| Sumario: | This article focuses on evaluating the image quality and diagnostic performance of virtual monoenergetic images (VMI) reconstructed at different kiloelectron volt (keV) levels using spectral computed tomography (CT) in patients with colorectal cancer (CRC). The study found that VMI at 40 keV (VMI40keV) reconstructed with standard-dose adaptive statistical iterative reconstruction (AV60) provided higher signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) than polychromatic images (PEI) and other VMI levels, improving lesion detection and diagnostic accuracy for CRC. Additionally, medium-strength deep learning image reconstruction (DLIRM) applied to reduced-dose CT scans combined with VMI40keV maintained comparable image quality to standard-dose PEI, suggesting potential for radiation dose reduction in CRC screening. The findings indicate that VMI40keV enhances radiologists’ sensitivity and negative predictive value in CRC diagnosis, although specificity showed a slight decrease, and further research is needed to optimize reduced-dose protocols and assess early-stage CRC detection.
|
|---|