• Nicky H G M Peters, Koen L Vincken, Maurice A A J van den Bosch, Peter R Luijten, Willem P Th M Mali, and Lambertus W Bartels.
• Department of Radiology, University Medical Center Utrecht, Heidelberglaan 100, E01.132 3584 CX Utrecht, the Netherlands. n.peters@umcutrecht.nl
• J Magn Reson Imaging. 2010 May 1; 31 (5): 1100-5.

PurposeTo evaluate the influence of the choice of different combinations of b-values on the ADC and on the diagnostic performance of quantitative diffusion weighted imaging (DWI) in breast lesions.Materials And MethodsSeventy-three patients (90 lesions) underwent 3 Tesla (T) breast MRI including a DWI-scan using b-values 0, 150, 499, and 1500 s/mm(2) and histological analysis. Five combinations of b-values were used to calculate the ADC, each with different sensitivities to perfusion and diffusion effects. The median ADC of benign lesions, noninvasive carcinomas and invasive carcinomas and the diagnostic performance of the five methods were compared.ResultsEighty-eight lesions were analyzed (37 benign, 13 noninvasive carcinomas, 38 invasive carcinomas). The median ADC was highest in benign lesions, intermediate in noninvasive carcinomas and lowest in invasive carcinomas for all methods. Calculating the ADC with the lowest 2 b-values yielded the highest ADC for all lesions types; the highest 2 b-values yielded the lowest ADC. The area under the receiver operating characteristic curve was approximately equal for all methods.ConclusionThe ADC of breast lesions varied substantially with the choice of different b-values, indicating that absolute ADC threshold values to differentiate benign and malignant lesions should be interpreted with caution. However, the diagnostic performance of quantitative DWI was not affected by the choice of different b-values.Copyright 2010 Wiley-Liss, Inc.

24 keywords...

hide…