Evaluation of model based iterative reconstruction IMR in computed tomography. (September 2016)

• Record Type: Journal Article
• Authors: Paruccini, N.; Pasquali, C.; Villa, R.; Spadavecchia, C.; Baglivi, A.; Radice, A.; Crespi, A.
• Is Part Of: Physica medica. Volume 32(2016)Supplement 3
• Page Start: 206
• Publication Date: 2016-09
• Subjects: Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2016.07.697
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract : Introduction: In the last decades the population dose increase is mainly due to CT. Recently iterative reconstruction algorithms were introduced to obtain dose reduction without compromising the diagnostic performance. Purpose: To investigate the effect of the model-based IMR algorithm in terms of image quality and patient dose. Materials and methods: CATPHAN phantom was scanned varying CTDIvol values from 0.8 to 31.8ṁGy., FBP and IMR algorithms were used to reconstruct the images. Different IMR reconstruction curves were analysed: Routine, Soft and SharpPlus. The homogeneous insert was used to analyse noise in terms of standard deviation and NPS; Low Contrast Detectability with statistical method was investigated too. MTF was evaluated using a tungsten bead, TTF with a homemade phantom. Patient dose indexes have been collected and analysed. Results: The averaged noise reduction for IMR routine and soft images is 45% and 66% respectively; IMR noise reduction works better for highly noisy images. LCD curve are better with IMR especially for small inserts and low doses. An averaged improvement of 35%–45% of contrast detectability with different curves has observed for 5 mm detail; these results are compatible with a reduction of 50% of CTDIvol on patient scans. IMR significantly improved MTF: MTF 50 increases by 12%, 24%, 64% with Soft, Routine and SharpPlus algorithm respectively; TTF curves confirm this trend. Dose reduction of 25% in cerebral exams and 50% in thorax-abdominal scans is confirmed by median values of CTDIvol from patient data. Conclusion: IMR reduces noise, improves low contrast detectability and reduces dose.