Differentiation between high-grade gliomas and solitary brain metastases based on multidiffusion MRI model...

Differentiating high-grade gliomas (HGGs) from solitary brain metastases (SBMs) using conventional magnetic resonance imaging (MRI) remains challenging due to their similar imaging features. This study aimed to evaluate the diagnostic performance of advanced diffusion models, such as neurite orientation dispersion and density imaging (NODDI) and mean apparent propagator magnetic resonance imaging (MAP-MRI), incomparison to traditional techniques like diffusion-weighted imaging (DWI), diffusion tensor imaging (DTI), and diffusion kurtosis imaging (DKI) for distinguishing HGGs from SBMs.

In total, 17 patients with HGGs and 26 patients with SBMs were prospectively recruited based on the established inclusion and exclusion criteria. Structural MRI sequences and diffusion spectrum imaging (DSI) were utilized to assess quantitative parameter models, including NODDI, MAP-MRI, DWI, DTI, and DKI. Quantitative parameters were measured for both the tumor parenchymal area and the peritumoral edema area. The quantitative parameters of the two patient groups were compared using either the independent Student's t-test or the Mann-Whitney U test. The effectiveness of each model was evaluated using receiver operating characteristic (ROC) curves and calculating the area under the curve (AUC). Finally, the DeLong test was employed to compare the diagnostic performance of each model through pairwise comparisons of ROC curves.

Isotropic volume fraction (V_iso) based on NODDI; mean squared displacement (MSD) and the return to plane probabilities (RTPP) based on MAP-MRI; radial diffusivity (RD_k) and mean diffusivity (MD_k) based on DKI; and axial diffusivity (AD), radial diffusivity (RD), and mean diffusivity (MD) based on DTI of the peritumoral edema tumor were significantly different between HGGs and SBMs (p < 0.05). The optimal single discriminant parameters for each model are NODDI_V_iso, MAP-MRI_MSD, DKI_MD_k, and DTI_AD. Among these, the AUC of V_iso (0.809) exceeds that of MSD (0.733), MD_k (0.718), and AD (0.779). The combined model, which incorporates DTI_AD, DKI_RD, and NODDI_V_iso, demonstrated superior diagnostic performance (0.897).

Advanced diffusion MRI quantitative parameters derived from NODDI, such as V_iso, have the potential to enhance the differentiation between HGGs and SBMs. The integrated utilization of these models is anticipated to enhance diagnostic accuracy and refine MRI protocols for brain tumor assessment.