Reply: SPECT/CT

REPLY: We thank Dr. Knoll and coworkers for their comments regarding our recent continuing education article on hybrid imaging with SPECT/CT (1). In their letter to the editor, they report on their experience with a new software tool for image fusion (Volumetrix Suite; GE Healthcare). In more than 50 patients examined with this new tool, Knoll et al. observed an improvement of the diagnostic information in more than 89% (50/56) of imaging studies. As outlined in our article, software-based image fusion is still attractive. It would be highly interesting to learn how this new software approach affects various imaging situations in SPECT. The authors are therefore encouraged to submit an original article reporting their results in more detail.

Current software algorithms already allow the highly accurate coregistration of anatomic and functional images, therefore representing a regular component in daily clinical practice, for example, for image-guided surgery or radiation treatment planning. Whereas software-based image fusion seems appropriate in certain clinical situations, we hypothesize that hardware image fusion using integrated SPECT/CT scanners outperforms software-based algorithms in several scenarios. It has been previously demonstrated that motion artifacts arising from separate acquisition of CT and SPECT may seriously affect the accuracy of image fusion in the thorax, abdomen, or pelvis (2,3). Typical functional images of the thorax or the abdomen contain insufficient anatomic landmarks for their correlation to anatomic reference points. Differences in patient positioning and respiratory motion make the correct alignment of anatomic and functional images from separate devices even more complicated. Because of these issues, software algorithms did not reach widespread clinical use for image coregistration of the abdomen or the thorax. In the future, however, this technology may play an important role in allowing correction of misregistration due to patient motion or breathing artifacts, which may also arise from integrated SPECT/CT.

Besides anatomic referencing, the added value of CT coregistration is also based on the attenuation correction capabilities of CT. Cardiac imaging poses a particular problem in attenuation correction because of respiratory and cardiac motion in the thorax. Individual CT-based attenuation correction of brain studies using SPECT may also lead to improved image quality and more accurate data evaluation. Furthermore, radionuclide treatment planning using attenuation correction of imaging data and assessment of organ or target volumes derived from simultaneously performed CT may be more accurate and potentially allows safe and effective therapy.

A similar discussion on the need for integrated hybrid scanners has already been raised after the introduction of hybrid PET/CT systems to clinical medicine. As indicated for PET/CT, image fusion is faster, more reliable, and more accurate using an integrated scanner than using separately performed imaging modalities (4). In addition to these technical issues, hybrid image acquisition of both modalities in a single clinical visit (1-stop) offers apparent logistic advantages and is obviously more comfortable for the patient. PET/CT scanners represent the imaging modality with the most rapid growth worldwide and play an increasing role in routine patient care, especially in oncologic applications. Yet, there is a lack of evidence that the same holds true for hybrid SPECT/CT systems. CT coregistration, however, has been recognized to result in higher specificity and sensitivity of scintigraphic imaging and to markedly reduce the number of indeterminate findings. The superiority of SPECT/CT over planar scintigrams or SPECT has been clearly demonstrated for imaging skeletal diseases, parathyroid adenomas, and neuroendocrine cancers and for mapping sentinel lymph nodes in various cancers (1). Studies demonstrating superiority in other clinical applications are lacking; however, pilot studies encourage the use of SPECT/CT in cardiac and neurologic imaging.

Regarding the growing number of studies demonstrating an added value of hybrid SPECT/CT over separately performed imaging modalities (1), it appears likely that this promising technique will gain an important role in clinical routine practice. The broad spectrum of existing SPECT tracers and their widespread availability suggests SPECT/CT as a complementary imaging modality to PET/CT procedures. In summary, we agree with Knoll and colleagues that advanced software-based coregistration procedures do have a legitimate relevance for image fusion, particularly if no hybrid technology is available. However, we believe that hardware-based hybrid acquisition offers several apparent advantages regarding accuracy, reliability, logistics, and comfort for the patient, which cannot be easily outweighed by software-based image fusion approaches.