Related Experiment Video
Updated: Jun 6, 2026
11:00
Visualization, Quantification, and Mapping of Immune Cell Populations in the Tumor Microenvironment
Published on: March 25, 2020
An automatic algorithm for the segmentation and morphological analysis of microvessels in immunostained histological
C C Reyes-Aldasoro1, L J Williams, S Akerman
1Department of Oncology, Cancer Research UK Tumour Microcirculation Group, The University of Sheffield, School of Medicine, U.K.
Journal of Microscopy
|December 2, 2010
Summary
A new algorithm automatically segments and analyzes microvessels in tumor histology images using color and shape. This tool accurately quantifies microvessel morphology, aiding in cancer research.
Area of Science:
- Histopathology
- Computational Biology
- Medical Imaging
Background:
- Accurate analysis of tumor microvessel morphology is crucial for understanding cancer progression.
- Manual segmentation and analysis are time-consuming and subjective.
Purpose of the Study:
- To develop and validate a fully automatic algorithm for microvessel segmentation and morphological analysis in CD31 immunostained histological tumor sections.
- To compare microvessel morphology between different tumor types using the developed algorithm.
Main Methods:
- Utilized a region-growing method in the 3D HSV color model based on distinctive hues of stained cells and background.
- Applied post-processing morphological tasks: joining, closing, and splitting segmented objects.
- Validated segmentation accuracy against hand-segmented images (96.3% pixel classification).
- Calculated morphometric parameters: vascular area (VA), lumen-to-vascular area ratio (lu/VA), eccentricity (e), and roundness (ro).
Main Results:
- The algorithm achieved high accuracy (96.3 ± 0.9%) in segmenting microvessels.
- Significant differences in lu/VA, e, and ro were observed between SW1222 colorectal carcinomas and mouse fibrosarcomas (MFs).
- The algorithm provides detailed morphometric data for each segmented vessel and overall tumor vascularity (rVA).
Conclusions:
- The developed algorithm provides an accurate and automated method for microvessel analysis in histological tumor sections.
- Morphological differences in microvessels exist between different tumor types, which can be quantified by the algorithm.
- The algorithm, named CAIMAN, is publicly available online for researchers.
