Quantitative analysis of mean apparent propagator-magnetic resonance imaging for distinguishing glioblastoma from solitary brain metastasis

Purpose: Distinguishing glioblastoma (GBM) and solitary brain metastasis (SBM) is vital for determining the optimal treatment. GBM and SBM present similar imaging characteristics on conventional magnetic resonance imaging (MRI). The aim of this study was to evaluate the efficacy of quantitative analysis of mean apparent propagator (MAP)-MRI for distinguishing GBM and SBM.

Method: Eighty-nine patients were enrolled. Regions of interest (ROIs), including the enhancing area (EA), peritumoural high signal intensity area (PHA), and maximum abnormal signal area (MASA), were manually delineated. The following MAP parameters for each region were measured: mean square displacement (MSD), non-Gaussianity (NG), NG axial (NGAx), NG vertical, Q-space inverse variance, return to origin probability (RTOP), return to axis probability (RTAP), and return to plane probability (RTPP). Normalised MAP values from each region were compared between the GBM and SBM groups, and their diagnostic efficiency was assessed. Multivariate logistic regression analysis was used to create the most accurate model.

Results: Compared with the SBM group, the MSD was significantly lower in the GBM group, whereas the RTAP, RTOP, and RTPP were significantly higher in each region, except for RTAP_PHA (all P < 0.05). RTPP_PHA, MSD_EA, and RTPP_MASA showed the most significant differences (all P < 0.01). The best logistic regression model combined RTPP_PHA, MSD_EA, and NGAx_MASA (area under the curve, 0.840).

Conclusions: Quantitative analysis of MAP-MRI is useful for distinguishing GBM from SBM. Multivariate analysis combined with multiple ROIs can improve diagnostic performance.