Keywords:
Breast, MR-Diffusion/Perfusion, Diagnostic procedure, Cancer
Authors:
H. Caylak1, N. Köse1, S. Bulut2; 1SİVAS/TR, 2Sivas/TR
DOI:
10.26044/ecr2019/C-0471
Methods and materials
Patient Selection:
Between 01.09.2013 and 15.02.2018,
79 female patients who underwent Dynamic Contrast Enhanced (DCE) and DWI Breast MRI before biopsy or surgical operation with positive diagnosis at MG,
US and/or clinical findings were included in the study and studied retrospectively. Non-mass lesions and lesions smaller than 10 mm were not included in the study. Also the patients whom treated with neoadjuvant chemotherapy and radiotherapy were not included in the study.The minimum age of the patients in the study was 21 years,
the maximum age was 76 years and the age values were 45.02 ± 11.02.
The median age is 44 years. 46 benign and 37 malign lesions,
a total of 83 lesions were included in the study.
Breast MRI Protocol:
Breast MRI examination was performed with 1.5 T MRI device (Simens Magnetom Aera 1.5 T,
Software: Syngo MR D13),
included 8-channel double superficial breast coil that cover the entire breast.
The patient was placed on the table in the prone position and both breast imaging were performed simultaneously.
MRI was performed between 7 - 14 days of menstrual cycle in premenopausal patients.The MRI protocol consisted of axial DCE T1-weighted,
axial and sagittal T2-weighted,
DW-images with b-values of 50,
400 and 800 s/mm2and pre-contrast ADC maps were automatically calculated from the b=50 and b=800 s/mm2.
Analysis of Breast MRI:
Breast MR images of the cases were evaluated using PACS (Pictures Archieving and Communication System) and Osirix Lite®v9.5. BI-RADS scores and ADC value of the lesions were determined. The longest dimensions of the lesions were measured in axial post-contrast T1-weighted sequences (Figure 1A and Figure 2A).
ADC values were determined from the section where the longest dimension of the lesion was measured or from the level closest to that slice.
Two different ROI methods (a ROI that covers whole lesion (W- ROI) and average of three small ROI (S-ROI) that placed the most aggressive subregion) were used to determine ADC of lesions and compared diagnostic performance(Figure 1C and Figure 2C). Hemorrhagic and / or necrotic areas within the lesion were especially avoided using pre-contrast T1A and T2A sequences.
W-ROIs were drawn manually.
The size of the S-ROIs was adjusted to between 3-5 mm2(4-6 pixels).To determine the most restiricted area in the lesion ,
we used colored ADC maps.( (Figure 1B and Figure 2B).ADC measurements were obtained from normal fibroglandular tissues (NFD) from the normal breast of each patient,
especially from the retroareolar area.
Figure 1:A) A 43-year-old female patient,
pathological diagnosis of fibroadenoma in the left breast showed on the fat-saturated DCE-T1 series.
B) Colored ADC map used for positioning of S-ROIs.
C) W-ROI and S-ROIs positioned on lesion and ADC values were determined.
Figure 2:A) A 66-year-old female patient,
pathological diagnosis of invasive ductal carsinoma (NST) in the left breast showed on the fat-saturated DCE-T1 series.
B) Colored ADC map used for positioning of S-ROIs.
C) W-ROI and S-ROIs positioned on lesion and ADC values were determined.
Statistical analysis:
For statistical analysis IBM SPSS Statistics 22 was used. Since parametric test assumptions are fulfilled in the evaluation of data (by using Kolmogorof-Simirnov); Differences between the mean ADC values of benign and malignant breast lesions were tested for significance by independent sample t test.
The area under the curve (AUC) was calculated using receiver operating characteristic (ROC) analysis for the mean ADC values of W-ROI and S-ROI.
P-value of less than 0.05 was considered to indicate statistically significant difference.