Quality control in sonography: Possibility of defining acceptability criteria

Recently, several authors and institutions proposed quality control procedures for 2-D B-mode sonographic image analysis. The aim of this study was to assess if it is possible to define a range of acceptability of the different parameters measured.
Phantoms for 2-D B-mode acquisitions were employed with 11 of the same model high-level ultrasound scanners. Uniformity, axial and lateral resolution, axial distance accuracy, geometric distortion, image contrast, dead zone and penetration depth were measured. All images were acquired by experienced operators and saved by frame grabbing the analog video output signal. A small subset of images was also saved directly in dicom 3.0 format on a magneto optical disk. Images were analysed with in-house developed software. Quality control test results for 2-D B-mode images were demonstrated to be related to the probe employed. We also found that the results were not dependant on the operator and type of imaging capture method used. By selecting data from the same type of scanner and dividing probes into two groups (convex medium-low frequency and linear medium-high frequency) it is possible to find a reasonably narrow range of variability for almost all of the parameters studied. Mean values of the different parameters measured, 95% confidence limits and proposed intervals of acceptability are presented.
Therefore, it is possible to establish range of acceptability for the results of a quality control test of an ultrasound scanner operated in 2-D B-mode given the type of scanner and the type of probe. This is particularly useful both for quality assurance and maintenance programmes.