TY - JOUR T1 - Tutorial for optimal utilization of SNMMI fully automated cloud-based PET Phantom Analysis Tool for PET/CT Quality Control JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 1195 LP - 1195 VL - 61 IS - supplement 1 AU - John Sunderland AU - Bonnie Clarke Y1 - 2020/05/01 UR - http://jnm.snmjournals.org/content/61/supplement_1/1195.abstract N2 - 1195Objectives: Recently, the SNMMI launched a fully-automated cloud-based phantom analysis tool (PAT) at http://www.snmmi.org/PAT, available freely to dues-paying members of SNMMI. Phantom analysis modules are available for the ACR PET phantom, the standard 30cm cylindrical PET uniform phantom, The NEMA Image Quality phantom, and the CTN oncology phantom. The primary purpose of the tool is to facilitate enhancement of scanner quality control programs by providing easy-to-use, robust, and reproducible quantitative tools to measure and document scanner performance. The primary objectives of this exhibit are to 1) provide a visual tutorial of how to access and use the software, 2) present phantom imaging protocols and describe image upload and analysis procedures 3) describe the quantitative measurements provided for each phantom, 4) review the report options. Methods: First, accessing the cloud-based software and the generation of an individual username and password for PAT are described. Available instructional material describing the phantom program and how to use the software and understand the results are presented. Second, detailed descriptions of phantom fill and acquisition instructions are presented for each phantom. One of the primary purposes of the analysis software to meet the needs of technologists and physicists generating data to fulfill PET/CT scanner accreditation requirements (i.e., ACR, Joint Commission). Fill and acquisition instructions are provided to optimally meet these needs. Alternative fill and acquisition combinations are described for use with other non-accreditation use-cases (i.e., low contrast lesion detectability, image quality, reconstruction optimization⋯). Image upload, radioactivity fill data input, and analysis initiation are described. Third, the quantitative output of each phantom analysis is described. Depending upon the phantom, reported results include calibration accuracy, noise, axial and radial clinical reconstructed resolution, recovery coefficient curves, or axial and radial uniformity. The methods used to generate each of the metrics from the image data are presented to provide background and context to the reports. Lesion detectability and low contrast resolution are two assessments necessary for Joint Commission accreditation and require expert reader interpretations, rather than automated analysis. A formalism using Likert scales for consistent reporting of these metrics are presented. Results: Fourth, available downloadable phantom analysis reports are described for each of the phantoms. PDF documents are generated with phantom-specific measurements, plots and screen shots of representative reconstructed slices. These are designed to present necessary information to meet accreditation requirements. An Excel spreadsheet is also generated for each phantom analysis, with a summary quantitative analysis. Finally, a Java script Object Notation (.json) file generated for all analyses is described. This file contains all data used in the analyses, including relevant DICOM header information, phantom fill data, feature coordinates, and quantitative image data calculations. Conclusion: Instructions for use of a fully automated, cloud-based, publicly available PET/CT phantom analysis tool is described. The primary purpose of the tool is to support more robust PET/CT scanner quality control programs and facilitate compliance with national accreditation programs. ER -