Background: Computed tomography (CT) is a cornerstone of modern diagnostic imaging; however, concerns regarding radiation exposure and image noise persist. Conventional filtered back projection (FBP) reconstruction is limited by increased noise at reduced radiation doses. Advanced reconstruction techniques, including iterative reconstruction (IR) and deep learning reconstruction (DLR), have been developed to enhance image quality while enabling radiation dose optimization.

Objective: To evaluate the impact of advanced reconstruction algorithms on CT image quality and to assess their potential role in radiation dose reduction compared with conventional FBP.

Materials and Methods: This prospective observational study included 150 adult patients undergoing routine CT chest or CT abdomen examinations. All datasets were reconstructed using FBP, IR, and DLR techniques. Radiation dose parameters, including CT dose index volume (CTDIvol) and dose-length product (DLP), were recorded. Objective image quality was assessed using signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). Subjective image quality evaluation was independently performed by two experienced radiologists using a five-point Likert scale. Statistical analysis was conducted to compare reconstruction techniques, with p < 0.05 considered significant.

Results: Advanced reconstruction algorithms demonstrated significant improvements in image quality compared with FBP. Both IR and DLR showed higher SNR and CNR values (p < 0.001), with DLR achieving the greatest noise reduction and contrast preservation. Radiation dose reductions of approximately 20–30% were achievable without compromising diagnostic image quality. Subjective assessments revealed superior image noise characteristics, anatomical detail, and diagnostic confidence with DLR, while all IR- and DLR-reconstructed images were deemed diagnostically acceptable.

Conclusion: Advanced CT reconstruction algorithms significantly enhance image quality and enable meaningful radiation dose reduction. Deep learning reconstruction, in particular, offers superior performance over conventional and iterative methods, supporting its integration into routine clinical CT imaging to improve diagnostic accuracy and patient safety.

Keywords: Computed Tomography; Iterative Reconstruction; Deep Learning Reconstruction; Image Quality