Image quality standards in digital mammography: A consensus statement

Evaluation of image quality in screening mammograms can suffer from significant inter- and intra-observer variability leading to inconsistent quality assessment. The purpose of this study is to establish a standardized framework for describing image quality, thereby reducing subjectivity and enhancing observer consistency.

A scoping literature review was conducted to compile a preliminary list of 19 image quality criteria. This preliminary list was evaluated by a multidisciplinary panel of 23 experts through a modified Delphi survey. Experts ranked each criterion based on its "importance for image quality" and "agreement with the phrasing." Criteria for which consensus was not initially achieved were subsequently revised through a focus group discussion.

The panel reached consensus on image quality criteria regarding image identification; artefacts; uniformity; symmetry; contrast; radiation dose; sharpness; fibroglandular triangle depiction; nipple-pectoralis distance; depiction of posterior, medial and lateral breast regions in craniocaudal views; and depiction of the pectoralis muscle, inframammary fold and superior breast regions in mediolateral views. However, additional attention was asked to provide clear guidelines on additional, alternative images in screening settings; defining a proper description of good, suboptimal and insufficient compression force; and providing detailed advice on which folds and air projections warrant a retake.

This study resulted in a checklist for mammogram image quality assessment, along with cut-off values for good, suboptimal, and insufficient quality, with the clinical relevance of each quality criterion highlighted.

A standardized image quality assessment tool is offered to improve the consistency of mammographic image quality interpretation in screening programmes.