied. However, CAD does not meet the clinical needs of cirrhosis and few researchers are concerned with it at present. Cirrhosis is characterized by the presence of widespread fibrosis and regenerative nodules in the hepatic, leading to different texture patterns of different stages. So, extracting texture feature is the primary task. Compared with typical gray level cooccurrence matrix (GLCM) features, texture classification from random features provides an effective way, and we adopt it and propose CCTCRF for triple classification (normal, early, and middle and advanced stage). CCTCRF does not need strong assumptions except the sparse character of image, contains sufficient texture information, includes concise and effective process, and makes case decision with high accuracy. Experimental results also illustrate the satisfying performance and they are also compared with typical NN with GLCM.

/pdf/cirrhosis-classification-based-on-texture-classification-of-33x8ppawwc.pdf

Cirrhosis classification based on texture classification of random features.

Original intensity preserved inhomogeneity correction and segmentation for liver magnetic resonance imaging

This paper presents a deep random walk technique for drusen segmentation from fundus images. It is formulated as a deep learning architecture which learns deep representations from fundus images and specify an optimal pixel-pixel affinity. Specifically, the proposed architecture is mainly composed of three parts: a deep feature extraction module to learn both semantic-level and low-level representation of image, an affinity learning module to get pixel-pixel affinities for formulating the transition matrix of random walk and a random walk module which propagates manual labels. The power of our technique comes from the fact that the learning procedures for deep image representations and pixel-pixel affinities are driven by the random walk process. The accuracy of our proposed algorithm surpasses state-of-the-art drusen segmentation techniques as validated on the public STARE and DRIVE databases.

Hui Liu

Papers

Feature Selection Combined with Neural Network Structure Optimization for HIV-1 Protease Cleavage Site Prediction

A new background distribution-based active contour model for three-dimensional lesion segmentation in breast DCE-MRI

Cirrhosis classification based on texture classification of random features.

Original intensity preserved inhomogeneity correction and segmentation for liver magnetic resonance imaging

Deep Random Walk for Drusen Segmentation from Fundus Images