Influence of post-bolus acquisition duration on estimated percentage of signal recovery from dynamic susceptibility contrast MRI in brain tumors

Evaluating the percentage of signal recovery (PSR) from DSC MRI facilitates the differential diagnosis of brain tumors. While prior studies have illustrated PSR dependency on pulse sequence parameters and preload, the influence of the post-bolus acquisition duration remains under-investigated. In this technical note, we examined 8 brain tumor patients with particularly long DSC acquisitions (∼5min). DSC time-intensity curves (TICs) were trimmed to different lengths after bolus arrival time (BAT). PSR measurements were highly dependent on TIC length. Rank consistency (ρ=0.85) and value consistency (ICC=0.86), compared to a reference PSR calculated closer to the steady state (4min 12s from BAT), was satisfactory for TIC lengths ≥2 minutes after BAT. This advocates for TIC length standardization, to improve the accuracy and generalizability of PSR measurements for differential diagnosis. A DSC sequence duration around 3 minutes should be preferred, to ensure ≤1 minute for BAT plus ≥2 minutes of data collection after BAT.