A combination of integrated circuit detectors and iterative reconstruction (IR) utilization in CT exams has proven effective at significantly reducing image noise in phantom subjects and could lead to less patient radiation exposure, according to results of a new study published in the journal Radiology.

While modern CT scanners are not particularly vulnerable to image noise within the detector system itself, factors such as patient obesity and substantial reductions in radiation dosages in typical patients can present problems for image quality.

The adoption of new integrated circuit detectors in favor of traditional direct circuit detector systems, as well as the growing prevalence of IR over filtered-back projection (FBP), could lead to reductions in overall image noise and patient exposure to radiation when used together, said lead author Ajit H. Goenka, MD, and his colleagues from the Mayo Clinic in Cleveland.

“While IR is replacing [FBP] as the standard of care in many areas of clinical practice, studies have not yet established whether the combination of IC with IR matches or improves on the diagnostic performance of a conventional [direct circuit] detector and FBP at standard and reduced radiation exposures for low-contrast lesion detection,” they wrote. “In abdominal imaging, the detection of low-contrast low-attenuation liver lesions (lesions that have a small attenuation difference and are hypo-attenuating relative to the liver background) is a critical diagnostic task that is highly sensitive to image noise.” 

For this reason, Goenka and his team set out to assess the differences between two techniques—discrete circuit detectors with FBP and integrated circuit detectors with IR—and their respective impacts on image noise, detectability and radiation exposure in CT of low-contrast, low-attenuation liver lesions in semi-anthropomorphic phantoms.

To do so, they conducted CT exams on a phantom subject with and without a fat-mimicking ring five centimeters thick containing liver inserts with four spherical lesions. They performed five scans each on two separate CT scanners, one equipped with a direct circuit detector and the other with an integrated circuit detector, at five different exposure settings. The resulting images were reconstructed using FBP for the direct circuit scanner and IR for the integrated circuit scanner, measuring image noise and lesion contrast-to-noise ratio. Additionally, the presence of lesions was evaluated by four radiologists using a five-point diagnostic confidence scale.

The researchers found that the combination of an integrated circuit detector and IR significantly reduced image noise and improved reader confidence in lesion detection, regardless of phantom size or exposure setting. No difference was found in the diagnostic abilities of either combination, however.

“The discrepancy between the effects on objective image quality metrics and low-contrast, low-attenuation liver lesion detectability emphasizes that task-specific measures are a critical feature in determining the clinical utility of newer noise reduction technologies,” the authors concluded. “[Low-contrast, low attenuation] lesion detection in patients with larger girths is more sensitive to reduced radiation exposure, notwithstanding the use of advanced noise reduction technologies.”