Hover-Net: Simultaneous segmentation and classification of nuclei in multi-tissue histology images

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.

The advent of digital pathology provides us with the challenging opportunity to automatically analyze whole slides of diseased tissue in order to derive quantitative profiles that can be used for diagnosis and prognosis tasks. In particular, for the development of interpretable models, the detection and segmentation of cell nuclei is of the utmost importance. In this paper, we describe a new method to automatically segment nuclei from Haematoxylin and Eosin (H&E) stained histopathology data with fully convolutional networks. In particular, we address the problem of segmenting touching nuclei by formulating the segmentation problem as a regression task of the distance map. We demonstrate superior performance of this approach as compared to other approaches using Convolutional Neural Networks.

Segmentation of Nuclei in Histopathology Images by Deep Regression of the Distance Map

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

Time–frequency localized three-band biorthogonal wavelet filter bank using semidefinite relaxation and nonlinear least squares with epileptic seizure EEG signal classification

Automatic cell nuclei segmentation and classification of breast cancer histopathology images

The invention discloses a mammary glandular cell segmentation method based on a multi-scale growth and double-strategy adhesion-removing model. The method comprises the following steps: firstly, inputting a mammary glandular tissue image and converting the image into a gray image; secondly, enhancing the contrast ratio; thirdly, carrying out cell positioning by using wavelet decomposition; fourthly, carrying out multi-scale region growth; fifthly, realizing primary segmentation of a cell region through voting and selecting; sixthly, judging whether the segmented region has cell adhesion or not; if the cell adhesion does not exist, determining that the segmented region is a single cell region, and outputting a segmentation result; if the cell adhesion exists, determining that the segmented region is an adhesion region, and carrying out adherent cell segmentation; and finally, carrying out adherent cell segmentation by using the double-strategy adhesion-removing model constructed by morphological corrosion-expansion operation and a corner detection segmentation algorithm until all the cells are segmented. By virtue of the method, the influences on mammary glandular cell segmentation, caused by a complicated background of a mammary glandular tissue slice image, are effectively inhibited; and the identification precision of an adherent cell segmentation line is improved and the segmentation precision of the adherent cells is further improved.

Mammary glandular cell segmentation method based on multi-scale growth and double-strategy adhesion-removing model

The invention discloses a thoracic impedance signal processing method in cardio-pulmonary resuscitation. The thoracic impedance signal processing method comprises the following steps of: S1, pre-processing, and removing high-frequency noise and baseline drift; S2, marking waveforms based on a multi-resolution window search method, and extracting width features L(i), amplitude features H(i) and adjacent waveform feature differences Dif(i) of each waveform; S3, decomposing wavelets, and extracting wavelet features Dw(i) of each waveform; S4, constituting a four-dimensional feature vector by using the L(i), H(i), Dif(i) and Dw(i), and performing normalization processing; S5, classifying and identifying; S6, calculating quality evaluation indexes of cardio-pulmonary resuscitation. The thoracic impedance signal processing method in cardio-pulmonary resuscitation has the significant effects that: pressing and aerating waveforms can be accurately identified, pseudo pressing waveforms are effectively removed, the positive detection rate and sensitive of the algorithm to be very ideal, the requirement of processing in real time can be achieved, various output evaluation indexes can be fed back to cardio-pulmonary resuscitation operation staffs in real time, thus being convenient for operation staffs to adjust operation of cardio-pulmonary resuscitation and achieving the purpose of providing guidance to a CPR process.

Liu Qianqian

Papers

Automatic cell nuclei segmentation and classification of breast cancer histopathology images

Mammary glandular cell segmentation method based on multi-scale growth and double-strategy adhesion-removing model

Thoracic impedance signal processing method in cardio-pulmonary resuscitation