Diffusion–relaxation correlation spectrum imaging for predicting tumor consistency and gross total resection in patients with pituitary adenomas: a preliminary study

Objective To evaluate the ability of diffusion–relaxation correlation spectrum imaging (DR-CSI) to predict the consistency and extent of resection (EOR) of pituitary adenomas (PAs). Methods Forty-four patients with PAs were prospectively enrolled. Tumor consistency was evaluated at surgery as either soft or hard, followed by histological assessment. In vivo DR-CSI was performed and spectra were segmented following to a peak-based strategy into four compartments, designated A (low ADC), B (mediate ADC, short T2), C (mediate ADC, long T2), and D (high ADC). The corresponding volume fractions ( f A , f B , f C , f D ) along with the ADC and T2 values were calculated and assessed using univariable analysis for discrimination between hard and soft PAs. Predictors of EOR > 95% were analyzed using logistic regression model and receiver-operating-characteristic analysis. Results Tumor consistency was classified as soft ( n = 28) or hard ( n = 16). Hard PAs presented higher f B ( p = 0.001) and lower f C ( p = 0.013) than soft PAs, while no significant difference was found in other parameters. f B significantly correlated with the level of collagen content ( r = 0.448, p = 0.002). Knosp grade (odds ratio [OR], 0.299; 95% confidence interval [CI], 0.124–0.716; p = 0.007) and f B (OR, 0.834, per 1% increase; 95% CI, 0.731–0.951; p = 0.007) were independently associated with EOR > 95%. A prediction model based on these variables yielded an AUC of 0.934 (sensitivity, 90.9%; specificity, 90.9%), outperforming the Knosp grade alone (AUC, 0.785; p