X-ray computed tomography (CT) as a modality in molecular imaging provide anatomical information – either in diagnostic cases or monitoring morphological changes during follow-up examinations for patients. Depending on the specific case the for the admitted patient the CT imaging provides broad anatomical information in relation to tracer uptake, attenuation correction (AC) for the tracers administered or high-quality morphological information of the investigated tissue with or without iodinated contrast material. The image quality should be suited for the purpose of the scan i.e. a low dose CT scan without contrast enhancement is considered suitable for attenuation correction and for rough localisation but is insufficient for diagnostic purposes in most cases. Depending on the indication a high dose CT scan is performed with contrast enhancement when there is no recent sufficient scan available. The scanning protocols are in general similar to those used in the radiology department and images is often interpreted by radiologists together with a nuclear medicine physician. One limitation is that when it is possible to improve fusion with the PET image, the CT scan is often performed with shallow breathing rather than maximal inspiration and then afterwards a CT of the chest at maximal inspiration of the lung is carried out. At our institution, a PET/CT with a low dose CT scan only is generally performed in two situations: first, when a prior CT scan is available, and second, when the nuclear medicine examination is the main factor in staging or diagnosis and the purpose of the CT scan only serves to correct attenuation and provide anatomical localization. When a dual modality examination is performed at our institution, we take the clinical problem or diagnosis into consideration and plan the scans accordingly: If a diagnostic CT scan sufficient for the clinical indication has been performed within approximately one month - for some aggressive malignancies no more than two weeks - prior to the dual modality examination we perform a low dose CT scan for attenuation and localization purposes in most cases. For malignancies like ovarian cancer where peritoneal carcinomatosis are likely present and important to precisely localize a high dose CT scan with contrast enhancement is performed.

Dual modality imaging optimization considers not only the quality of the CT and Positron emission tomography (PET)/Single-photon emission tomography (SPECT) images, but also the quality of the combined examination, which is heavily influenced by the clinical question that needs to be answered. Since the diagnostic CT scan is carried out in a dual modality setting it frequently serves as the baseline scan for subsequent stand-alone CT examinations conducted for treatment evaluation and therefore it should be appropriate for this purpose. On top of this clinical optimization comes the more detailed radiation dose calculations and optimal CT quality for both high and low dose CT scans.1,2

One general challenge for imaging practices is managing a wide plethora of CT scanning protocols. Currently there is no current standard for a master protocol design and therefore a degree of variation is seen between scanning parameters for the same medical indications. This variation can be ascribed to differences in scanner fleet and local practice by the self-governing medical staff – i.e. radiologists, technologists, radiographers, and physicians. National implementation of COUNCIL DIRECTIVE 2013/59/EURATOM3- addresses the national standards for protection against exposure from ionizing radiation. While the inspection of quality assurance systems for protocols is mandatory, there is no legal demand for the review of CT protocols to be carried out in a specific manner or for the details of such a review to be specified. The advent of new technology, including tube current modulation (TCM), automatic exposure control (AEC), iterative reconstruction (IR) algorithms, deep learning (DL) and the introduction of artificial intelligence (AI) in pre- and post-processing, can lead to the optimization of image quality in various ways across CT scanners with disparate technical characteristics. This has made the standardization of quality metrics a challenging endeavor for a nuclear medicine department that houses two or more scanners from different vendors.4, 5, 6, 7, 8 The primary objective of technology advancement is to provide the best quality images (highest possible diagnostic confidence) at the lowest possible dose governed by the ALARA principle for dose optimization - technological advances from vendors are largely driven by the need to provide the best possible image quality with the lowest possible radiation dose. National Legal Guides covering CT acceptance and commissioning tests for CT systems requires annual testing such as modulation transfer function (MTF), variance in slice thickness, HU-values, noise variation and dose linearity for CT systems for baseline information regarding performance and quality control in regard to a set metric.9 The guideline does not prompt any proposed method for intersystem quality matching. Alas the information gathered has little clinical relevance in regard for testing clinical protocols where more variables are included – i.e. difference in reconstruction kernels or difference between systems’ automatic exposure control which are widely utilized in clinical protocols. The proposed methods are largely based on basic system quantities and more advanced features are not included. i.e. metrics for benchmarking CT models based on more advanced characteristic are not dictated by the governing offices and a lot of these new frontier technologies are black boxes for end users, which poses challenges when handling a broad scanner fleet across different departments. Verdun et. al. describe the daunting task of providing an overview of image quality considerations when handling CT benchmarking of clinical tasks,10 where others focus on bridging the distance between model observers and human observer studies11 in narrowing in on relevant parameters for clinical method validation.

IEC standard 61223-3-5:201912 proposes a method for evaluation AEC characteristics for the CT system – but not in relevance to a clinical protocol and hence there is a need for scrutinizing methods for inter-comparison of these clinical relevant metrics. The American Association of Physics in Medicine (AAPM) proposed a guideline for inspecting CT Protocols13 in a said manner as support for the quality assurance program. The purpose of the guidelines is to ensure that diagnostic information is balanced with applied radiation dose to the patient and that these protocols are under periodical review to warrant optimized use of the CT systems and their characteristics. AAPM also proposed several performance tests for operational performance in general for CT scanners in addition to a basic CT QC14 ranging from noise inspection, CT numbers, and physical condition to AEC and analysis of task-based detection. Winslow et al. have looked at different metrics for evaluating image quality across platforms15,16, while Szczykutowicz et al. has proposed a method for defining a master protocol design whilst discussing the validity of investigation of the AEC when applying protocol optimization in a clinical setting17, 18, 19, 20 for aligning scanner performance and general optimization of the relation between tube output, table speed, pitch and patients´ weight categories. Berta et. al. has proposed a method for comparing new protocols with current standards,21 while others have proposed model-observer based evaluation of image quality.22, 23, 24, 25 As for CT imaging in relations to the field of molecular imaging Bertolini et. al provided a review on current use of CT protocols in PET/CT provided information on different optimization strategies used in the literature.26 Depending on the clinical indication and local practice a radiologist might be involved in the description and hence integrated in the imaging protocol feedback.

This review outlines pertinent CT imaging parameters and explores a potential approach to establish a more effective method for conducting CT protocol reviews. A nuclear medicine department might have of a set of standard protocols supplied by vendors, which the clinical staff can use to improve the image quality and adjust the protocols to align with local clinical needs. Further and ongoing improvement hereof provides continues optimized diagnostic accuracy for your practice.