No / Clinical morphological analysis of histopathological specimens is a successful manner for diagnosing benign and malignant diseases. Analysis of glandular architecture is a major challenge for colon histopathologists as a result of the difficulty of identifying morphological structures in glandular malignant tumours due to the distortion of glands boundaries, furthermore the variation in the appearance of staining specimens. For reliable analysis of colon specimens, several deep learning methods have exhibited encouraging performance in the glands automatic segmentation despite the challenges. In the histopathology field, the vast number of annotation images for training the deep learning algorithms is the major challenge. In this work, we propose a trainable Convolutional Neural Network (CNN) from end to end for detecting the glands automatically. More specifically, the Modified Res-U-Net is employed for segmenting the colorectal glands in Haematoxylin and Eosin (H&E) stained images for challenging Gland Segmentation (GlaS) dataset. The proposed Res-U-Net outperformed the prior methods that utilise U-Net architecture on the images of the GlaS dataset.

L'un des problèmes les plus complexes dans l'analyse des images histologiques est l'évaluation de l¿organisation spatiale des structures histologiques dans le tissu. En fait, les sections histologiques peuvent contenir un très grand nombre de cellules de différents types et irrégulièrement réparties dans le tissu, ce qui rend leur contenu spatial indescriptible d'une manière simple. Les méthodes fondées sur la théorie des graphes ont été largement explorées dans cette direction, car elles sont des outils de représentation efficaces ayant la capacité expressive de décrire les caractéristiques spatiales et les relations de voisinage qui sont interprétées visuellement par le pathologiste. On peut distinguer trois familles principales de méthodes des graphes utilisées à cette fin: analyse de structure syntaxique, analyse de réseau et analyse spectrale. Cependant, un autre ensemble distinctif de méthodes basées sur la morphologie mathématique sur les graphes peut être développé et adapté pour ce problème. L'objectif principal de cette thèse est le développement d'un outil capable de fournir une évaluation quantitative des arrangements spatiaux des structures histologiques en utilisant la morphologie mathématique basée sur les graphes. / One of the most challenging problems in histological image analysis is the evaluation of the spatial organizations of histological structures in the tissue. In fact, histological sections may contain a very large number of cells of different types and irregularly distributed, which makes their spatial content indescribable in a simple manner. Graph-based methods have been widely explored in this direction, as they are effective representation tools having the expressive ability to describe spatial characteristics and neighborhood relationships that are visually interpreted by the pathologist. We can distinguish three main families of graph-based methods used for this purpose: syntactic structure analysis, network analysis and spectral analysis. However, another distinctive set of methods based on mathematical morphology on graphs can be additionally developed for this issue. The main goal of this dissertation is the development of a framework able to provide quantitative evaluation of the spatial arrangements of histological structures using graph-based mathematical morphology.

yes / Coinciding with advances in whole-slide imaging scanners, it is become essential to automate the conventional image-processing techniques to assist pathologists with some tasks such as mitotic-cells detection. In histopathological images analysing, the mitotic-cells counting is a significant biomarker in the prognosis of the breast cancer grade and its aggressiveness. However, counting task of mitotic-cells is tiresome, tedious and time-consuming due to difficulty distinguishing between mitotic cells and normal cells. To tackle this challenge, several deep learning-based approaches of Computer-Aided Diagnosis (CAD) have been lately advanced to perform counting task of mitotic-cells in the histopathological images. Such CAD systems achieve outstanding performance, hence histopathologists can utilise them as a second-opinion system. However, improvement of CAD systems is an important with the progress of deep learning networks architectures. In this work, we investigate deep YOLO (You Only Look Once) v2 network for mitotic-cells detection on ICPR (International Conference on Pattern Recognition) 2012 dataset of breast cancer histopathology. The obtained results showed that proposed architecture achieves good result of 0.839 F1-measure.