Brock L, Liebert A, Schreiter H, Skwierawska D, Ehring C, Eberle J, Laun FB, Uder M, Kapsner L, Ohlmeyer S, Hadler D, Knoll F, Bickelhaupt S (2024)
Publication Type: Conference contribution
Publication year: 2024
Publisher: Springer Science and Business Media Deutschland GmbH
Book Volume: 15249 LNCS
Pages Range: 127-136
Conference Proceedings Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Event location: Marrakesh, MAR
ISBN: 9783031734793
DOI: 10.1007/978-3-031-73480-9_10
Breast magnetic resonance imaging (MRI) increasingly incorporates diffusion-weighted imaging (DWI) sequences. DWI allows to calculate apparent diffusion coefficient (ADC) maps which can supplement differentiation between malignant and benign breast lesions. However, artifacts in DWI are not infrequent, e.g., due to patient motion, pulsation or other sources, which can cause shifts between the different b-value acquisitions and affect the accuracy of the ADC map. This IRB-approved, retrospective study includes n = 141 women with breast lesions undergoing breast MRI including an ultra-high b-value DWI acquisition (b = 1500s/mm2). Thirteen different rigid and non-rigid co-registration methods facilitated by the ANTs library were evaluated. The targets for co-registration were different b-value DWI. ADC maps were calculated from the co-registered DWI images and analyzed quantitatively, measuring mean, standard deviation, and within-lesion coefficient of variance (CoV), using repeated-measures ANOVA alongside Dunnett post-hoc tests in manual segmentations of the target lesions performed on ultra-high b-value images. In addition, we evaluated between-patient CoV and area under the curve (AUC) for differentiation of benign and malignant lesions, as well as sensitivity, specificity, and accuracy to measure clinical efficacy. Amongst the n = 13 co-registration techniques, the iterative SyNCC co-registration method with ultra-high b-value DWI images as target demonstrated the most notable improvement in standard deviation and within-lesion CoV while maintaining consistent mean ADC values. However, none of the co-registration method significantly improved the differentiation of lesion types. Nonetheless, the study offers a detailed comparison of co-registration methods for DWI images in ADC map generation.
APA:
Brock, L., Liebert, A., Schreiter, H., Skwierawska, D., Ehring, C., Eberle, J.,... Bickelhaupt, S. (2024). Comparative Study on Co-registration Techniques for Diffusion-Weighted Breast MRI and Improved ADC Mapping. In Marc Modat, Žiga Špiclin, Alessa Hering, Ivor Simpson, Wietske Bastiaansen, Tony C. W. Mok (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 127-136). Marrakesh, MAR: Springer Science and Business Media Deutschland GmbH.
MLA:
Brock, Luise, et al. "Comparative Study on Co-registration Techniques for Diffusion-Weighted Breast MRI and Improved ADC Mapping." Proceedings of the 11th International Workshop on Biomedical Image Registration, WBIR 2024, held in conjunction with the 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024, Marrakesh, MAR Ed. Marc Modat, Žiga Špiclin, Alessa Hering, Ivor Simpson, Wietske Bastiaansen, Tony C. W. Mok, Springer Science and Business Media Deutschland GmbH, 2024. 127-136.
BibTeX: Download