Archives of Hellenic Medicine

Last update:

26-Oct-2010

Arch Hellen Med, 27(5), September-October 2010, 811-817

ORIGINAL PAPER

Pitfalls in the production of enhancement kinetic curves
in 3T dynamic magnetic resonance mammography and means of avoidance

E. LAVDAS,¹ P. MAVROIDIS,² V. ROKA,¹ D.L. ARVANITIS,³ I.V. FEZOULIDIS,¹ K. VASIOU¹
¹Department of Radiology-Medical Imaging, University Hospital of Larissa, Medical School, University of Thessaly, Larissa, Greece,
²Department of Medical Radiation Physics, Karolinska Institutet and University of Stockholm, Stockholm, Sweden
³Department of Anatomy, University Hospital of Larissa, Medical School, University of Thessaly, Larissa, Greece

OBJECTIVE The purpose of this study was to investigate means for the elimination or reduction of errors that appear in enhancement kinetic curves in 3Telsa (3T) dynamic magnetic resonance (MRI) mammography from breathing or kinetic artifacts, in order to avoid pitfalls in the accurate determination of the curve type.

METHOD A magnetic resonance unit 3-T (Signa, GE Healthcare) was used. Patients were imaged in the prone position with an eight-channel body phased array surface coil. The following sequences were used for MRI of both breasts: (a) axial Τ2-TSE, (b) axial STIR and (c) Vibrant (six phases) unenhanced and contrast-enhanced axial T1-weighted fat saturation with dynamic acquisition after intravenous injection of a bolus of 0.1 mmoL/kg of gadodiamide. During the assessment of the kinetic curves, it was observed that there was no consistency in the location of the selected region of interest (ROI) in all phases of dynamic acquisition. The ROIs were positioned manually, measuring distances from stable reference points (skin, sternum) in all the phases and the enhancement kinetic curves were plotted based on the signal intensity.

RESULTS In the 10 patients examined, 50 abnormalities were investigated for each of which an enhancement kinetic curve was derived (from the images of the multi-phasic sequence, Vibrant), both semi-automatically (using the calculation software of the system) and manually. In 31 (62%) of the cases of abnormality, a change of the enhancement kinetic curve type was found for both methods. In these cases, it was observed that with the semi-automated method 6 enhancement kinetic curves were of type III, whereas using the manual calculation they were of type I; with the semi-automated method 9 were of type III and using the manual calculation of type II, and with the semi-automated method 10 curves were of type II and using the manual calculation of type I. Of the cases for which no change of the enhancement kinetic curve type was found, 8 cases were curves of type I and 11 cases were curves of type II.

CONCLUSIONS During the derivation of enhancement kinetic curves, it should be ensured that the ROI is at the same location of the lesion in all the phases of the dynamic study (enhancement). More attention should be given in the case of lesions that change size depending on the phase of enhancement. In these cases stable reference points should be noted (for example, the distance between the lesion and skin or sternum). In cases where there is no consistency of the ROI location in the different phases of the multi-phasic sequence, it should be placed manually using stable reference points and subsequently (based on the signal intensity values) the enhancement kinetic curves should be derived.

Key words: Dynamic magnetic resonance mammography, Enhancement kinetic curves, 3 Tesla.