Evaluation of image quality and patient exposure in fluoroscopy using a phantom: Is there any clinical relevance?

• Published in: European journal of radiology Vol. 138; p. 109607
• Main Authors: Tsalafoutas, I.A., Tsapaki, V., Triantopoulou, I.
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01.05.2021
• Subjects: Digital fluoroscopy; Fluoroscopy; Humans; Image artifacts; Image processing algorithms; Image Processing, Computer-Assisted; Image quality; Phantoms, Imaging; Radiation Dosage; Radiation Exposure; Image quality; Radiation exposure; Image artifacts; Image processing algorithms; Digital fluoroscopy
• ISSN: 0720-048X; 1872-7727
• DOI: 10.1016/j.ejrad.2021.109607

•Modern X-ray systems have numerous X-ray preset protocols; some are a “black box”.•X-ray system manufacturers offer more radiation optimization tools than in the past.•Protocol selection has large impact on image quality and patient radiation dose.•There are limitations on how software reproduces visual image quality evaluations.•Performance of protocols under clinical conditions cannot be assessed by phantoms. To investigate the impact of X-ray preset acquisition protocol settings on fluoroscopy image quality (IQ) and radiation exposure. A quality control (QC) phantom was imaged with a modern digital C-arm system, using various preset fluoroscopy protocols. IQ was assessed using human observers and in-house software for automated evaluation, based on contrast-to-noise ratios of details and their background. Patient radiation exposure was evaluated using the displayed Incident Air-Kerma and Kerma-Area Product values. Protocol selection affects radiation exposure by a factor of about 3. IQ evaluation showed that acquisition protocols produce images with quite different characteristics. The visual IQ evaluation method was time consuming and cumbersome. The automated method, utilized the visual IQ evaluation results for calibration of detection thresholds. However, it failed to reproduce these results for all images and details types. In some images, digital image processing created artifacts which affected the pixel value distributions around details in a way that could be handled only by the human vision. Manufacturers provide many preset protocols designated for specific clinical uses, which have large impact on IQ characteristics and radiation exposure. However, protocol settings’ selection rationale is essentially a “black box” for the end user. Though QC phantoms are currently used for IQ evaluation, they are not appropriate for drawing firm conclusions concerning the expected performance of each protocol in clinical practice. Currently, there is no consensus on the optimum technical characteristics of preset protocols for specific procedures. More work is needed in this area.