A review of graph-based methods for image analysis in digital histopathology (original) (raw)
Related papers
Automatic image analysis of histopathology specimens using concave vertex graph
Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention, 2008
Automatic image analysis of histopathology specimens would help the early detection of blood cancer. The first step for automatic image analysis is segmentation. However, touching cells bring the difficulty for traditional segmentation algorithms. In this paper, we propose a novel algorithm which can reliably handle touching cells segmentation. Robust estimation and color active contour models are used to delineate the outer boundary. Concave points on the boundary and inner edges are automatically detected. A concave vertex graph is constructed from these points and edges. By minimizing a cost function based on morphological characteristics, we recursively calculate the optimal path in the graph to separate the touching cells. The algorithm is computationally efficient and has been tested on two large clinical dataset which contain 207 images and 3898 images respectively. Our algorithm provides better results than other studies reported in the recent literature.
Histographs: graphs in histopathology
Medical Imaging 2020: Digital Pathology
Spatial arrangement of cells of various types, such as tumor infiltrating lymphocytes and the advancing edge of a tumor, are important features for detecting and characterizing cancers. However, convolutional neural networks (CNNs) do not explicitly extract intricate features of the spatial arrangements of the cells from histopathology images. In this work, we propose to classify cancers using graph convolutional networks (GCNs) by modeling a tissue section as a multi-attributed spatial graph of its constituent cells. Cells are detected using their nuclei in H&E stained tissue image, and each cell's appearance is captured as a multi-attributed high-dimensional vertex feature. The spatial relations between neighboring cells are captured as edge features based on their distances in a graph. We demonstrate the utility of this approach by obtaining classification accuracy that is competitive with CNNs, specifically, Inception-v3, on two tasks-cancerous versus non-cancerous and in situ versus invasive-on the BACH breast cancer dataset.
2010
The characterization and quantitative description of histological images is not a simple problem. To reach a final diagnosis, usually the specialist relies on the analysis of characteristics easily observed, such as cells size, shape, staining and texture, but also depends on the hidden information of tissue localization, physiological and pathological mechanisms, clinical aspects, or other etiological agents. In this paper, Mathematical Morphology (MM) and Machine Learning (ML) methods were applied to characterize and classify histological images. MM techniques were employed for image analysis. The measurements obtained from image and graph analysis were fed into Machine Learning algorithms, which were designed and developed to automatically learn to recognize complex patterns and make intelligent decisions based on data. Specifically, a linear Support Vector Machine (SVM) was used to evaluate the discriminatory power of the used measures. The results show that the methodology was successful in characterizing and classifying the differences between the architectural organization of epithelial and adipose tissues. We believe that this approach can be also applied to classify and help the diagnosis of many tissue abnormalities, such as cancers.
Graph Run-Length Matrices for Histopathological Image Segmentation
The histopathological examination of tissue specimens is essential for cancer diagnosis and grading. However, this examination is subject to a considerable amount of observer variability as it mainly relies on visual interpretation of pathologists. To alleviate this problem, it is very important to develop computational quantitative tools, for which image segmentation constitutes the core step. In this paper, we introduce an effective and robust algorithm for the segmentation of histopathological tissue images. This algorithm incorporates the background knowledge of the tissue organization into segmentation. For this purpose, it quantifies spatial relations of cytological tissue components by constructing a graph and uses this graph to define new texture features for image segmentation. This new texture definition makes use of the idea of gray-level run-length matrices. However, it considers the runs of cytological components on a graph to form a matrix, instead of considering the runs of pixel intensities. Working with colon tissue images, our experiments demonstrate that the texture features extracted from "graph run-length matrices" lead to high segmentation accuracies, also providing a reasonable number of segmented regions. Compared with four other segmentation algorithms, the results show that the proposed algorithm is more effective in histopathological image segmentation.
Histopathological Image Analysis: A Review
IEEE Reviews in Biomedical Engineering, 2009
Over the past decade, dramatic increases in computational power and improvement in image analysis algorithms have allowed the development of powerful computer-assisted analytical approaches to radiological data. With the recent advent of whole slide digital scanners, tissue histopathology slides can now be digitized and stored in digital image form. Consequently, digitized tissue histopathology has now become amenable to the application of computerized image analysis and machine learning techniques. Analogous to the role of computer-assisted diagnosis (CAD) algorithms in medical imaging to complement the opinion of a radiologist, CAD algorithms have begun to be developed for disease detection, diagnosis, and prognosis prediction to complement the opinion of the pathologist. In this paper, we review the recent state of the art CAD technology for digitized histopathology. This paper also briefly describes the development and application of novel image analysis technology for a few specific histopathology related problems being pursued in the United States and Europe.
Computerized Medical Imaging and Graphics, 2011
In this paper, we present a graph-based multi-resolution approach for mitosis extraction in breast cancer histological whole slide images. The proposed segmentation uses a multi-resolution approach which reproduces the slide examination done by a pathologist. Each resolution level is analyzed with a focus of attention resulting from a coarser resolution level analysis. At each resolution level, a spatial refinement by label regularization is performed to obtain more accurate segmentation around boundaries. The proposed segmentation is fully unsupervised by using domain specific knowledge. (V. Roullier). histopathological images is a very challenging task . First, the produced images are relatively huge and their processing requires computationally efficient tools. Second, the biological variability of the objects of interest makes their extraction difficult. As a consequence, few works in literature have considered the processing of whole slide images and most of these works rely only on machine learning techniques .
HistoCartography: A Toolkit for Graph Analytics in Digital Pathology
ArXiv, 2021
Advances in entity-graph based analysis of histopathology images have brought in a new paradigm to describe tissue composition, and learn the tissue structure-to-function relationship. Entity-graphs offer flexible and scalable representations to characterize tissue organization, while allowing the incorporation of prior pathological knowledge to further support model interpretability and explainability. However, entity-graph analysis requires prerequisites for image-to-graph translation and knowledge of state-of-the-art machine learning algorithms applied to graph-structured data, which can potentially hinder their adoption. In this work, we aim to alleviate these issues by developing HistoCartography, a standardized python API with necessary preprocessing, machine learning and explainability tools to facilitate graph-analytics in computational pathology. Further, we have benchmarked the computational time and performance on multiple datasets across different imaging types and histo...
A Brief Study on Histopathological Images
INTERNATIONAL JOURNAL OF RECENT TRENDS IN ENGINEERING & RESEARCH
The process of examining biological tissue under a microscope for detecting the severity of the disease is called histology, it is an essential technique in biomedical research and clinical practice. While slide preparation and imaging is increasingly becoming automated, but the analysis of histology images still require the trained eyes of a pathologist to examine under a microscope. Processing tissues from histopathological images has become now fully computerized, significantly increasing the speed, the labs can produce tissue slides for viewing images digitally. Digitizing these slides, allows pathologist to view these slides on a computer rather than on microscope. routine analysis of tissues selection will be very difficult, manual task that can be completed only by trained pathologists at a huge cost. In the clinical domain, these methods could improve the accuracy and consistency of diagnoses and hence pathologists can focus on the most difficult cases. This research domain could complete the tasks that are time-consuming for humans, and discover new diseases from millions of whole-slide images (WSIs) or precisely delineating tissues within a tumor, allowing for a quantitative comparison of tumours grown under different conditions.
The pathological image segmentation is important in cancer diagnosis and grading. In human body, tissues are characterized with the organization of their components. Cancer causes the changes in these organization. In order to diagnose the cancer disease, pathologist visually examine the changes in the tissue. This examination mainly relies on the visual interpretation. It may lead to considerable amount of observer variability. Hence, they may or may not identify the abnormal tissue. To avoid this problem robust algorithms are introduced for segmentation. Graph Run Length Method (GRLM), Gray Level Co-occurrence Matrix (GLCM) provides efficient way to segment the abnormal tissue. To a pathological image color graph was automatically generated by using Graph Run Length Method (GRLM). Gray Level Co-occurrence Matrix (GLCM) provides texture features of pathological image. The graph provides the arrangement of cells and structure of cells in a tissue. Based on the arrangement of cells, structure of cells, GLCM based texture features we can segment the abnormal tissue efficiently.
SPATIAL INTERACTION ANALYSIS WITH GRAPH BASED MATHEMATICAL MORPHOLOGY FOR HISTOPATHOLGY
Exploring the spatial interactions between tumor and the inflammatory microenvironment using digital pathology image analysis can contribute to a better understanding of the immune function and tumor heterogeneity. We address this by providing tools able to reveal various metrics describing spatial relationships in the cancer ecosystem. The approach comprises nuclei segmentation and classification, using supervised learning algorithm, to detect lymphoid aggregates and tumor patterns, and spatial distribution quantification using sparse sets' mathematical morphology. Tumor patterns were classified into three groups: surrounded by lympho-cytes, close to lymphoid aggregates or distant and might be protected from immune attack. The approach provides statistical assessment and comprehensive visual representation of the inflammatory tumor microenvironment.