E13. How to set up an MRI service

Abstract

Breast magnetic resonance imagine (MRI) has become one of the mainstays in breast radiology, and as with every tool that is incorporated in the radiologist’s equipment the physician has to learn the intricacies of the technique in order to make the most of it for his patients. These aspects include: the technical requirements, the diagnostic criteria, and the clinical management of the imaging findings. Technical requirements and diagnostic criteria [1−4] Magnet hardware: The field strength of the magnet should not be lower than 1.5 Tesla, and the stronger the gradient (>20 mT/m) the better. The breast coil should be dedicated for breast imaging and, if possible, should be prepared for vacuum-assisted procedures. Multi-channel coils and parallel imaging offer the best signal-to-noise ratio. Patient preparation and position: Premenopausal patients should be scanned on days 6−13 of the menstrual cycle, even when oral contraception is being used. In the case of hormone replacement therapy, MRI should be performed at least 4 weeks after discontinuation of treatment. In urgent cases (i.e. breast cancer staging) this schedule protocol can be waived. The patient should be positioned prone with her arms placed next to her body, in order to increase the anatomical coverage of the breast coil. The usual field of view (FOV) should be 280–320 mm with a matrix of 512×256 or 512times;512 mm, depending on the native imaging plane. The images should cover from the supraclavicular region to the inframammary fold and, of course, both breasts. Imaging plane: Any of the three usual planes (sagittal, axial or coronal) is correct, as long as the spatial and temporal resolution requirements are met (see below). It is also important that the phase-encoding direction is correctly oriented in order to minimize the movement artifacts. In the sagittal plane, the phase-encoding direction should be from head to feet, in the axial plane from left to right, and in the coronal plane from head to feet. Temporal and spatial resolution: The enhancement peak in malignant lesions is maximum usually 1−3 minutes after contrast injection, and therefore temporal resolution should be less than 120 seconds. Continuous acquisition of 60−90-s temporal sequences along a total acquisition time of 6−9 minutes is enough to configure the enhancement curve and show uptake and wash-out features. These features will give us the kinetic aspects of the MR image. Spatial resolution is the second requisite, as it will give us the morphology of the lesions (margins and internal architecture), but it should be taken into account that, when spatial resolution increases, the price to pay is time. The pixel size should be no greater than 0.5−0.8 mm in the x and y planes (in-plane resolution) and the slice thickness should be less than 4 mm, ideally 1−3 mm. Basic imaging sequences: Pre-contrast images should be T2-weighted and diffusion images. T2-weighted images (T2WI) are useful to delineate the fluid content of lesions, ducts, and also axillary regions. Repetition time (TR) and echo time (TE) should be greater than 2 s and 80 ms, respectively, and the optimal sequence is the fast-spin echo sequence (FSE). Short T1 inversion recovery (STIR) images are an alternative to T2 FSE when the fat signal is to be suppressed, and the advantage over T2 FSE with fat saturation is that the field homogeneity requirements are not as strict. Diffusion sequences should have a diffusion factor of b = 0 and 700 s/mm 2 . The slice thickness in these sequences should ideally be lower than 3 mm in order to reach a good correlation with T1-weighted dynamic images.