Intra-individual comparison of coronary calcium scoring between photon counting detector- and energy integrating detector-CT: Effects on risk reclassification

Abstract

PurposeTo compare coronary artery calcium volume and score (CACS) between photon-counting detector (PCD) and conventional energy integrating detector (EID) computed tomography (CT) in a phantom and prospective patient study.MethodsA commercially available CACS phantom was scanned with a standard CACS protocol (120 kVp, slice thickness/increment 3/1.5 mm, and a quantitative Qr36 kernel), with filtered back projection on the EID-CT, and with monoenergetic reconstruction at 70 keV and quantum iterative reconstruction off on the PCD-CT. The same settings were used to prospectively acquire data in patients (n = 23, 65 ± 12.1 years), who underwent PCD- and EID-CT scans with a median of 5.5 (3.0–12.5) days between the two scans in the period from August 2021 to March 2022. CACS was quantified using a commercially available software solution. A regression formula was obtained from the aforementioned comparison and applied to simulate risk reclassification in a pre-existing cohort of 514 patients who underwent a cardiac EID-CT between January and December 2021.ResultsBased on the phantom experiment, CACSPCD–CT showed a more accurate measurement of the reference CAC volumes (overestimation of physical volumes: PCD-CT 66.1 ± 1.6% vs. EID-CT: 77.2 ± 0.5%). CACSEID–CT and CACSPCD–CT were strongly correlated, however, the latter measured significantly lower values in the phantom (CACSPCD–CT: 60.5 (30.2–170.3) vs CACSEID–CT 74.7 (34.6–180.8), p = 0.0015, r = 0.99, mean bias –9.7, Limits of Agreement (LoA) –36.6/17.3) and in patients (non-significant) (CACSPCD–CT: 174.3 (11.1–872.7) vs CACSEID–CT 218.2 (18.5–876.4), p = 0.10, r = 0.94, mean bias –41.1, LoA –315.3/232.5). The systematic lower measurements of Agatston score on PCD-CT system led to reclassification of 5.25% of our simulated patient cohort to a lower classification class.ConclusionCACSPCD–CT is feasible and correlates strongly with CACSEID–CT, however, leads to lower CACS values. PCD-CT may provide results that are more accurate for CACS than EID-CT.