Related Experiment Video
Updated: Oct 25, 2025
09:06
Quantitative Analysis of Cellular Composition in Advanced Atherosclerotic Lesions of Smooth Muscle Cell Lineage-Tracing Mice
Published on: February 20, 2019
8.5K
Vesseg: An Open-Source Tool for Deep Learning-Based Atherosclerotic Plaque Quantification in Histopathology
Jacob M Murray1,2,3, Phillip Pfeffer4, Robert Seifert5,6,7
1Department of Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany (J.M.M., H.-P.S., J.K.).
Arteriosclerosis, Thrombosis, and Vascular Biology
|August 12, 2021
Summary
VesseG, a new deep learning tool, automates atherosclerotic plaque segmentation in microscopy images, significantly reducing research time and improving consistency for atherosclerosis studies.
Area of Science:
- Cardiovascular Research
- Medical Imaging Analysis
- Computational Pathology
Background:
- Manual plaque segmentation in atherosclerosis research is labor-intensive and prone to observer variability.
- Accurate segmentation is crucial for quantifying disease progression and evaluating treatments.
Purpose of the Study:
- To introduce VesseG, an open-source tool utilizing deep learning for automated atherosclerotic plaque segmentation.
- To provide a user-friendly and efficient solution for analyzing microscopy images in atherosclerosis research.
Main Methods:
- Developed VesseG, a containerized tool featuring two deep learning models.
- Trained and validated models on 1089 hematoxylin-eosin stained mouse brachiocephalic artery sections.
- Compared model performance against three human expert raters.
Main Results:
- Achieved high segmentation accuracy with mean Soerensen-Dice scores of 0.91 for plaque and 0.97 for lumen.
- Demonstrated excellent overall mean accuracy of 0.98.
- VesseG provides time savings of over 10 minutes per slide and is in active use.
Conclusions:
- VesseG integrates state-of-the-art deep learning into atherosclerosis research.
- The tool offers substantial time savings and enhances consistency in plaque segmentation.
- VesseG supports continuous model improvement and pipeline development.
