Breast MRI in the era of diffusion weighted imaging: do we still need signal-intensity time curves?

Dietzel M, Ellmann S, Schulz-Wendtland R, Clauser P, Wenkel E, Uder M, Baltzer PAT (2020)


Publication Type: Journal article

Publication year: 2020

Journal

Book Volume: 30

Pages Range: 47-56

Journal Issue: 1

DOI: 10.1007/s00330-019-06346-x

Abstract

OBJECTIVE: Dynamic contrast-enhanced imaging of the initial (IP) and delayed phase (DP) is an integral part of any clinical breast MRI protocol. Furthermore, DWI is increasingly used as an add-on sequence by the breast-imaging community. We investigated whether DWI could be used as a substitute DP. MATERIAL AND METHODS: One hundred thirty-two consecutive patients with equivocal or suspicious findings at ultrasound and/or mammography received a full diagnostic breast MRI according to international recommendations. Histopathological verification served as reference standard. We evaluated three sections of the MRI protocol: IP, DP, and apparent diffusion coefficient (ADC) maps derived from DWI. Circular ROIs (regions of interest, mean size 5-10 mm2) were drawn into the enhancing parts of the lesion (first postcontrast). ROIs were transferred to the corresponding location on ADC maps and IP and DP images. Mean ROI values were investigated signal intensity (SI): (1) Initial-phase enhancement = (SI(IP) - SI(precontrast))/SI(precontrast); (2) Delayed-phase enhancement = (SI(DP) - SI(IP))/SI(IP); (3) ADC. Multiparametric combinations were computed using logistic regression analysis: (1) IP+: Initial-phase enhancement and ADC; (2) Curve: Initial-phase enhancement and delayed-phase enhancement; (3) Curve+: Curve and ADC. The diagnostic performances of these feature combinations to diagnose malignancy were compared by the area under the receiver-operating characteristics curve (AUC). RESULTS: One hundred thirty-two patients (age: mean = 57.1 years, range 23-83 years) with 145 lesions were included (malignant/benign 101/44). IP+ (AUC = 0.877) outperformed Curve (AUC = 0.788, p = 0.03). Curve+ was not superior to IP+ (p = 1). CONCLUSION: DWI could substitute DP. Because DWI is typically used as an add-on to IP and DP, our results might help to abbreviate and to simplify current practice of breast MRI. KEY POINTS: • DWI provides similar but superior diagnostic information for diagnosis of malignancy in enhancing breast lesions compared to DP. • Adding DP to DWI does not provide incremental information to distinguish benign from malignant lesions. • DWI could substitute DP. As DWI is typically used as an add-on to IP and DP, our findings might help to abbreviate and to simplify current breast MRI practice.

Authors with CRIS profile

Involved external institutions

How to cite

APA:

Dietzel, M., Ellmann, S., Schulz-Wendtland, R., Clauser, P., Wenkel, E., Uder, M., & Baltzer, P.A.T. (2020). Breast MRI in the era of diffusion weighted imaging: do we still need signal-intensity time curves? European Radiology, 30(1), 47-56. https://dx.doi.org/10.1007/s00330-019-06346-x

MLA:

Dietzel, Matthias, et al. "Breast MRI in the era of diffusion weighted imaging: do we still need signal-intensity time curves?" European Radiology 30.1 (2020): 47-56.

BibTeX: Download