| Background and Purpose:1)Breast Imaging Reporting and Data System (MR BI-RADS) has been issued by the Association of American College of Radiology in2003. It was considered that the mass with type II or type III curves in the dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) as one of important suspected malignant standards in the in breast MRI, but a great quantity of studies have shown that DCE-MRI for the differential diagnosis in breast lesions was high sensitivity and low specificity, in the literatures, they were reported, the specificity was in the range of37%-97%.Breast diffusion-weighted imaging (DWI) is a random Brownian motion which is a functional imaging reflects of water molecules diffusion in the tissue and lesions, it is a possible method to improve the specificity of DCE-MR. But in the current breast DWI studies were mainly focused on the mono-exponential DWI model, calculated the ADC values and determined the threshold between the benign and malignant lesions, thereby embedding an assumption of diffusion are present of the parity between the intracellular space and the extracellular one, they have been documented they were certain overlap in the differential diagnosis of breast lesions. And in the bi-exponential and multi-exponential DWI models, it can be observed ADC values have an inverse relationship with the b-values, but positively associated with the effect of perfusion and T2shine through effect. Furthermore, they have pointed out diffusion was disparity between the intracellular space and the extracellular one. Therefore there are some drawbacks of the mono-exponential ADC values in the differential diagnosis. Based on the above theories, we have performed the preliminary study of differential diagnosis to draw the attenuation rate of ADCb=800/0mm2/s and ADC b=1500/0mm2/s (ARADC) in breast mass with type Ⅱ or type Ⅲ curves.2) Technically, the currently applied DWI protocol with echo planar imaging (EPI) sequence is remained subject to some pitfalls in the breast examination, which are primarily comprised of patient motion and/or eddy current-induced distortions, susceptibility artifacts and non-sufficient chemical shift saturation. This phenomenon is present both in the1.5Tesla and3.0T MR scanners, especially the latter. We presume that might reduce the reliability of results in the processing of the ADCs calculation. Susceptibility artifacts from main magnetic field in-homogeneity are occurred generally in the areas of pre-pectoral and the margin of the breast. In fact, the poor-resolution of breast DWI had the partial volume effects in the imaging plane, and the non-sufficient chemical shift saturation may intensify this effects. Eventually the measured ADC values of lesions might be over-or underestimated, particularly in evaluating the tiny lesions less than1.0cm.We modify the applied protocols to minimize the artifact’s contributions.Method:1) In this study,41women cases were retrospectively reviewed, of which pathologically confirmed (age22-79years, mean age45.61±7.89years), the masses in TIC with type Ⅱ or Ⅲ curve were included. 2) All patients were examined with a3.0T MR scanner (GE HDxt American) using a eight-channel breast commercial coil, firstly conventional protocols were performed:2D axial STIR T2WI,2D axial T1WI, the2D oblique FS T2WI. And then DWI was preformed.①The axial unilateral DWI of b-values800/0mm2/s was performed using a echo-planar imaging (EPI) sequence which the center of FOV was located on the center between the breast. The axial unilateral DWI acquisition parameters were as the following:TR/TE,5000ms/67ms, FOV,28-32cm2, acquisition matrix,128*128mm2, NEX,6.0,slice thickness of4.0-5.Omm,slice gap,1.0mm,duration time,2:05-3:20min.②if DWI acquisition artifacts of b-values800,0mm2/s were visually occurred, and the DWI acquisition of b-values1500/0mm2/s, the modified protocol were employed. The following centers were approximately located in the same focus, the unilateral FOV center, the unilateral shimming block and the unilateral breast center. The parameters were as follow:FOV,26cm2, shimming block,12-16cm2, TR/TE:5000ms/67ms, matrix,128mm*128mm, NEX,6.0, slice thickness,3.0mm, slice gap,1.0mm; duration time,2:05-3:20min.③the dynamic enhancement contrast was using3D axial LAVA fat-suppressed sequence in the axial plane. TR/TE,7.4ms/1.5ms, Flip anger,14°, Matix,416mm*320mm,FOV,28-32cm2, slice thickness:1.4mm, duration time:10:04min. Dosage of Gd-DTPA0.2mmol/kg, injection rate of3ml/sec, and then flushed with saline10ml, injection rate of2ml/s. One-way ANOVA were used to compare the of ADCs two groups of b-value, and ARADC between the benign and malignant lesions. Further the ROC curve analysis to determine the threshold, consequently the chi-square test was to determine the sensitivity and specificity of the differential diagnosis.3) In the pre-experiment,97patients were performed of two groups of b value (b=800/0mm2/s and b=1500/0mm2/s), when artifacts that acquisition with bilateral DWI protocol were not occurred, we introduced the modified protocol of reduced of FOVs and active shimming blocks, the paired t-test were compared ADC and SNR values Results:1) The lesions with type II and type III in the TIC,20cases are benign lesions, including:fibroadenomas (12,29.27%of), intraductal papillomas (3,7.32%), abscesses (3,7.32%) and benign phyllodes (2,4.88%).21cases are malignant lesions, including invasive ductal carcinoma (16,39.02%), intraductal papillary carcinomas (3,7.32%) and malignant phyllodes (2,4.88%). This group of cases with type II curve are19cases,8cases (46.34%) are benign lesions (19.51%) of ADC values b=800/Omm2/s and b=1500/0mm2/s are1.29±0.09mm2/s,1.02±0.08mm2/s respectively. and11cases (26.83%) cases are malignant lesions of ADC values b=800/0mm2/s and b=1500/0mm2/s are1.12±0.13mm2/s,0.99±0.14mm2/s respectively. This group of cases with type II curve is22cases (53.66%),12cases (29.27%) are benign lesions of ADC values b=800/0and b=1500/0are1.19±0.15mm2/s,1.07±0.13mm2/s,10cases (24.39%) are malignant lesions of ADC values b=800/0mm2/s and b=1500/0mm2/s were1.15±0.12mm2/s,1.01±0.14mm2/s respectively. Analysis of variance among the four groups, p>0.05, there is no significant statistical difference. Histological type:in the group of21cases are malignant of ADC values b=800/0mm2/s and b=1500/0mm2/s are1.13±0.33mm2/s,1.03±0.12mm2/s.20cases are benign lesions of ADC values b=800/0mm2/s and b=1500/0mm2/s are1.29±0.41mm2/s,1.00±0.23mm2/s, there is no significant statistical difference. Malignant group of ADC b=800/0mm2/s and ADC b=1500/0mm2/s attenuation rate is0.14±0.04, the benign group of ADC b=800/Omm2/s and ADC b=1500/0mm2/s attenuation rate is0.23±0.06, p<0.05. Take the threshold of0.19, differential diagnosis of the benign and malignant groups, sensitivity is90.5%, specificity is95.0%, the accuracy is92.68%.2) In the pre-experiment of the97cases, DWI acquisition with the modified protocol is measured ADC values and SNR values. Paired t-test, p-value>0.05, and it is a statistically significant difference. In this study, the bilateral collection CNR b= 800/0mm2/s is4.97±0.06; unilateral acquisition CNRb=1500/0mm2/s is5.67±0.98, paired t-test shows p value=0.014. Unilateral acquisition of DWI effectively eliminates the artifacts.Conclusions:1) The ARADC is benefited in the differential diagnosis of type II or III curve, adjust the BI-RADS classification of some lesions to guide clinical treatment.2) Unlateral acquisition of DWI protocol effectively eliminates the artifacts... |
PDF Full Text Request