HomeNewsHigher image quality and diagnostic accuracy of contrast-enhanced CT scans of the upper abdomen obtained by applying a deep learning image reconstruction algorithm compared to adaptive statistical iterative reconstruction

Higher image quality and diagnostic accuracy of contrast-enhanced CT scans of the upper abdomen obtained by applying a deep learning image reconstruction algorithm compared to adaptive statistical iterative reconstruction

    In a prospective, single-center study, 50 consecutive oncologic patients with 130 hypovascular liver lesions underwent abdominal contrast-enhanced CT with images reconstructed using two reconstruction algorithms: deep learning image reconstruction (DLIR) with three intensity levels (low, medium, and high), and conventional adaptive statistical iterative reconstruction (ASiR-V) with ten strength levels from 10% to 100%. All images were assessed for objective and subjective image quality, and the diagnostic accuracy of the DLIR algorithm was calculated.
    DLIR at medium level was found to have the highest subjective image quality compared to ASiR-V, as well as better diagnostic accuracy (93.8%) than ASiR-V at 50% for the assessment of liver lesions.