In 170 patients, coronary artery calcium scores derived from PCCT PureCalcium virtual noncontrast images showed excellent correlation with true noncontrast (TNC) calcium scoring (ICC 0.98). Median Agatston scores were not significantly different: 4.8 vs 2.7, respectively (p = 0.99).
High correlation did not translate into complete interchangeability. PureCalcium correctly classified 125/170 patients (74%) into CAD-RADS plaque-burden categories, with excellent categorical agreement (κ = 0.88), but discrepancies were concentrated around the clinically important P0/P1 boundary.
Eliminating the dedicated TNC acquisition could reduce total cardiac CT radiation exposure by approximately 20%, since the TNC scan contributed 19.7% ± 8.8% of total examination dose. PureCalcium therefore shows considerable potential for single-scan CCTA plus calcium assessment, although caution remains necessary in patients with very low calcium burden.
Access publication
Figure 2. Representative images of coronary artery calcifications in true noncontrast coronary artery calcium scoring (CACS) (left column), PureCalcium images from coronary CT angiography (CCTA) (middle column), and contrast-enhanced CCTA (right column). All patients underwent CCTA for evaluation of coronary artery disease, with reconstruction in the axial orientation. (A) Images in a 47-year-old female patient with no coronary artery calcifications, P0 in the Coronary Artery Disease Reporting and Data System, or CAD-RADS 2.0. (B) Images in a 70-year-old male patient with mild coronary artery calcifications, P1 in CAD-RADS 2.0. (C) Images in a 72-year-old female patient with moderate coronary artery calcifications, P2 in CAD-RADS 2.0. (D) Images in a 65-year-old male patient with severe coronary artery calcifications, P3 in CAD-RADS 2.0. PureCalcium is a customized algorithm for virtual noncontrast image reconstruction derived from photon-counting CCTA specifically for the purpose of CACS.
Protocol
Detail | Value |
|---|
Scanner | NAEOTOM Alpha, VA50 |
Software | syngo.via VB60 |
Scan area | Coronary arteries |
TNC scan mode | High-pitch |
CCTA scan mode | Flash, sequential, or spiral depending on heart rate/rhythm |
Acquisition mode | Standard-resolution PCCT |
Tube voltage [kV] | 120 |
IQ level | 100 |
Dose modulation | Automatic tube current modulation |
Matrix | 512 × 512 |
TNC slice thickness / increment [mm] | 3.0 / 1.5 |
CCTA slice thickness / increment [mm] | 0.4 / 0.2 |
TNC kernel | Qr36 |
TNC iterative reconstruction | None |
TNC reconstruction energy | 70 keV |
PureCalcium slice thickness / increment [mm] | 2.0 / 1.5 |
PureCalcium kernel | Qr40 |
PureCalcium iterative reconstruction | QIR 3 |
Calcium scoring thickness | Automatic resampling to 1.5 mm |
Calcium threshold | 130 HU; minimum area 0.5 mm² |
Contrast agent | Iohexol, Accupaque 300 |
Contrast volume | 50–65 mL |
Injection rate | 4 mL/s |
Contrast timing | Test bolus |
Heart-rate control | Up to 15 mg IV metoprolol if required |
Vasodilation | 0.4 mg sublingual nitroglycerin if not contraindicated |
Key Results
Parameter | Main finding |
|---|
Study population | 170 patients; 111 with CACS >0 |
Median Agatston score | TNC 4.8 vs PureCalcium 2.7, p = 0.99 |
Score correlation | ICC 0.98 (95% CI 0.97–0.99) |
Plaque-burden classification | 125/170 (74%) correctly classified; κ = 0.88 |
P0/P1 discrepancies | 20 patients classified P1 instead of P0; 14 classified P0 instead of P1 |
Radiation contribution of TNC | DLP 32.5 ± 11.5 mGy·cm; 19.7% ± 8.8% of total examination dose |
Source:
Haag NP, Michael AE, Lennartz S, Panknin C, Niehoff JH, Borggrefe J, Shahzadi I, Zwanenburg A, Kroeger JR. Coronary Artery Calcium Scoring Using Virtual Versus True Noncontrast Images From Photon-Counting Coronary CT Angiography. Radiology. 2024;310(3). doi:10.1148/radiol.230545.
