Automated segmentation of breast in 3 d mr images using a robust atlas
Download as DOCX, PDF0 likes132 views
This paper introduces a robust atlas-based segmentation algorithm for the breast boundary in 3D MR images, combining probabilistic atlas and atlas selection methods. The algorithm utilizes phase congruency maps to create a versatile atlas that works across different MR imaging sequences without intensity-based registration. Results show high accuracy, with median Dice similarity coefficient values of 94% for intra-sequence and 87% for inter-sequence segmentations.
Related topics:1 of 1
Download to read offline
More Related Content
Similar to Automated segmentation of breast in 3 d mr images using a robust atlas (20)
Ad
More from I3E Technologies (20)
Ad
Automated segmentation of breast in 3 d mr images using a robust atlas
- 1. AUTOMATED SEGMENTATION OF BREAST IN 3-D MR IMAGES USING A ROBUST ATLAS ABSTRACT This paper presents a robust atlas-based segmentation (ABS) algorithm for segmentation of the breast boundary in3-DMR images. The proposed algorithm combines the well- knownmethodologies of ABS namely probabilistic atlas and atlas selectionapproaches into a single framework where two configurationsare realized. The algorithm uses phase congruency maps to createan atlas which is robust to intensity variations. This allows anatlas derived from images acquired with one MR imaging sequenceto be used to segment images acquired with a different MR imaging sequence and eliminates the need for intensity-basedregistration. Images acquired using a Dixon sequence were usedto create an atlas which was used to segment both Dixon images(intra-sequence) and T1-weighted images (inter-sequence). Inboth cases, highly accurate results were achieved with the medianDice similarity coefficient values of 94% ± 4% and 87 ± 6.5%,respectively.