Charles E. Willis, PhD, medical physicist with MD Anderson Cancer Center teamed up with MD Anderson's S. Jeff Shepard and John Romlein, clinical engineer with Xtria Healthcare Systems, to provide SCAR attendees with a comprehensive overview of quality control (QC) and quality assurance (QA) in the digital environment. Willis led with a discussion of QA in CR and DR, followed by Shepard's tips for selecting and maintaining displays. Finally, Romlein looked at QA issues from the teleradiology perspective.
QA in CR and DR
The first step in QA is to differentiate QA and QC, according to Willis. QA ensures maximum, consistent performance mandated by ACR standards and includes personnel and equipment; QC entails technical procedures that ensure a satisfactory product. In the digital environment, facilities need to consider the entire digital process, identify points where errors may occur and take action to avert and detect and correct errors. Willis walked participants through the process, identifying common errors throughout the DR/CR process.
The digital process begins when the patient arrives; identification and patient information such as allergies and pregnancy must be verified, and the tech must make sure the anatomy matches the request. The tech may err in the next step by misidentifying the patient to the imaging system, resulting in a broken studies or an orphan. The end-result is an unavailable image.
Procedures to remedy patient and exam identification errors include automated association such as a RIS interface, bar code scanner augmentation and DICOM worklist management. Willis cautioned that no remedy is perfect. For example, the DICOM worklist may not account for unscheduled exams or the tech may not notice if the bar code scanner does not catch a patient's medical record number.
The next round of potential errors occurs at the positioning level and includes mispositioning, movement, improper collimation and incorrect exam selection. QC should be performed at the acquisition device to identify positioning errors and complete repeat exams as necessary.
The next level of the QA plan involves reject/repeat analysis. Digital imaging sites need to develop a method for capturing rejected and repeated studies, collecting and analyzing the data, reporting results to management and staff, implementing training and sharing results with vendors.
Inadequate erasure of previous exams, artifacts, incorrect gain adjustment and inadequate correction of non-uniformity can occur as the image receptor captures the projection. Projection errors are generally detected at the acquisition station where they may be corrected; however, some errors necessitate a retake.
Willis stressed the importance of active QC countermeasures to avoid errors before they occur rather than correct them post-exam. Countermeasures include prophylactic erasure of all CR plates at the start of a shift, periodic checks of non-uniformity corrections, periodic gain recalibrations, periodic checks of phototimer calibration as well as cleaning, thorough acceptance testing and a comprehensive technique guide for techs.
Like in previous steps, the next phase of the digital process-rendering of the captured projection for viewing-is ripe for QA/QC. Potential problems include incorrect exposure field recognition, incorrect grayscale recognition and inappropriate histogram analysis-all errors that are typically corrected with a retake or re-processing. Willis noted that many digital vendors have incorporated methods for automating QC in the CR and DR solutions. For example, most systems provide an exposure indicator. "There is a documented tendency to overexpose CR and DR, but QC programs based on exposure indicators are successful," Willis said.
Vendors are increasingly integrating QC workstations into the acquisition device to further streamline digital workflow. Image processing can be modified at the QC workstation to correct for some errors. Other functions completed at the QC workstation include image rotation, annotations, demographic overlays and conjoining images.
A transmission failure may occur as the image is pushed to the archive; the study may be deleted from the local cache or information may be omitted from the transmitted image. Errors that may occur at the final step of image display for viewing include incorrect grayscale calibration, inadequate spatial resolution and incorrect or missing