Impact of model-based iterative reconstruction on low-contrast lesion detection and image quality in abdominal CT: a 12-reader-based comparative phantom study with filtered back projection at different tube voltages

Abstract

• MBIR improved quantitative image quality but not lesion detection compared with FBP. • Increased CNR by low tube voltages did not improve lesion detection. • Changes in image noise and CNR do not directly influence diagnostic accuracy.

Effects of various generations of iterative CT reconstruction algorithms on low-contrast detectability as a function of the effective abdominal diameter: A quantitative task-based phantom study

“…This is the case when comparing different CT systems, or when comparing various image reconstruction algorithms on the same CT scanner. Our results confirmed the findings of previous human subjective analyses of statistical IR (ASiR and SA-FIRE) [26,27] and partial model-based IR (ASiR-V and ADMIRE) [25,28] using an objective methodology based on a task-specific concept.…”

“…Other investigators have also used CT phantoms to assess the potential improvement of LCD with various IR algorithms. Subjective analyses with human readers (ROC, 4AFC experiment) have indicated that IR techniques (SAFIRE, ASiR, and AIDR) do not statistically improve LCD [9,[23][24][25] or only slightly improve LCD at the same dose level with various IR algorithms [26][27][28]. These researchers have also applied quantitative image quality parameters (noise and CNR) to validate their use.…”

3D printing of anatomically realistic phantoms with detection tasks to assess the diagnostic performance of CT images

“…The observed contrast increase at a lower tube voltage was reinforced by the iodine content of the phantoms. The results are in line with previous observations in phantoms and patients after contrast medium administration [12][13][14], underlining that the phantoms simulate the clinical situation adequately. Contrast increase at reduced tube voltage is less pronounced for soft tissues without contrast medium enhancement [15], which may thus limit the suitability of the phantoms for studying tube voltage effects in situations where CT scans are acquired without contrast medium administration.…”

Visual grading evaluation of reconstruction methods and dose optimisation in abdominal Computed Tomography

“…Since the non-linear effects of IR lead to smoothing of the anatomical features and change in appearance of the anatomy in the images, some loss of information is observed when applying higher strengths of the algorithm [92,93]. Low-contrast detectability is compromised when using higher strengths of ADMIRE [94] as was demonstrated by the inferior image quality for criterium "liver parenchyma" in Papers II and IV when comparing the two strengths of the algorithm. Although some improvement in image quality was seen with thinner slices of the higher strength, a general recommendation for ADMIRE 5 in clinical practice was not indicated (Paper IV).…”