Abstract
BACKGROUND AND OBJECTIVES: Generating computed tomography (CT) angiography (CTA) 3-dimensional (3D) volume-rendered (3DVR) images can be time consuming without specialized technical staff or artificial intelligence solutions. However, their role in aneurysm detection in patients with subarachnoid hemorrhage is not known. Our aim was to assess the diagnostic utility of 64-detector row CTA with 3DVR (CTA+3DVR) vs without 3DVR (CTA-3DVR) in intracranial aneurysm detection.
METHODS: A retrospective analysis of patients presenting with spontaneous subarachnoid hemorrhage (regardless of location) who underwent 64-detector row CTA and subsequent digital subtraction angiography (DSA) between 2013 and 2020 was performed. DSA was the reference standard. Almost all DSAs were performed with 3D rotational angiography. Two neuroradiologists blinded to the DSA results separately reviewed CTA source and maximum intensity projection images without 3DVR images (CTA-3DVR) and then immediately followed by the 3DVR images (CTA + 3DVR). Disagreements were resolved by consensus review. Aneurysm size was measured on DSA.
RESULTS: In total, 200 patients were included in the study. 140 aneurysms in 114 patients were detected on DSA. CTA-3DVR detected 135 of 140 aneurysms (96.4%), and CTA+3DVR detected 136 of 140 aneurysms (97.1%). All missed aneurysms measured 2-4 mm. Three of four missed aneurysms were associated with multiple aneurysms, and 1 was a singly occurring, 2-mm M3 segment aneurysm. The 1 additional aneurysm detected by CTA + 3DVR was a 3-mm left A3 segment aneurysm. No false positives occurred with or without 3DVR.
CONCLUSION: CTA+3DVR detected 1 additional distal A3 aneurysm vs CTA-VR. 3DVR does not seem to substantially improve the detection of aneurysms but may be useful in the emergent setting for aneurysm morphology characterization.