Evaluation of optimal ClearInfinity artificial intelligence deep learning reconstruction algorithm weights with a...

Abdominal computed tomography angiography (CTA) often requires a high tube voltage and relatively large contrast volume to ensure adequate vascular enhancement, which increases radiation exposure and contrast-related risks. Deep learning reconstruction techniques have demonstrated potential to improve image quality under low-dose acquisition conditions. This study evaluated a 70-kVp abdominal CTA protocol using low tube voltage and low contrast medium volume ("dual-low"), reconstructed with ClearInfinity (Neusoft Medical) at different weighting levels.

A prospective study was conducted on 100 patients undergoing abdominal CTA who were divided into two groups: group A (70 kVp, 40-45 mL of contrast medium, and a flow rate of 2.0-2.5 mL/s) and group B (conventional 100-kVp protocol, 60-70 mL of contrast medium, and a flow rate of 3.0-3.5 mL/s). Group B served as standard-dose reference, and its images were reconstructed using the vendor-recommended ClearView iterative reconstruction. Images in Group A were reconstructed with ClearInfinity at four weighting levels-30%, 50%, 70%, and 90%-which corresponded to groups A1 to A4, respectively. Radiation dose parameters, including computed tomography dose index volume (CTDIvol), dose-length product (DLP), and effective dose (ED) were recorded and analyzed for both groups. Objective evaluation included CT attenuation, image noise (SD), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR). Subjective image quality was independently evaluated by two radiologists.

The CTDIvol, DLP, and ED in group A were significantly lower than those in group B (P<0.001). Mean vascular attenuation did not differ significantly between the groups (P>0.05). As the ClearInfinity weighting increased from 30% to 90%, image noise progressively decreased, whereas both the SNR and CNR progressively increased. Compared with the iterative reconstruction group (group B), group A4 exhibited a lower SD and higher SNR and CNR across major abdominal vessels (all P values <0.05). Subjective image quality increased from A1 to A2 and then remained stable at higher ClearInfinity weights, with no significant differences across groups A2-A4. Interobserver agreement for subjective scoring was excellent (κ>0.81).

A 70-kVp abdominal CTA protocol reconstructed with 90% ClearInfinity weighting provides substantial radiation dose reduction while preserving diagnostic vascular image quality despite reduced contrast volume.