Clin Res Cardiol (2022). https://doi.org/10.1007/s00392-022-02002-5

Influence of slice thickness and iterative reconstruction on coronary artery calcification quantification
N. Mekkhala1, für die Studiengruppe: UKE
1Med. Klinik 2 - Kardiologie, Angiologie, Universitätsklinikum Erlangen, Erlangen;

Introduction: Reconstruction parameters including Iterative reconstruction algorithms and slice thickness influence the quantification of coronary artery calcification. Hence, with modern scanner platforms and increasing use of iterative reconstruction algorithms for image noise reduction, assessment of calcified disease burden and coronary calcium quantification will ultimately be affected. We sought to analyze the influence of slice thickness and iterative reconstruction on quantitative parameters of coronary calcification.

Methods: Consecutive patients referred for coronary CT angiography for suspected coronary artery disease were included in this analysis. In all patients, native data sets for quantification of coronary calcifications were acquired prior to contrast enhancement. All native CT data were acquired using a third-generation dual source CT system (Somatom Force, Siemens Healthineers, Forchheim, Germany) using prospectively ECG-triggered spiral acquisition at 60% of the cardiac cycle. Besides the standard reconstruction (3.0 mm thickness, 1.5 increment and medium sharp reconstruction kernel), 5 additional reconstructions were rendered using 2 mm and 1 mm slice thickness as well as 3 mm, 2 mm and 1 mm slice thickness with iterative reconstruction (IR) level 2 (Admire®, Siemens Healthineers, Forchheim, Germany). Data sets were transferred to a dedicated workstation and calcium scoring was performed using commercially available software (Ziostation2, Ziosoft inc., Tokyo, Japan). The Agatston method as well as the volume method were used for quantification of coronary calcification using a standard threshold of 130 Hounsfield units (HU) for calcium detection. All reconstructions were compared to the standard reconstruction of 3 mm without IR concerning differences in Agatston score and calcium volume.

Results: 104 patients (624 reconstructions) were included (mean age 67±10 years, 62 males). Compared to standard 3 mm reconstructions, Agatston score and calcium volume showed a significant positive correlation using all other reconstruction parameters (r 0.96, p<0.00001 for all). With decreasing slice thickness, median Agatston score was significantly higher compared to standard reconstruction (median [IQR]: 25 [0;460] for 3 mm vs. 48 [0;512] for 2 mm and 92 [5, 566] for 1 mm, p< 0.0001). With the use of iterative reconstruction level 2, a significant underestimation of Agatston score was observed compared to standard 3.0 mm reconstruction (median [IQR]: 25 [0;460] for 3 mm without IR, vs. 18 [0; 364] for 3 mm + IR, p<0.0001 and 60 [2; 477] for 1 mm + IR, p=0.002). Bland Altman analysis showed significant systematic overestimation of the Agatston score when thin slices were used for reconstruction (mean bias [95% limits of agreement]: -31 [-350 to 287] for 2 mm and mean bias -99 [-491 to 292] for 1 mm). Furthermore, significant underestimation of coronary calcification was noticed when IR was added (mean bias 80 AU [-235 to 296] for 3.0 mm + IR). This systematic underestimation was slightly compensated with thinner slices however with wide limits of agreement (mean bias 32 AU [-183 to 248] for 2.0 mm + IR and mean bias -7 AU [-251 to 236] for 1.0 mm + IR). Similar patterns were observed using the volume method.

Conclusion: Slice thickness and iterative reconstruction significantly influence the quantification of coronary calcification. Iterative reconstruction cause significant underestimation of calcified disease.


https://dgk.org/kongress_programme/jt2022/aP1232.html