Volume perfusion CT imaging of cerebral vasospasm: diagnostic performance of different perfusion maps

In this study, we aimed to evaluate the diagnostic performance of different volume perfusion CT (VPCT) maps regarding the detection of cerebral vasospasm compared to angiographic findings.

Forty-one datasets of 26 patients (57.5 ± 10.8 years, 18 F) with subarachnoid hemorrhage and suspected cerebral vasospasm, who underwent VPCT and angiography within 6 h, were included. Two neuroradiologists independently evaluated the presence and severity of vasospasm on perfusion maps on a 3-point Likert scale (0-no vasospasm, 1-vasospasm affecting <50 %, 2-vasospasm affecting >50 % of vascular territory). A third neuroradiologist independently assessed angiography for the presence and severity of vasospasm on a 3-point Likert scale (0-no vasospasm, 1-vasospasm affecting < 50 %, 2-vasospasm affecting > 50 % of vessel diameter). Perfusion maps of cerebral blood volume (CBV), cerebral blood flow (CBF), mean transit time (MTT), and time to drain (TTD) were evaluated regarding diagnostic accuracy for cerebral vasospasm with angiography as reference standard. Correlation analysis of vasospasm severity on perfusion maps and angiographic images was performed. Furthermore, inter-reader agreement was assessed regarding findings on perfusion maps.

Diagnostic accuracy for TTD and MTT was significantly higher than for all other perfusion maps (TTD, AUC = 0.832; MTT, AUC = 0.791; p < 0.001). TTD revealed higher sensitivity than MTT (p = 0.007). The severity of vasospasm on TTD maps showed significantly higher correlation levels with angiography than all other perfusion maps (p ≤ 0.048). Inter-reader agreement was (almost) perfect for all perfusion maps (kappa ≥ 0.927).

The results of this study indicate that TTD maps have the highest sensitivity for the detection of cerebral vasospasm and highest correlation with angiography regarding the severity of vasospasm.