Related Experiment Video
Updated: Aug 11, 2026
14:08
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
Published on: April 13, 2013
Liver segmentation for CT images using GVF snake
Fan Liu1, Binsheng Zhao, Peter K Kijewski
1Department of Medical Physics, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021, USA. liuf@mskcc.org
Medical Physics
|February 16, 2006
Summary
This study presents a novel semiautomatic method for liver segmentation in CT scans, improving accuracy for challenging cases. The developed algorithm achieves a median volume difference of 5.3% compared to radiologist delineations.
Area of Science:
- Medical Imaging
- Computer-Aided Diagnosis
- Radiology
Background:
- Accurate liver segmentation in computed tomography (CT) is crucial but challenging due to similar tissue densities and edge lesions.
- Existing methods struggle with precise delineation in complex anatomical regions.
Purpose of the Study:
- To develop and evaluate a semiautomatic method for accurate liver contour and volume segmentation on contrast-enhanced CT images.
- To enhance the performance of snake algorithms for liver segmentation using gradient vector flow (GVF).
Main Methods:
- A semiautomatic liver segmentation method using a snake algorithm with a gradient vector flow (GVF) external force.
- Edge map modification using Canny edge detector, liver template, and concavity removal to refine segmentation.
- Slice-by-slice segmentation with adjacent slice constraint for volumetric analysis.
Main Results:
- The method achieved a median volume difference ratio of 5.3% (range 2.9%-7.6%) compared to manual radiologist delineations.
- Successfully segmented 551 liver images from 20 volumetric datasets with colorectal metastases.
- Demonstrated improved performance in challenging segmentation scenarios.
Conclusions:
- The proposed semiautomatic liver segmentation method offers a promising solution for accurate delineation on CT images.
- The integration of GVF snake, modified edge maps, and slice-by-slice constraints enhances segmentation precision.
- This technique has potential applications in clinical settings for liver volume assessment.
