Related Experiment Video
Updated: Nov 7, 2025
Multi-modal Pulmonary Imaging: Using Complementary Information from CT and Hyperpolarized 129Xe MRI to Evaluate Lung Structure-Function
Published on: April 12, 2024
Semantic Cardiac Segmentation in Chest CT Images Using K-Means Clustering and the Mathematical Morphology Method
Beanbonyka Rim1, Sungjin Lee1, Ahyoung Lee2
1Department of Software Convergence, Soonchunhyang University, Asan 31538, Korea.
Insights
This study introduces an unsupervised method for whole-heart segmentation in CT scans, combining K-Means clustering and mathematical morphology. The approach aims to improve efficiency and accuracy in detecting cardiovascular diseases like coronary artery disease (CAD).
Area of Science:
- Medical Imaging
- Cardiovascular Imaging
- Artificial Intelligence in Medicine
Background:
- Whole cardiac segmentation in chest CT images is crucial for identifying cardiovascular diseases (CVDs) and related conditions like coronary artery disease (CAD).
- Manual segmentation is labor-intensive, time-consuming, and requires specialized radiologist expertise for accurate ground truth labeling.
- The scarcity of annotated data and the need for expert annotators present significant challenges for developing automated segmentation methods.
Purpose of the Study:
- To develop an unsupervised semantic whole-heart segmentation technique for chest CT images.
- To overcome the limitations of manual annotation and data scarcity in cardiac segmentation.
- To evaluate the proposed method's performance in segmenting cardiac structures from CT data.
Main Methods:
- A novel unsupervised semantic whole-heart segmentation approach was developed.
- The method integrates K-Means clustering as a threshold criterion within a mean-thresholding framework.
- Mathematical morphology operations were employed to enhance threshold shifting and refine segmentation results.
Main Results:
- Experiments were conducted on 500 subjects using two data subsets: 56 slices (full heart) and 30 slices (upper heart).
- The K-Means method yielded an average silhouette score of 0.4130 across both cases.
- Segmentation of 56 slices achieved an overall accuracy (OA) of 34.90% and mean intersection over union (mIoU) of 41.26%.
- Segmentation of the first 30 slices demonstrated higher performance with OA of 55.10% and mIoU of 71.46%.
Conclusions:
- The proposed unsupervised method offers a viable alternative to manual cardiac segmentation in CT images.
- The technique shows potential for aiding in the detection of cardiovascular abnormalities, particularly in datasets with limited annotations.
- Performance varied between full and partial heart scans, suggesting further refinement may be needed for comprehensive whole-heart analysis.
Abstract:
Whole cardiac segmentation in chest CT images is important to identify functional abnormalities that occur in cardiovascular diseases, such as coronary artery disease (CAD) detection. However, manual efforts are time-consuming and labor intensive. Additionally, labeling the ground truth for cardiac segmentation requires the extensive manual annotation of images by the radiologist. Due to the difficulty in obtaining the annotated data and the required expertise as an annotator, an unsupervised approach is proposed. In this paper, we introduce a semantic whole-heart segmentation combining K-Means clustering as a threshold criterion of the mean-thresholding method and mathematical morphology method as a threshold shifting enhancer. The experiment was conducted on 500 subjects in two cases: (1) 56 slices per volume containing full heart scans, and (2) 30 slices per volume containing about half of the top of heart scans before the liver appears. In both cases, the results showed an average silhouette score of the K-Means method of 0.4130. Additionally, the experiment on 56 slices per volume achieved an overall accuracy (OA) and mean intersection over union (mIoU) of 34.90% and 41.26%, respectively, while the performance for the first 30 slices per volume achieved an OA and mIoU of 55.10% and 71.46%, respectively.
