Influence of image reconstruction parameters on cardiovascular risk reclassification by Computed Tomography Coronary Artery Calcium Score
- Authors: Mantini, C.; Maffei, E.; Toia, P.; Ricci, F.; Seitun, S.; Clemente, A.; Malagò, R.; Runza, G.; La Grutta, L.; Midiri, M.; Cotroneo, A.; Forte, E.; Cademartiri, F.
- Publication year: 2018
- Type: Articolo in rivista (Articolo in rivista)
- OA Link: http://hdl.handle.net/10447/280512
Objective: To investigate the influence of different CT reconstruction parameters on coronary artery calcium scoring (CACS) values and reclassification of predicted cardiovascular (CV) risk. Methods: CACS was evaluated in 113 patients undergoing ECG-gated 64-slice CT. Reference CACS protocol included standard kernel filter (B35f) with slice thickness/increment of 3/1.5 mm, and field-of-view (FOV) of 150â180 mm. Influence of different image reconstruction algorithms (reconstructed slice thickness/increment 2.0/1.0â1.5/0.8â3.0/2.0â3.0/3.0 mm; slice kernel B30f-B45f; FOV 200â250 mm) on Agatston score was assessed by Bland-Altman plots and concordance correlation coefficient (CCC) analysis. Classification of CV risk was based on the Mayo Clinic classification. Results: Different CACS reconstruction parameters showed overall good accuracy and precision when compared with reference protocol. Protocols with larger FOV, thinner slices and sharper kernels were associated with significant CV risk reclassification. Use of kernel B45f showed a moderate positive correlation with reference CACS protocol (Agatston CCC = 0.67), and yielded significantly higher CACS values (p <.05). Reconstruction parameters using B30f or B45f kernels, 250 mm FOV, or slice thickness/increment of 2.0/1.0 mm or 1.5/0.8 mm, were associated with significant reclassification of CV risk (p <.05). Conclusions: Kernel, FOV, slice thickness and increment are major determinants of accuracy and precision of CACS measurement. Despite high agreement and overall good correlation of different reconstruction protocols, thinner slices thickness and increment, and sharper kernels were associated with significant upward reclassification of CV risk. Larger FOV determined both upward and downward reclassification of CV risk.