Related Experiment Video
Updated: Jun 6, 2026
10:25
Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
Published on: September 25, 2019
A Clustering Algorithm for Liver Lesion Segmentation of Diffusion-Weighted MR Images
Abhinav K Jha1, Jeffrey J Rodríguez, Renu M Stephen
1College of Optical Sciences, University of Arizona, Tucson, AZ, USA.
Proceedings. IEEE Southwest Symposium on Image Analysis and Interpretation
|December 15, 2010
Summary
Accurate liver lesion segmentation in diffusion-weighted MRI is crucial for therapy response assessment. A new clustering algorithm incorporating spatial information and geometric constraints improves segmentation accuracy compared to existing methods.
Area of Science:
- Medical Imaging
- Radiology
- Biomedical Engineering
Background:
- Accurate segmentation of liver lesions in diffusion-weighted magnetic resonance imaging (DW-MRI) is essential for calculating the apparent diffusion coefficient (ADC).
- The ADC parameter indicates lesion response to therapy.
- Segmentation is challenging due to low signal-to-noise ratio (SNR), fuzzy boundaries, and artifacts like speckle and motion.
Purpose of the Study:
- To develop an improved algorithm for segmenting liver lesions in DW-MRI.
- To address the challenges of low SNR, fuzzy boundaries, and artifacts in liver lesion segmentation.
- To enhance the accuracy of ADC calculation for therapy response assessment.
Main Methods:
- A novel clustering algorithm was developed.
- The algorithm incorporates spatial information into the segmentation process.
- A geometric constraint was integrated to improve segmentation accuracy.
Main Results:
- The proposed algorithm demonstrated improved accuracy in segmenting liver lesions.
- The enhanced segmentation facilitates more reliable ADC computation.
- Performance was superior compared to existing segmentation algorithms.
Conclusions:
- The developed clustering algorithm effectively overcomes common challenges in DW-MRI liver lesion segmentation.
- This method offers a more accurate approach for calculating ADC, aiding in therapy response evaluation.
- The algorithm shows promise for clinical application in oncology and radiology.
