SCIENTIFIC PAPER An open source automatic quality assurance (OSAQA) tool for the ACR MRI phantom Jidi Sun • Michael Barnes • Jason Dowling • Fred Menk • Peter Stanwell • Peter B. Greer Received: 21 June 2014 / Accepted: 6 November 2014 Ó Australasian College of Physical Scientists and Engineers in Medicine 2014 Abstract Routine quality assurance (QA) is necessary and essential to ensure MR scanner performance. This includes geometric distortion, slice positioning and thickness accu- racy, high contrast spatial resolution, intensity uniformity, ghosting artefact and low contrast object detectability. However, this manual process can be very time consuming. This paper describes the development and validation of an open source tool to automate the MR QA process, which aims to increase physicist efficiency, and improve the con- sistency of QA results by reducing human error. The OSAQA software was developed in Matlab and the source code is available for download from http://jidisun.wix.com/ osaqa-project/. During program execution QA results are logged for immediate review and are also exported to a spreadsheet for long-term machine performance reporting. For the automatic contrast QA test, a user specific contrast evaluation was designed to improve accuracy for individuals on different display monitors. American College of Radi- ology QA images were acquired over a period of 2 months to compare manual QA and the results from the proposed OSAQA software. OSAQA was found to significantly reduce the QA time from approximately 45 to 2 min. Both the manual and OSAQA results were found to agree with regard to the recommended criteria and the differences were insignificant compared to the criteria. The intensity homogeneity filter is necessary to obtain an image with acceptable quality and at the same time keeps the high contrast spatial resolution within the recommended crite- rion. The OSAQA tool has been validated on scanners with different field strengths and manufacturers. A number of suggestions have been made to improve both the phantom design and QA protocol in the future. Keywords Quality assurance Á ACR MRI phantom Á MR QA Á Magnetic resonance imaging Á Medical imaging Introduction Magnetic resonance (MR) images have superior soft tissue contrast over computed tomography images (CT), allowing clinicians to diagnose abnormal tissue with higher accuracy [1]. No ionising radiation is required during an MR scan. However, as a complex imaging modality a routine quality assurance (QA) is essential to validate MR image quality and monitor the performance of the scanner. Aspects including spatial distortion, image uniformity, high con- trast spatial resolution, slice profiles and image artefacts need to be quantified [2, 3] and the clinical scanner needs to achieve the acceptance criteria [3–5]. In addition to the vendor provided QA phantoms, sev- eral phantoms have been developed commercially to examine the general performance of MRI scanners. Examples include the ACR (American College of Radiol- ogy) MRI phantom [4, 5], MagIQ phantom (Leeds Test Objects Ltd., Boroughbridge, UK), ADNI (Alzheimer’s Disease Neuroimaging Initiative) phantom [6] and the Spin Safety Test Object (Spin Safety, Rennes, France). The ACR MRI phantom was purchased at our institution because it is a well-established MRI QA system [4]. It has J. Sun (&) Á F. Menk Á P. Stanwell Á P. B. Greer University of Newcastle, Newcastle, NSW, Australia e-mail: jidi.sun@uon.edu.au M. Barnes Á P. B. Greer Calvary Mater Newcastle, Newcastle, NSW, Australia J. Dowling CSIRO Australian E-Health Research Centre, Brisbane, QLD, Australia 123 Australas Phys Eng Sci Med DOI 10.1007/s13246-014-0311-8