The prototype combined PET/CT scanner was developed from 1995 to 1998 at the University of Pittsburgh. After being installed there in 1998, approximately 330 patients were evaluated with the device over a 3-year period. In mid 2001, the Food and Drug Administration approved the clinical use of PET/CT, primarily based on the favorable data derived from these initial patient investigations. Since FDA approval, more than 1,000 scanners have been purchased and implemented in centers worldwide, with approximately 95% to 98% of all PET scanners in the last year being sold as part of combined PET/CT scanners. The overwhelming market forces and the large decrease in demand for dedicated PET scanners will likely make PET/CT the dominant imaging modality for oncology.


There are many reasons why combining PET with CT rather than PET with any other anatomical imaging modality became the ultimate choice for a fusion instrument. Although a combined PET/MR device could in theory be designed, it made more logistical sense to combine PET with CT, since approximately 90% of oncology patients are at some point evaluated with CT. However, there are many more advantages of combining PET with CT, including the use of CT for attenuation correction (AC). By using the CT portion of the examination for AC, it obviates the need to perform a separate transmission scan, ultimately reducing the time it takes to perform a whole-body PET/CT scan significantly compared to dedicated PET alone (by approximately 15 to 25 minutes). In addition, PET/CT scanners were designed with a large enough patient port to accommodate a flat radiation therapy planning palate, making it possible to develop an integrated anatomical and functional PET/CT-based radiation

treatment plan, rather than the traditional anatomical based plan (CT) that has been employed throughout the years prior to PET/CT. There is also improved patient convenience by consolidating the usual imaging workup of an oncology patient (PET and CT) into a single examination.


Although there are many benefits of combining the two modalities, a number of logistical issues were not well addressed prior to implementing the modality into clinical practice, the most important being that a new “fusion” modality requires expertise in both cross-sectional CT, as well as PET imaging. The majority of PET studies were read by nuclear medicine physicians; however, in the early clinical use of PET/CT, there were very few interpreting physicians who were trained in both radiology and nuclear medicine or PET. Centers performing PET/CT have dealt with this problem in a number of ways. The centers that had the rare dually trained physician were in the best position and typically had those interpreting physicians read their scans. Centers without a dually trained physician have often relied on a radiologist to read the CT portion of the examination, while a nuclear medicine physician would read the PET portion of the examination. This method is acceptable as long as there is corroboration by the interpreting physicians as to the final interpretation. In some centers, PET/CT has been interpreted solely by a nuclear medicine physician who will typically use the CT portion of the examination for AC and localization and issue a report disclaimer that the CT portion of the examination is not being interpreted for diagnostic purposes. This last scenario is suboptimal for a number of reasons, the most important being that most relatively or non-FDG avid tumors will not show up on the PET portion of a PET/CT, while there is often enough information to make the diagnosis on the CT portion of the examination, regardless of the technique used. The general consensus in the radiology community is that there is poor quality and good quality CT, but if there is enough information on the CT to make the diagnosis, it is, by definition, a diagnostic CT. Nuclear medicine physicians have two options in this situation; have a radiologist look at the CT portion of the examination, or do not archive the CT portion of the examination, because it is unlikely that a disclaimer will protect them from not detecting a large non-FDG avid lung cancer that is obviously present on the CT.

Figure 1. A snapshot from the interactive PET/CT atlas shows axial and coronal PET/CT fused images with a cursor pointing to the area of FDG uptake in the bilateral supraclavicular areas corresponding to brown fat. Hovering over the structure displays is the name of the structure in gray. Clicking on the structure brings up a dialogue box explaining the typical FDG uptake patterns observed with the structure. Along the bottom part of the window is the gallery, which provides additional examples of atypical patterns and variant patterns of FDG uptake within the selected structure. Double clicking on the gallery images displays an enlarged image with an explanation of the additional case example. (Click the image for a larger version.)


Given all of these considerations and the rapid increase in popularity of combined PET/CT imaging, it is understandable why there has been an urgent shift in focus toward determining appropriate training requirements for physicians interpreting PET/CT. The Society of Nuclear Medicine (SNM) and American College of Radiology (ACR) recently released joint recommendations for training of interpreting physicians, with the goal of providing reasonable training recommendations that would enable both radiologists and nuclear medicine physicians to interpret an entire PET/CT study. The training requirements are based on the background training achieved and the physician’s level of experience. The joint recommendations have formulated a schedule of supervised scans and CME activities that serve as the basis of the training recommendations. Although no time frame was given for when the recommendations would become requirements or whether they would be enforced, in general, past recommendations from the ACR have become standards and often requirements in the field of radiology.


The release of training recommendations is important for the continued success of combined PET/CT. However, currently there are no training programs available to interpreting physicians that could fulfill all of the proposed CME and supervised scan recommendations as a single comprehensive course. There are many preceptorship programs available throughout the United States that can provide portions of the recommendations, and combined with CME credits obtained from a PET and/or PET/CT seminar, interpreting physicians can attain the recommendations. However, other educational materials are certainly needed. With this need in mind, the interactive PET/CT atlas was developed.


Figure 2. A snapshot shows an example of additional case examples that are displayed when participants click on the image gallery in the bottom of the atlas window. (Click the image for a larger version.)

The interactive PET/CT atlas is an educational activity for PET/CT interpreting physicians geared toward providing training in cross-sectional anatomy as well as education in FDG PET and PET/CT imaging. The atlas includes images from an entire patient dataset, with PET, CT, and fused PET/CT images available in axial and coronal orthogonal planes. One of the primary functions of the atlas is to provide cross-sectional CT images labeled with anatomical structures important in oncologic imaging. When participants use the mouse to hover over a structure, the name of the structure is displayed. Nuclear medicine physicians will find this feature useful when reading PET/CT. If they encounter a structure they are not familiar with, they can use the atlas to point to the correlative structure on the atlas and the name will be displayed for them. Due to the limitations of using a 512 x 512 pixel image set, not all structures in the body are labeled.

If the participant left clicks on the structure, the typical normal and variant FDG uptake patterns are provided (Figure 1). In addition, important citations related to the FDG uptake patterns are provided for further reference. Many structures also have additional examples of variant or atypical patterns of physiologic, altered physiologic, and benign causes of FDG uptake provided in the gallery section of the atlas (Figure 2).

Figure 3. A snapshot from an image-based question from the final examination. The quizzes are broken down into head and neck, chest, abdomen, and pelvis. (Click the image for a larger version.)

Quizzes are also provided for self-assessment (Figure 3). After interacting with the different regions of the body, there are interactive quizzes, divided up by body region. The four sections include quizzes on the head and neck, chest, abdomen, and pelvis. A final examination is also provided for the participant, which includes 70 image-based questions.

Future versions of the interactive PET/CT atlas include more detailed anatomy, organ-based anatomical/functional modules, additional whole body patient data sets, and quizzes. The atlas is sold through, an online CME provider.

Todd M. Blodgett, MD, is clinical intructor, University of Pittsburgh Department of Radiology, where he is completing an oncologic imaging fellowship.