Development and validation of image quality scoring criteria (IQSC) for pediatric CT: a preliminary study

Abstract

Objective To develop and assess the value and limitations of an image quality scoring criteria (IQSC) for pediatric CT exams. Methods IQSC was developed for subjective assessment of image quality using the scoring scale from 0 to 4, with 0 indicating desired anatomy or features not seen, 3 for adequate image quality, and 4 depicting higher than needed image quality. Pediatric CT examinations from 30 separate patients were selected, five each for routine chest, routine abdomen, kidney stone, appendicitis, craniosynostosis, and ventriculoperitoneal (VP) shunt. Five board-certified pediatric radiologists independently performed image quality evaluation using the proposed IQSC. The kappa statistics were used to assess the interobserver variability. Results All five radiologists gave a score of 3 to two-third (67%) of all CT exams, followed by a score of 4 for 29% of CT exams, and 2 for 4% exams. The median image quality scores for all exams were 3 and the interobserver agreement among five readers (acceptable image quality [scores 3 or 4] vs sub-optimal image quality ([scores 1 and 2]) was moderate to very good (kappa 0.4–1). For all five radiologists, the lesion detection was adequate for all CT exams. Conclusions The image quality scoring criteria covering routine and some clinical indication-based imaging scenarios for pediatric CT examinations has potential to offer a simple and practical tool for assessing image quality with a reasonable degree of interobserver agreement. A more extensive and multi-centric study is recommended to establish wider usefulness of these criteria.

Developing and implementing a multi-modality imaging optimization study in paediatric radiology: Experience and recommendations from an IAEA coordinated research project

“…In addition to an additional IQ level clearly and unambiguously indicating images having been acquired with better-thannecessary IQ for future optimization studies of this nature, greater standardisation of subjective image quality requirements would be of great value. In this sense the use of a 5-level scale, as proposed in the literature for relevant studies [33,34] would be beneficial as it can better demonstrate the quality of the acquired images, including those with image quality unnecessary high for the diagnostic task. This would require the commitment of the clinical staff participating in this optimization exercise, since it would increase the complexity and thus the time required for the evaluation of the images.…”

Development of image quality related reference doses called acceptable quality doses (AQD) in paediatric CT exams in Qatar

“…We did not have direct automatic access to equivalent body diameter or water equivalent diameter or size-specific dose estimates (SSDE) which is the case in most countries and thus weight is still the most convenient parameter till such time electronic estimates of It should be understood that the score of 4 depends upon training and orientation of the readers (radiologists). In good centres, where awareness is created about acceptance of images with some noise rather than crisp images, the radiologists can discern images of higher than necessary quality [11,16]. Thus, the score of 4 may be assigned more often by dose conscious reader than others.…”

Higher patient doses through X-ray imaging procedures

“…In addition, the MDCT's use of narrow collimation caused a decrease in geometric efficiency and thus an increase in image noise and was countered through the use of increased dose [19]. Use of higher than needed image quality, with a preference for crisp images rather than those with some noise present, has also contributed to the higher CT doses [9,20]. As such, radiologists must be willing to use images with some noise (salt and pepper appearance) [20].…”

“…Use of higher than needed image quality, with a preference for crisp images rather than those with some noise present, has also contributed to the higher CT doses [9,20]. As such, radiologists must be willing to use images with some noise (salt and pepper appearance) [20].…”

“…Unfortunately, cases of variable and higher exposure have not been confined to abnormal patient attenuation or nonstandard cases. A recent study [20] examining image quality among the most frequent pediatric CT exams found that 16-40% of the CT images assessed were of higher than necessary quality acquired with higher settings (kVp, mAs) leading to higher radiation dose for adequate diagnostic interpretation. Similar issues have also been identified in other studies [68][69][70] and in fluoroscopy.…”