Impact of model-based iterative reconstruction (MBIR) on image quality in cerebral CT angiography before and after intracranial aneurysm treatment

Abstract

PurposeTo subjectively and objectively assess the impact of model-based iterative reconstruction(MBIR) on image quality in cerebral computed tomography angiography compared to adaptive statistical iterative reconstruction (ASIR). Methods107 patients (mean age: 58 ± 14 years) were included prior to (n = 38) and after (n = 69) intracranial aneurysm treatment. Images were acquired using a routine protocol and reconstructed with MBIR and ASIR. Image noise, signal-to-noise (SNR) and contrast-to-noise (CNR) ratios in the internal carotid and middle cerebral arteries were compared between MBIR and ASIR using the Wilcoxon signed-rank test. Additionally, two neuroradiologists subjectively assessed noise, artefacts, vessel sharpness and overall quality using a semi-quantitative assessment scale. ResultsObjective assessment revealed that MBIR reduced noise (p < 0.0001) and additionally improved SNR (p < 0.0001) and CNR (p < 0.0001) compared to ASIR in untreated and treated patients. Subjective assessment revealed that in untreated patients, MBIR improved noise reduction, artefacts, vessel sharpness and overall quality relative to ASIR (p < 0.0001). In the treated groups, noise and vessel sharpness were improved (p < 0.0001) with no change in artefacts on images reconstructed with MBIR compared to ASIR. ConclusionMBIR significantly improves noise, SNR, CNR and vessel sharpness in untreated and treated patients with intracranial aneurysms. MBIR does not reduce artefacts generated by metallic devices following intracranial aneurysm treatment.