Selective Size Imaging via Diffusion MRI for Preoperative IDH Mutation Status, Grading, and Progression‐Free Survival Risk Stratification in Adult‐Type Diffuse Glioma

ABSTRACT Background Selective size imaging using filters through diffusion times (SSIFT) MRI provides cell size‐weighted contrast, but its value in adult‐type diffuse gliomas remains unclear. Purpose To evaluate SSIFT MRI for preoperative isocitrate dehydrogenase (IDH) mutation status assessment, World Health Organization (WHO) central nervous system (CNS) grading and progression‐free survival (PFS) risk stratification were compared with apparent diffusion coefficient (ADC) metrics. Study Type Prospective, single‐center. Population Seventy‐one patients with pathologically confirmed adult‐type diffuse gliomas (median age, 52 years; 42 men, 29 women) with Grades 2, 3, and 4 gliomas ( n = 17, 12, and 42). Field Strength/Sequence 3 T; conventional MRI included T1‐weighted imaging, T2‐weighted FLAIR imaging, and diffusion‐weighted imaging with ADC mapping. Oscillating and pulsed gradient spin‐echo diffusion encoding. Assessment Whole‐solid‐tumor regions of interest were delineated, and extracted metrics were compared to pathologic and molecular findings. SSIFT‐derived incremental area under the curve (iAUC) quantified normalized diffusion‐weighted signal changes with diffusion time. Extracted metrics included mean and 90th‐percentile iAUC values (iAUC_mean, iAUC_90Percentile) and mean and 10th‐percentile ADC values (ADC_mean, ADC_10Percentile). Statistical Tests Kruskal–Wallis and Jonckheere–Terpstra tests, ROC analysis with paired DeLong tests, Cox regression, Kaplan–Meier analysis, and bootstrap comparison of integrated time‐dependent AUC (AUCint). p < 0.05 indicated statistical significance. Results Both iAUC metrics yielded AUCs of 0.839 for IDH assessment (95% confidence intervals [CIs], 0.731–0.938 and 0.731–0.941), significantly outperforming ADC_mean (0.697; 0.573–0.821). iAUC_90Percentile differentiated Grade 2 from Grade 3 gliomas (AUC = 0.789; 95% CI: 0.571–0.966) and outperformed ADC_10Percentile (AUC = 0.520; 95% CI: 0.284–0.722). During median follow‐up of 469 days, 26 patients progressed. iAUC_mean remained significantly associated with PFS after multivariable adjustment (HR = 1.206; 95% CI: 1.048–1.387). Adding iAUC_mean increased bootstrapped AUCint from 0.788 to 0.815, although the increase was not significant (ΔAUCint = 0.027; 95% CI: −0.003 to 0.078; p = 0.0899). Data Conclusion SSIFT‐derived iAUC might provide cell size‐related microstructural information complementary to ADC and may aid preoperative IDH mutation status assessment and glioma grading. Evidence Level 1. Technical Efficacy Stage 2.