A Deep Feature Extraction Method for HEp-2 Cell Image Classification

Abstract

The automated and accurate classification of the images portraying the Human Epithelial cells of type 2 (HEp-2) represents one of the most important steps in the diagnosis procedure of many autoimmune diseases. The extreme intra-class variations of the HEp-2 cell images datasets drastically complicates the classification task. We propose in this work a classification framework that, unlike most of the state-of-the-art methods, uses a deep learning-based feature extraction method in a strictly unsupervised way. We propose a deep learning-based hybrid feature learning with two levels of deep convolutional autoencoders. The first level takes the original cell images as the inputs and learns to reconstruct them, in order to capture the features related to the global shape of the cells, and the second network takes the gradients of the images, in order to encode the localized changes in intensity (gray variations) that characterize each cell type. A final feature vector is constructed by combining the latent representations extracted from the two networks, giving a highly discriminative feature representation. The created features will be fed to a nonlinear classifier whose output will represent the type of the cell image. We have tested the discriminability of the proposed features on two of the most popular HEp-2 cell classification datasets, the SNPHEp-2 and ICPR 2016 datasets. The results show that the proposed features manage to capture the distinctive characteristics of the different cell types while performing at least as well as the actual deep learning-based state-of-the-art methods in terms of discrimination.

Automatic Detection Method for Cancer Cell Nucleus Image Based on Deep-Learning Analysis and Color Layer Signature Analysis Algorithm

“…Song et al proposed using contour fragments produced from cell blocks to plot graphs of the minimal energy function, and fragments from the same cytoplasm were placed in the same set to achieve automatic segmentation of the overlapping cytoplasm [16]. Vununu et al proposed a deep feature extraction method for HEp-2 cell image classification [17]. Kucharski et al proposed a semi-supervised segmentation method to solve the problem of ground-truth images segmentation for the detection of nests of nevus cells in histopathological images of skin specimens [18].…”

Dealing with Lack of Training Data for Convolutional Neural Networks: The Case of Digital Pathology

“…Among feature learning methods, deep learning and, more specifically, Convolutional Neural Networks (CNNs) have now become a major trend in many computer vision and medical tasks [10][11][12]. In CNNs, a number of convolutional and pooling layers learn by backpropagation the set of features that are best for classification, thus avoiding the design of handcrafted texture descriptors.…”

Acute lymphoblastic leukemia cells image analysis with deep bagging ensemble learning

“…Recently, with developing of deep learning (DL) based techniques, researches have investigated the potential use of applying DL methods in cell image classification problems. Compared with conventional hand-craft features CV based solutions, DL based techniques have the advantages with automatically selecting representative image features of cells [12] and reduce the difficulty of solving ALL cells classification through a series of standard DL procedures.…”