PT - JOURNAL ARTICLE AU - David Fiorella AU - Aquilla Turk AU - Imran Chaudry AU - Raymond Turner AU - Jared Dunkin AU - Clemente Roque AU - Marily Sarmiento AU - Yu Deuerling-Zheng AU - Christine M Denice AU - Marlene Baumeister AU - Adrian T Parker AU - Henry H Woo TI - A prospective, multicenter pilot study investigating the utility of flat detector derived parenchymal blood volume maps to estimate cerebral blood volume in stroke patients AID - 10.1136/neurintsurg-2013-010840 DP - 2014 Jul 01 TA - Journal of NeuroInterventional Surgery PG - 451--456 VI - 6 IP - 6 4099 - http://jnis.bmj.com/content/6/6/451.short 4100 - http://jnis.bmj.com/content/6/6/451.full SO - J NeuroIntervent Surg2014 Jul 01; 6 AB - Purpose Newer flat panel angiographic detector (FD) systems have the capability to generate parenchymal blood volume (PBV) maps. The ability to generate these maps in the angiographic suite has the potential to markedly expedite the triage and treatment of patients with acute ischemic stroke. The present study compares FP-PBV maps with cerebral blood volume (CBV) maps derived using standard dynamic CT perfusion (CTP) in a population of patients with stroke. Methods 56 patients with cerebrovascular ischemic disease at two participating institutions prospectively underwent both standard dynamic CTP imaging followed by FD-PBV imaging (syngo Neuro PBV IR; Siemens, Erlangen, Germany) under a protocol approved by both institutional review boards. The feasibility of the FD system to generate PBV maps was assessed. The radiation doses for both studies were compared. The sensitivity and specificity of the PBV technique to detect (1) any blood volume deficit and (2) a blood volume deficit greater than one-third of a vascular territory, were defined using standard dynamic CTP CBV maps as the gold standard. Results Of the 56 patients imaged, PBV maps were technically adequate in 42 (75%). The 14 inadequate studies were not interpretable secondary to patient motion/positioning (n=4), an injection issue (n=2), or another reason (n=8). The average dose for FD-PBV was 219 mGy (median 208) versus 204 mGy (median 201) for CT-CBV. On CT-CBV maps 26 of 42 had a CBV deficit (61.9%) and 15 (35.7%) had a deficit that accounted for greater than one-third of a vascular territory. FD-PBV maps were 100% sensitive and 81.3% specific to detect any CBV deficit and 100% sensitive and 62.9% specific to detect any CBV deficit of greater than one-third of a territory. Conclusions PBV maps can be generated using FP systems. The average radiation dose is similar to a standard CTP examination. PBV maps have a high sensitivity for detecting CBV deficits defined by conventional CTP. PBV maps often overestimate the size of CBV deficits. We hypothesize that the FP protocol initiates PBV imaging prior to complete saturation of the blood volume in areas perfused via indirect pathways (ie, leptomeningeal collaterals), resulting in an overestimation of CBV deficits, particularly in the setting of large vessel occlusion.