In 1992, the American College of Radiology (ACR) Committee on Standards and Accreditation for the Commission of Neuroradiology and Magnetic Resonance recognized that there was increasing variation in the quality of MR images across the country. Some third party payors came to a similar conclusion and attempted to limit reimbursement on the basis of field strength. However, the committee realized that there was more than one characteristic which impacted image quality.

With the approval of the ACR leadership, a new committee set to work on the MRI Accreditation Program. Using the ACR Standard for Performance of MRI as a basis, the committee developed a program which incorporated a review of the qualifications of radiologists (or other physicians interpreting MRI), MR scientists/medical physicists, technologists, clinical image quality, and quality control testing through the use of a standardized phantom test object. The accreditation program was designed to raise the awareness of areas which needed to be improved and to provide recognition of quality whole body services.

All ACR accreditation programs use their successful Mammography Accreditation Program as a template. Ultimately, the ACR goal is to have one data base for accreditation programs so that only one application is necessary and the site need only indicate which module is appropriate for a particular practice location. In the interim, while the software is still under development, the MRI Accreditation Program will be treated individually, as are the other ACR accreditation programs.

Committee on MRI accreditation

The MR Accreditation Committee consists of members with wide-ranging experience and expertise in MRI, representing both academic and community practice orientations. R. Edward Hendrick, PhD, from the University of Colorado, is an MR physicist who also has had extensive experience with the ACR Mammography Accreditation Program. Dr. Hendrick was named Chairman of the Physics Subcommittee, which was charged with developing a phantom test object which simulates the human head. Jeffrey Weinreb, MD, whose group at NYU has extensive experience using objective criteria to compare MR images, was named Chairman of the Clinical Image Evaluation Subcommittee. Members of both subcommittees were then asked to develop criteria which could be used to set reasonable expectations for the quality of magnetic resonance images across the country.

During development, the MR Accreditation Committee and the two subcommittees met several times a year. An extensive pilot program which included nearly 50 sites was conducted to test program instructions and accreditation criteria. The pilot included MR magnets ranging in field strength from 0.064 to 1.5 Tesla. Based on pilot findings, documents were refined and finally submitted to the ACR Council Steering Committee for review and comment in Fall, 1996. Program documents were adjusted based on the input from the Council Steering Committee and then sent to the ACR Executive Committee, which reviewed all the information and approved the MR Accreditation Program for implementation in November 1996.

Program overview

The MRI Accreditation Program looks at the practice of MRI, which is performed using general purpose whole body scanners. When evaluating a program for accreditation, specific clinical examinations and quality control data are collected. Sites are asked to send their best examinations for the head, cervical spine, lumbar spine, and knee because it is estimated that these examinations cover 90% of the requests for MRI. Quality control data is then collected using a head phantom test object.

The application process

The entry application requests credentialing and CME information about radiologists and/or physicians who provide MR interpretations, medical physicists, and technologists who work at the practice site, all of whom are key to the performance of MR imaging. Each whole body magnet in operation at the site must be tested for the accreditation program; the site must submit the required clinical images and the quality control data obtained through testing the phantom designed for the program. Specialty magnets at the site (for example, magnets used for knee or extremity examinations only) are not currently included. However, they must be identified so that they can later be part of the accreditation process, when a module for specialty magnets is developed.

After the entry application has been reviewed and approved, the ACR sends a full application which requests more detailed information about the site's general MR practice. The full application package also contains labels for clinical and phantom images for each whole body magnet. Additionally, the full application package contains instructions for obtaining clinical images, purchasing the phantom from the manufacturer, and for performing the phantom testing itself.

Phantom evaluation

After application acceptance, sites are given the information to order the phantom directly from the manufacturer using an ID number assigned by the ACR. The total cost of the phantom is $540, paid to the manufacturer. The ACR designed the phantom to create economical, standardized testing tool and receives no profit from the sale of the phantom. Each site is required to submit phantom images using either the ACR protocol for obtaining T1- and T2-weighted conventional spin echo images or the routine T1- and T2-weighted scan protocol for head examinations at the site. The phantom images and testing data are used to assess: 1) limiting high contrast spatial resolution, 2) slice thickness accuracy, 3) geometric accuracy, 4) signal uniformity, 5) T1 ghost ratio, and 6) low contrast detectability. If the site normally uses lower bandwidth techniques and filtering for routine clinical scanning, these techniques may be used for accreditation submissions. The images and tapes are then evaluated by ACR physicists who will perform the appropriate measurements using the data produced by phantom testing.

Clinical image and phantom image review are intended to complement each other, providing a comprehensive evaluation of the quality of MRI services. The criteria for evaluation are independent of field strength and can be applied uniformly so that the performance of each general purpose whole body magnet is measured against a single standard.

Clinical image evaluation

The application site must provide images from the four required clinical examinations for every whole body magnet at the practice location being evaluated. This is part of an accreditation process for general MRI services. Specialty magnets in operation at a site requesting review for accreditation are unable to qualify for accreditation at this time. However, accreditation modules for examinations performed on magnets reserved for specialty examinations will be considered in future program development; sites which have achieved accreditation for the practice of general MRI services using whole body magnets will be notified automatically when additional accreditation modules become available for their specialty magnets.

Sites must send their best clinical images for brain, cervical spine, lumbar spine, and knee examinations. These images will be evaluated for filming technique, anatomic detail, artifacts, contrast and spatial resolution, and exam identification. Clinical images must be acquired within one week of phantom testing so that both the images and data are collected within the same two-week time period.

Site accreditation

The ACR MR Accreditation Program is completely voluntary. All applicants must be able to meet the same criteria. Based on existing ACR standards, there are alternate pathways to meeting credentialing qualifications for radiologists, other medical specialty physicians, and technologists. Input of an MR scientist/medical physicist is recommended.

Each site that applies for MRI accreditation will receive a documented final report. The report will delineate the accreditation evaluation. Specific recommendations for improvement will be suggested in the final report, where appropriate. In addition, information on initiating an appeals mechanism or reapplying in order to correct a problem that has been identified is provided with the final report.

Accredited sites will be issued an accreditation certificate for display at the practice location. In addition, the practice name and location will be published in the ACR Bulletin and provided to the general consumer as requested.

During the three-year period for which accreditation is granted, the sites are obligated to report changes in personnel and equipment to the ACR. This is so the site may receive the instructions to maintain current accreditation status.

As of this writing, more than 1600 applications have been sent to MRI sites. Applications have been completed on more than 400 magnets and are currently undergoing review. Those interested in receiving an MR Accreditation application should write to the American College of Radiology (1891 Preston White Drive, Reston, VA 2019 l-4397) to request an application or call 800-227-5463.      AR