Deep Learning Reconstruction Combined with Contrast-Enhancement Boost Technique in "Quadruple-low" CCTA Protocol:
Hua Zhou1, Hao Tian1, WenYue Chen2
1Department of Medical Imaging, Affiliated Hospital of Jiangnan University, Wuxi, Jiangsu 214122, China.
Rationale And Objectives:
To evaluate the impact of a deep learning reconstruction (DLR) algorithm combined with contrast-enhancement boost (CE-boost) technology on image quality and diagnostic performance in coronary CT angiography (CCTA) using a low tube voltage, low flow rate, low-contrast volume, and low concentration ("Quadruple-low") protocol.
Materials And Methods:
In this prospective study, 102 patients were randomly assigned to the routine-dose (RD, n = 51) or quadruple-low (QL, n = 51) group for CCTA. The RD group received 0.6mL/kg of contrast agent (350mgI/mL) at 5mL/s, while the QL group received 0.4mL/kg (320mgI/mL) at 3.5mL/s. The RD group was reconstructed using hybrid iterative reconstruction (HIR). The QL group underwent four reconstruction workflows (HIR, DLR, CE-boost+HIR, CE-boost+DLR). Objective parameters included signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of major coronary arteries. Subjective image quality was rated by two radiologists. Coronary plaque burden was assessed using CAD-RADS 2.0, segment stenosis score (SSS), segment involvement score (SIS), and coronary artery calcium score (CACS).
Results:
The QL protocol reduced radiation dose by 78.3% (1.1 ± 0.4 vs. 4.8 ± 2.2mSv, P<0.001) and contrast volume by 43% (P<0.01) compared to RD. QL-DLR-boost images achieved the highest SNR, CNR, and subjective quality scores (P<0.001). CAD-RADS, SSS, SIS, and CACS did not differ significantly among reconstruction methods (P = 0.62-0.85). Against invasive coronary angiography, QL-DLR-boost yielded the highest diagnostic accuracy (96.7%) with 100% sensitivity.
Conclusion:
DLR combined with CE-boost enables CCTA under the "Quadruple-low" protocol to achieve superior diagnostic image quality while markedly reducing radiation dose and contrast volume.
