Related Experiment Video
Updated: Mar 2, 2026
05:23
Author Spotlight: Noninvasive Cerebral Blood Flow Determination in Human Functional Brain Region for Diagnosis of Neurological Disorders
Published on: May 31, 2024
911
SU-E-I-16: Automated Liver Segmentation Method for CBCT Dataset by Probabilistic Atlas Construction
D Li1,2,2,2, H Li1,2,2,2, Y Yin1,2,2,2
1College of Physics and Electronics, Shandong Normal University, Ji nan, Shandong Province.
Medical Physics
|May 19, 2017
Summary
This study presents an automated liver segmentation method for adaptive radiation therapy using low-contrast cone-beam CT (CBCT) images. The developed technique achieves accurate liver contouring, crucial for precise daily treatment adjustments.
Area of Science:
- Medical Imaging
- Radiation Oncology
- Image Segmentation
Background:
- Accurate liver contouring is essential for adaptive radiation therapy (ART).
- Daily cone-beam CT (CBCT) images present low-contrast challenges for accurate segmentation compared to planning CT.
- Automated methods are needed to improve efficiency and precision in ART.
Purpose of the Study:
- To develop an automated method for accurate liver contour segmentation in low-contrast daily CBCT images.
- To enable precise liver structure identification for adaptive radiation therapy adjustments.
Main Methods:
- A probabilistic atlas was constructed from 50 contrast-enhanced planning CT images using iterative affine registration.
- Deformable registration based on edge-preserving scale space was employed to map CBCT images to the atlas.
- Liver contours were generated using the deformation map, incorporating intensity distribution to refine segmentation and remove irrelevant tissues.
Main Results:
- The algorithm demonstrated effective liver segmentation on low-contrast CBCT images.
- Volumetric overlap with manual segmentations by oncologists ranged from 87% to 94% on 10 test cases.
- Oncologists confirmed the segmentation results were comparable to manual segmentation and suitable for adaptive radiation therapy.
Conclusions:
- The proposed automated segmentation method is highly effective for low-contrast CBCT images.
- This technique is suitable for daily use in adaptive radiation therapy, improving treatment accuracy.
