This poster is published under an
open license. Please read the
disclaimer for further details.
Keywords:
Neoplasia, Imaging sequences, Diagnostic procedure, MR, Breast
Authors:
R. Rosa, A. M. Sarmento, M. D. L. G. F. R. Orvalho, H. Ferreira; Lisboa/PT
DOI:
10.1594/ecr2014/C-2101
Methods and materials
Sixteen women with clinical indication for MRI breast examination (suspicious lesion/staging) were enrolled in the study.
Exclusion criteria included: prior chemo-,
radio- and/or hormone therapy; prior breast surgery and mammary implants.
Furthermore,
MRI examinations were done between the 7th and the 14th day of the menstrual cycle in the case of pre-menopausal women,
and not before than 8 days after needle biopsy if applicable.
MRI examinations were done in a 1.5T scanner (Magnetom Avanto,
Siemens) using a 14-channel breast coil.
The conventional breast MRI protocol was followed,
including an axial T1 3D GRE (FLASH) sequence with SPAIR fatsat (TR/TE=4.0/1.69ms) for DCE-MRI.
Additionally,
a T1 3D FLASH sequence modified with dual-echo (TR/TE1/TE2=7.4/2.38/4.76ms) was also run prior to contrast injection and after conventional DCE-MRI.
The gadolinium contrast injection protocol was comprised of: 0.1 mmol/L/Kg Gadopentetate Dimeglumine (Magnevist,
Bayer) + 30 mL saline bolus at an injection rate of 2.5 mL/s. Five FLASH SPAIR image volumes were acquired post-contrast followed by the FLASH Dixon sequence.
Dixon fat and water images were then computed using the scanner's console.
Water and SPAIR images were evaluated qualitatively regarding fatsat homogeneity,
lesion conspicuity and axillary region visualization.
Quantitatively,
water and SPAIR images were evaluated regarding signal-to-noise ratio (SNR),
contrast-to-noise ratio (CNR),
and fatsat uniformity of lesion,
gland and adipose tissues in pre and post-contrast images.
For that purpose three regions-of-interest (ROIs) were placed in each of the different tissues and also in air for assesment of noise (Osirix,
Pixmeo).
ROIs’ signals mean values were then calculated and SNR,
CNR and fatsat uniformity computed (Excel,
Microsoft).
SNR was computed from mean tissue signals over the standard-deviation (SD) signal of air.
CNR was computed from the difference between the SNRs for the different tissues.
Fatsat uniformity was computed from the ROIs’ mean SD of the fat tissue signals.
The different fatsat techniques,
Dixon and SPAIR,
were then compared using paired t-test or Wilcoxon statistics.
The normality of the different variables was tested using the Shapiro-Wilk test.
The significance level used was p<0.05 (SPSS,
IBM).