MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

Evaluation of An Automated Deformable Mapping Technique with and Without External Fiducial Markers to Relate Corresponding Lesions in 3D X-Ray and Ultrasound Breast Images

C Green1*, M Goodsitt2 , J Lau3 , K Brock4 , C Davis5 , P Carson6 , (1) university Michigan, Ann Arbor, MI, (2) university Michigan, Ann Arbor, MI, (3) University of Michigan, Ann Arbor, Michigan, (4) UT MD Anderson Cancer Center, Houston, TX, (5) General Electric CRD, Niskayuna, NY, (6) The University of Michigan Health System, Ann Arbor, MI

Presentations

(Tuesday, 7/31/2018) 7:30 AM - 9:30 AM

Room: Room 202

Purpose: To compare the performance of an automated method for registering corresponding lesions in dedicated breast CT (bCT), digital breast tomosynthesis (DBT), and automated breast ultrasound (ABUS) images with and without the use of external fiducial markers.

Methods: A CIRS multi-modality breast phantom containing 10 lesions was employed and imaged with conventional CT to simulate coronal slice acquisition for bCT (without compression), DBT (upright positioning with cranial-caudal compression), and ABUS (supine positioning with anterior-to-chest wall compression). Six external fiducial markers (gel pads containing 1-mm glass beads) were attached to the breast phantom. The reconstructed images were segmented manually. An automated algorithm built, deformed, and related the resulting biomechanical models of the breast for the registration of lesions between modalities. Performance was assessed by the number of matched lesions and measures of distances between centers of mass (dCOM) of matched lesions.

Results: The maximum number of lesions that could be matched was 7 because 3 of the 10 lesions were not visible in ABUS images. For mapping bCT to ABUS without markers, 5 of the 7 lesions were matched and the mean dCOM was 10.6 ± 3.0 mm. With markers, 6 of the 7 lesions were matched and the mean dCOM was 6.4 ± 3.1 mm. For mapping DBT to ABUS without markers, 5 of the 7 lesions were matched and the mean dCOM was 6.5 ± 2.0 mm. With markers, 6 of the 7 lesions were matched and the mean dCOM was 5.7 ± 1.0 mm.

Conclusion: This work demonstrates the potential for using this deformable mapping technique to identify related lesions within various breast imaging modalities and shows improved lesion correlation with the use of external fiducial markers. Future work will include an IRB-approved proof of concept study with patient data for registration between DBT and ABUS images.

Funding Support, Disclosures, and Conflict of Interest: This work was supported in-part by a research grant (15-PAF04328) from GE Global Research. Crystal A. Green is supported by the Science Mathematics, and Research for Transformation (SMART) Scholarship for Service Program (HQ0034-16-C-0008). MMG is a collaborator on a grant funded by GE.

Keywords

Not Applicable / None Entered.

Taxonomy

Not Applicable / None Entered.

Contact Email