Abstract ID: 365 MR spectroscopy and multiparametric imaging of the breast

Purpose To focus on the current status of MR spectroscopy (MRS) and multiparametric MR imaging (mMRI) of breast lesions, from the standpoint of the medical physicist in a multidisciplinary clinical setting. Methods A meta-analysis of the available literature on the subject has been performed. Studies at 1.5 T as well as at 3.0 T were considered. Higher magnetic fields were also included for the sake of completeness. Results MRS of breast lesions has not yet entered the field of routine examinations due to technical difficulties and confounding factors that limit its sensitivity. Major limitations include patient motion, strong lipid signals in MR spectra, and suboptimal performances of dedicated breast coils. In a recent meta-analysis [1], pooled sensitivity and specificity were reported as 73% and 88%, respectively. On the other hand, mMRI shows increased sensitivity and specificity compared to single MRI modalities. A study on mMRI at 7 T has recently reported a sensitivity of 100% and a specificity of 88.2% when dynamic studies and DWI were simultaneously taken into account [2]. In another study, MRS was shown to improve the performance of a machine-learning classifier when added to mMRI based on morphology, dynamic imaging and DWI [3]. Conclusions MR spectroscopy of breast lesions might finally find its role in the clinical practice. Multiparametric MRI has shown promising results in the characterization of suspicious breast lesions and in the correlation to molecular subtypes of breast tumors. The role of the medical physicist in a multidisciplinary clinical setting is increasingly recognized as a key factor in order to guarantee the quality of the information provided by quantitative MRI imaging techniques.