REVIEW   

Journal of Radiological Review 2020 May-June;7(3):185-95

DOI: 10.23736/S2723-9284.20.00029-4

Copyright © 2020 EDIZIONI MINERVA MEDICA

language: English, Italian

Iterative model-based CT reconstruction algorithm: the background and added clinical value

Davide IPPOLITO ^{1, 2} ✉, Cesare MAINO ^{1, 2}, Luca RIVA ^{1, 2}, Anna PECORELLI ^{1, 2}, Andrea DE VITO ^{1, 2}, Sophie LOMBARDI ¹, Maria RAGUSI ^{1, 2}, Teresa GIANDOLA ^{1, 2}, Cammillo TALEI FRANZESI ^{1, 2}, Sandro SIRONI ^{2, 3}

¹ Department of Diagnostic Radiology, San Gerardo Hospital, Monza, Italy; ² School of Medicine, University of Milano-Bicocca, Milan, Italy; ³ Department of Diagnostic Radiology, Papa Giovanni XXIII Hospital, Bergamo, Italy

In recent years, many studies have investigated the reduction of the radiation dose exposure in the head and body CT and the increase in image quality by applying different reconstruction algorithms. Compared to previous generations of image reconstruction algorithms, model-based iterative algorithms (MBIR) are mathematically more complex and accurate, using a knowledge-based approach to process data statistics and system models that depict geometry and physical characteristics of the CT scanner. Recently, model-based reconstruction algorithms become more employed in clinical practice, thanks to the intrinsic superiority to obtain a better image quality and an important dose reduction. This review aims to evaluate the importance of the model-based approach in all anatomical districts.

KEY WORDS: Multidetector computed tomography; Radiation dosage; Computer-assisted image processing; Knowledge bases