Thank you. This next module has to do with PET-CT, especially the low-dose CT image that you get with the cardiac PET-CT perfusion set of images. My name is Randall Thompson. I'm a cardiologist in Kansas City and a member of ASNIC. I have nothing to disclose. Our objectives today are to understand the differences between CT and line source PET attenuation and the diagnostic content of the relevant attenuation maps to understand the value of visually estimated coronary artery calcium scoring and to understand the limitations of the low-dose CT attenuation maps and the relevant cardiovascular diagnoses that can and should be made from them. Here's our outline. We're gonna talk about the introduction of low-dose CT and PET acquisition techniques, processing and quality control, some issues on interpretation and what to report. Cardiac PET is commonly performed on either a dedicated PET scanner or a hybrid PET-CT scanner. Both utilize attenuation correction. And in both cases, the attenuation map is acquired sequentially rather than concurrently with the emission images. PET does have rather robust attenuation correction. With SPECT, attenuation correction is optional. With PET, it's not optional, it's required. And so all machines have some method of attenuation correction with PET. The dedicated PET scanners use a PET source such as germanium for attenuation correction while the hybrid PET-CT scanners use a CT source. The CT scanner's capable usually of a fully diagnostic CT scan, at least with PET, most of these scanners have a functioning CT that could be used for a lot of other purposes. But a low-dose attenuation map is all that's required and radiation should be kept as low as reasonably achievable. Here's an example. On the left is a dedicated PET scanner that uses a PET source for the attenuation correction. On the right is a PET-CT scanner that uses a CT scanner for the attenuation source. As you see at the bottom of the slide is an example of the image that they achieved. This is the attenuation map that you get to review. As you notice, the dedicated PET source has a rather low-resolution tomogram and the CT scanner has a much higher-resolution tomogram. So these attenuation maps, both types should be inspected for diagnostic content. The PET source attenuation map is sometimes very low-resolution, but you can see a few items that are relevant such as a large peripheral fusion, large hiatal hernias, and sometimes lung cancers if the mass is big enough. On the other hand, considerable diagnostic information is available on low-dose, even low-dose chest CT attenuation map. In the United States, as of January of 2020, cardiac PET studies performed on a PET-CT scanner are reported with different CPT codes and have different reimbursement than cardiac PET studies performed on a dedicated PET scanner. Inspection of those CT images, by the way, is part of the work description for the CPT codes, so it's part of the work that you're supposed to do. This is the list of the new codes as of 2020. Looks rather complicated. No one goes through all of those. It's beyond the scope here, but they are separated into codes that are from a PET-CT scan, dedicated PET scanner versus a PET-CT scanner. This is the most common code we're talking about from a dedicated PET scan of this one, and I'll blow it up. This is code 78492, myocardial imaging, positron emission, tomography, perfusion study, including ventricular wall motion, ejection fraction when performed, multiple studies at rest or stress. So this is the most common PET, myocardial perfusion imaging study on a dedicated PET scanner. On the other hand, we have a different code, a different way of reporting when you perform that service on a PET-CT scanner. That's this code, 78431, myocardial imaging, positron emission, tomography, perfusion study, including ventricular wall motion and or ejection fraction when performed, multiple studies at rest and stress, exercise or pharmacologic, with concurrent acquired computer tomography transmission scan. In other words, this is a myocardial perfusion imaging study performed on a PET-CT scan. And there's similar dual codes for metabolism, for a single myocardial perfusion study, and for a combined metabolism perfusion study, which is done, for example, for viability or for imaging sarcoid. CT attenuation maps should be acquired in a low radiation dose setting, usually eight to 25 milliamps with a mid-breath hold or free breathing with arms raised or rarely by the sides in the same position as the PET emission acquisition. So if you think about it, you want your attenuation scan to be as close as possible to the acquisition of your emission scan. In other words, if someone takes a deep breath during the CT portion, and they're doing free breathing or light breathing during the PET scan, it'll throw off the attenuation map, you'll get misregistration. Likewise, if the arms are usually above the head, occasionally we have to put one or two arms down by the sides. In either case, you should have the arms in the same position on the PET scan and the CT transmission scan. The manufacturer's default settings may be much higher than this low radiation dose, and you may need help from the application specialist or physicist to set the proper acquisition parameters to as low as reasonably achievable. And for quality control, the interpretive physician should indeed inspect the merged emission transmission images in appropriate quality control software. This was outlined in a different module. Artifact on PET studies related to the beam hardening, for example, from a metal artifact in a shoulder or pacemakers or metal hardware of the spine has been an issue in the past. And it's possible to get beam hardening artifact today, but newer reconstruction techniques, especially with the inter reconstruction made this much less of a problem than it was in the past. So you may read about this, but it turns out to be much less of a problem than it used to be. Here's a common PET CT rubidium perfusion acquisition. If you can start on the left, you'll see that we start with the CT scout scan, and then we give the radio tracer, then we acquire the rubidium emission images, and then perform a CT for attenuation correction. Now, rubidium has a short half-life, and so within a few minutes, you can give your pharmacologic stress agent, usually regadensin or dipyridamole, rarely adenosine. And during stress, the radio tracer is again given, and the rubidium stress emission scan is acquired. And then another CT attenuation scan is obtained, and this is all done fairly quickly. Now, in our laboratory, we routinely perform a separate acquisition on the CT scanner for calcium scoring. This is a gated acquisition at a set dosage, 120 kV. And to allow a lot of calcium to be scored in the fashion that's been done for a number of years, so that it's comparable to other coronary calcium scores. This takes an extra 20 seconds. So as you see from this, it really is one's able to obtain a rest, stress, PET, myocardial perfusion image study quite quickly, really within half an hour. In these days of COVID and social distancing, that's a separate advantage that we don't think about that much. The patient can get in, get the study done and get out, and there's less contact with other people. I'll give some examples of the occasional diagnostic information that can be obtained on the PET source from a cardiac PET studies. These are uncommon, but for example, here's a PET transmission map, and you see in these arrows, this is a large lung mass. Now, this was known before the patient had the scan, but you couldn't make it out on these low resolution images. Here's another set, patient has something very big underneath the heart. This is a huge hiatal hernia. We know that because the patient also had a CT scan in the same setting, and here's just this massive hiatal hernia compressed in the heart, probably causing the symptoms that generated the request for the stress test in the first place. When you review the quality control, look for proper registration, look for the presence of coronary and thoracic aortic calcifications, check for cardiac chamber sizes, pulmonary artery size and aortic aneurysm, and then check for important pulmonary and soft tissue abnormalities. It's important when you're getting started to have a checklist, although with a lot of practice, most reviewers can internalize that. It is important to have a systematic approach though. Let me talk a minute about adding a coronary calcium score to a myocardial diffusion imaging. It's easier if one has a PET-CT instrumentation. So this is a patient that has a nice normal myocardial diffusion imaging study. We would generally tell this patient the prognosis is good, they did not have obstructive coronary artery disease, and we would give them reassurance. If we obtain a CT coronary calcium score in the same setting, sometimes we get very different and very helpful information. So for example, this patient has a very high coronary calcium score. We tell them something different. We say, well, you have no obstructive coronary lesions that we can detect on a myocardial diffusion imaging study, but you do have a lot of preclinical atherosclerosis. This patient would be referred for a different follow-up algorithm for aggressive risk factor modification. And if they had a change in symptoms, they would be reevaluated more quickly than if it was the other way, if they did not have significant coronary calcium. So in my laboratory, and a number of laboratories around the United States, we recommend performing a coronary calcium score when possible in all diagnostic PET scans, or even all diagnostic myocardial diffusion imaging studies, either as a separate acquisition, or in this case with a PET-CT scanner, just the extra minute or so of time and a small amount of extra radiation. Here's one of the many pieces of data showing the powerful prognostic value of coronary calcium scoring. This is a well-traveled slide, perhaps the most famous slide in this field. But as you see, as a patient's coronary calcium score goes up, their risk of death over subsequent several years increases. So the more calcium you have, the more atherosclerosis you have, and the greater the risk of death. Thus, we believe that by adding this to your myocardial diffusion imaging study, you gain a lot in the diagnostic study. In fact, that's been shown. Now, for example, this one of a number of studies showing the combined quantitative assessment of myocardial diffusion in coronary calcium with a hybrid PET-CT scanner improves the detection of coronary artery disease. In this study, the area under the curve for the combined seen in red is greater, so one does pick up more important disease. Also, quite frankly, one subtly changes the interpretation. If you see a lot of coronary calcium, you tend to read for sensitivity. If you see none, you tend to read for specificity. That probably improves the accuracy of the read, even without the subjective additional measurement. Now you can train your eye to look at the attenuation maps to estimate the coronary calcium. You might do that for cost reasons because some CT scanners are not equipped with coronary calcium scoring or because the extra cost is not supported by insurance companies or in certain circumstances. This was a study that we performed a number of years ago and it showed that you can train your eye to estimate the calcium scoring from the low-dose attenuation maps, even the lower dose. Now this attenuation map shows a rather obvious coronary artery calcifications. Incorporating this information improved the diagnostic accuracy of the study. You can see these coronary calcifications here and here and here. This is a couple of paired images. On the left are the low-dose CT attenuation maps. On the right are the proper coronary calcium scores. On the top is a patient that has a high coronary calcium score, greater than 400. On the bottom is one that has very little coronary calcium, less than 10. Your eye can obviously see that pretty well. This is the output slide from that study. If you look along the diagonal, this is the diagonal line of identity between the visually estimated coronary calcium and the Agatsin score on the proper calcium score study. And most of the patients fell on that line of, this vertical line. But more patients fell below the line than above the line. So your eye is not totally sensitive enough sometimes. Sometimes it's hard to see the calcium if the study is non-gated or if the patient's obese. There's an improvement in the diagnostic and prognostic content of the test when the MPI is combined with the coronary artery calcium score and we've said for a number of years that it's probably likewise true when the coronary artery calcium seen on the CT attenuation map is inspected and incorporated. Indeed, some people have shown this. Now, this is a very nice study from Intermountain Health. They looked and recorded whether calcium on a PET-CT scan looking at the attenuation map was visible or not, and they incorporated that into their decision-making. They had a very large study, over 11,000 patients. After all the exclusions, they had over 3,000 patients in their calcium scoring absent group and over 2,500 in the calcium present group. The top is the percent free of event where the calcium score was absent or present. Obviously, what we'd expect based on the coronary artery calcium score. In the bottom left are their number of patients, the percentage of patients that had coronary angiography, high-grade obstruction, revascularization based on whether the visually estimated calcium score was present or not. I'll blow this up a little bit more. As you see, when calcium was not present or visibly present in these scans, the odds of a patient having coronary revascularization was very low. I was reviewing the study and writing an editorial on it, and that very next day, I had a patient who had both the coronary calcium score and a CT attenuation scan. The patient had had a coronary calcium score and it was 121. The patient had symptoms, went on to have a microdiffusion imaging study with a scanner that did not have calcium capability. It was a non-gated study, and you can't see the calcium in this patient. Now, this is sort of the extreme case. The attenuation map here is done at very low radiation, 20 kV, and the patient's morbidly obese and it's non-gated. So at least in this extreme circumstance, you might not be able to see the calcium. If you see it, you definitely can count on it being atherosclerosis. If you don't see it in the patient's obese, it might be just that there's not that much calcium. It's hard to see. There are also a number of relevant cardiovascular diagnoses on the CT transmission map. I've listed them here. Many of these have to do with why the patient may have had the test ordered in the first place or they're highly relevant medical diagnoses that a cardiologist or a cardiac imager should make, including coronary calcium, thoracic aortic calcifications, that would diagnose atherosclerosis, aortic aneurysm, pericardial effusions, pericardial calcifications, pleural effusions, paralyzed diaphragms, coronary anomaly, and you can read the rest. There, a physician reviewing these scans should inspect these. In fact, it's even part of our American CPT codes for that. Here's some examples. Most of these were discovered incidentally. This patient had a normal myocardial effusion. I'm really excited, but he does have a large ascending aortic aneurysm. This is nearing the size where surgery is necessary and should be attended to. Here's a patient who has a pericardial effusion and a pleural effusion. The arrow points to the pericardial effusion. This was an interesting case. This patient was being evaluated for kidney transplant, but was not yet on dialysis, came for a pre-transplant evaluation, did have some symptoms, shortness of breath, and this pericardial effusion was diagnosed on the CT attenuation map of a myocardial effusion imaging PET CT scan. The MPI was negative, but the patient went from the imaging suite to the dialysis suite for urgent dialysis or his eugenic pericarditis. Now, this patient has pericardial constriction, at least, excuse me, pericardial calcifications, also relevant, may have something to do with why the test was ordered in the first place. Here's a patient with an atrial septal closure device. Here's a patient with a moderate-sized hiatal hernia. Here's a patient where I believe the cause of the chest pain was on the CT attenuation scan. Here's a pill that's lodged in the crypts of a large hiatal hernia. These hurt, pills that are stuck in the esophagus hurt a lot. That's probably why the patient had chest pain. The coronary calcium score was zero, and the MPI study was negative. Here's a patient that has gallstones, also possibly relevant. Here's a patient who has breast cancer. Here's where cardiologists get nervous. We're not really trained to look outside the chest or the heart, so we need to develop algorithms where cardiologists, after reading the scans, become accustomed to looking for non-cardiovascular sorts of things, or have regular or even routine ovaries. Here's a patient with interstitial lung disease presenting with shortness of breath. Interstitial lung disease was more severe than appreciated initially by the clinician. Here's a patient with a large pleural effusion. Here's a patient with heavy coronary calcifications and pleural effusions. And here's a patient with, you notice the liver is too high on the right compared to the left. If you look at the scalp film, you see the elevation of the right hemidiaphragm. Interestingly, you would think that a patient coming through an internist or a cardiologist office would have had a chest X-ray most of the time, but not always. Sometimes we pick up this abnormality on the CT portion. Here's an interesting patient that I saw recently, an 82-year-old who had chest pain, but also a cough. He had a new diagnosis of cavitary lung mass made. Obviously, a major finding that you wouldn't want to miss, so you do need to inspect these CT scans, of course. In this era of COVID, we need to look for typical ground glass appearances. We do have patients who don't know they have COVID who come in for these studies. One should look at the CT attenuation scan before the patient leaves the facility, ideally, or certainly soon. This patient had the diagnosis of COVID made from the scan. Indeed, was having a cough while we caught him before he could expose too many other people. Here's a 71-year-old patient who had prior echocardiograms, which are normal, and has a large coronary sinus. It helps to know the anatomy. This is the coronary sinus draining across into the right atrium. And this was, if you track it up, patent left superior vena cava. An incidental finding, but has some clinical relevance. So a cardiologist could, can, and should make these mostly cardiovascular diagnoses. A systematic approach is important. And ASNIC, there are no current guidelines or policy statements on this, although ASNIC has a couple in the works, one that I expect to be published fairly soon regarding the issues of training and what should be reported in the systematic approach. Now, there are limitations of the CT attenuation map. As you saw, this is not a full CT scan, although it has a lot of diagnostic information. It does not cover the whole chest. You can't see small nodules. You don't see the entire lung fields. You can estimate the coronary calcium, but most people prefer a number. It helps to have a proper number. It's sometimes difficult to distinguish no calcium from mild coronary calcium on these low-dose attenuation maps, especially when they're non-gated. And I frequently suggest a full non-contrast chest CT scan. Now, I look at these myself. If there's a nodule or some other abnormality, a proper full chest CT scan also gets the patient with the radiologic follow-up if that's necessary, and the report goes in the usual systematic manner. So reporting a scan, the cardiac PET-CT scan, the usual rules for reporting apply. Plus, you should include in the CT review. You should mention it in the body of the report, and if they are important findings, you should mention that in the conclusions. The cardiac PET-CT, CPT code, billing codes, their work descriptors do include a view of the CT attenuation map, so it's part of what one is getting paid for, and the relevant CT findings deserve to be mentioned in the study summary or inclusion. For example, you might report that the myocardial diffusion imaging study is normal, the left ventricular systolic function is normal, and item number three, the limited field of view, CT attenuation map did demonstrate substantial coronary calcifications. Now, another question comes up. If you have a coronary calcium score at a PET study, should you have one report or two? And I would say from personal experience, the definition of one, and the reason is that we all have electronic medical records these days. If you have two reports, they're gonna appear in the referring physician's inbox at different times invariably, sometimes even further apart than you would think, a day apart even, and that leads to confusion. One report integrating both avoids that confusion. So the key teaching points to review what we've touched on is, number one, there's important diagnostic content in the CT attenuation map of a PET CT scan. More than a dozen cardiovascular diagnoses can and should be made from reviewing those scans. You can train your eye to estimate the coronary calcium and it's helpful to do so. These CT attenuation scans are not diagnostic chest CT scans and frequently a diagnostic proper chest CT scan may need to be ordered separately. Thank you very much for your attention.

CT attenuation map

cardiac PET-CT imaging

differences between CT and line source PET attenuation

diagnostic content of attenuation maps

visually estimated coronary artery calcium scoring

limitations of low-dose CT attenuation maps