Identification of a Single-Dose, Low-Flip-Angle–Based CBV Threshold for Fractional Tumor Burden Mapping in Recurrent Glioblastoma

Histogram of CCC across each patient, ranging from 0.52 to 0.96. Seventy-seven percent of patients exhibit a CCC greater than 0.8.

The area under the ROC curve (AUC) for the optimal LFA-based thresholds is found to be 0.95 and 0.96 for sRCBV <1.0 and >1.56, respectively.

Boxplots (with individual datapoints) showing the consistency between percentage of tumor voxels (sRCBV >1.0) in the enhancing tumor for the double-dose, MFA and single-dose, LFA protocols.

A–C, The patient shown is a 73-year-old woman with grade-IV glioblastoma presenting 15 months after surgical resection. D–F, The patient shown is a 34-year-old man with grade-IV glioblastoma presenting 20 months after surgical resection. Images include anatomic postcontrast T1-weighted images (A and D), FTB maps for the single-dose, LFA protocol (B and E) by using 1.0 and 1.37, and the reference double-dose, MFA protocol (C and F) by using 1.0 and 1.56 superimposed on the contrast-enhanced T1-weighted images. Blue, yellow, and red voxels represent PTRE (FTBlow, sRCBV <1.0), tumor/treatment effect admixture (FTBmid, 1.0 > sRCBV <1.37 [LFA], 1.56 [MFA]), and high tumor cell probability (FTBhigh, sRCBV >1.37 [LFA], 1.56 [MFA]), respectively. The Dice similarity coefficients comparing the LFA and MFA sRCBV for PTRE and tumor recurrence for the patient in (A–C) are 0.85 and 0.87, respectively. For the patient in (D–F), the coefficients are 0.77 for PTRE and 0.96 for tumor recurrence.