Cardiac CT Scan - Anatomy and Basics of Cardiac imaging

Cardiac CT is one of six modalities we commonly use to image the heart. Each of these 6 modalities, including cardiac CT, has different sets of use cases, pros & cons. Echocardiography is much more widely available than cardiac CT and can be performed at the patient's bedside. 

Watch this excellent video on cardiac CT imaging basics with anatomy explained:


Transthoracic echo is performed with the ultrasound probe on the patient's skin surface, while transesophageal echo is performed with the ultrasound probe within the esophageal lumen. Transesophageal echo offers better image quality that transthoracic echo, though is associated with the risks that come with imaging a patient with an esophageal ultrasound probe, namely the risks associated with sedation, aspiration, and esophageal perforation. Cardiac echo can provide us with structural information about the heart... things like identifying if there's a mass or thrombus in a cardiac chamber, the size and wall thickness of cardiac chambers, septal defects, valve disorders, and pericardial effusions. Cardiac echo also can provide us with... hemodynamic information, such as identifying cardiac wall motion abnormalities, cardiac output, and identifying and quantifying cardiac valvular disorders. Cardiac nucs studies allow us to qualitatively assess myocardial perfusion in different regions of the left ventricular wall. Instead of imaging the heart with ultrasound waves like in echo, with cardiac nucs studies, we study how radiotracer uptake in the myocardium looks when the patient is at rest compared to when they heart is being stressed. That stress can be achieved through making the patient exercise or through pharmacologic techniques. When we use SPECT to create the images, the radiotracers used are ones that emit gamma rays, while when we use PET to create the images, the radiotracers are ones that emit positrons. Cardiac nucs studies are inexpensive and, unlike most cardiac CT studies, don't impart any renal risks since we don't use iodinated contrast. However, like CT, cardiac nucs studies do expose patients to ionizing radiation. Cardiac nucs provides us with perfusion information about the heart, helping us identify regions where the vascular supply of left ventricular myocardium may be compromised, and showing us what regions of myocardium may or may not benefit from coronary artery bypass surgery. When we're considering a revascularization procedure like coronary artery bypass surgery, we like to know what's the condition of the myocardium in different regions of the left ventricle. Is the myocardium normal, or is it ischemic, stunned, hibernating, or infarcted? Ischemic myocardium occurs due to inadequate blood supply and is reversible if blood flow is restored. Stunned myocardium refers to temporary dysfunction following ischemia with time, stunned myocardium generally recovers its function and returns to normal. Hibernating myocardium is chronically underperfused but still viable myocardium - restoring blood flow to hibernating myocardium can lead to functional recovery and improved contractile performance. Infarcted myocardium results from prolonged and severe ischemia leading to tissue death - infarcted myocardium has lost its contractile function permanently and is replaced by scar tissue over time, and does not benefit from having its perfusion restored. The function vs. stress curves of these 5 myocardial states are different from each other... which means we can distinguish them from each other by comparing perfusion during stress and perfusion at rest. Cardiac MR imaging is a versatile way of studying the heart, but comes with the drawbacks associated with most MR studies, such as cost, limited availability, and reliance on highly skilled operators... not to mention, the inability of some patients with claustrophobia to tolerate it, our inability to use it in some patients with metal in their body, and problems with gadolinium contrast in folks with renal insufficiency. When these hurdles are overcome, you've got a modality that allows you to assess cardiac... structure - like identifying if there's a mass or thrombus in a cardiac chamber, the size and wall thickness of cardiac chambers, myocardial viability, septal defects, valve disorders, and congenital heart disease. Cardiac MR also allows us to assess other facets of the heart, but may be somewhat less competitive relative to other imaging modalities. For example, although cardiac MR allows us to assess... hemodynamics, echocardiography does an acceptable job and is way more accessible. Cardiac MR also allows us to evaluate myocardial... perfusion, though cardiac nucs does an okay job and is more accessible too. And while cardiac MR allows us to assess the... coronary arteries, cardiac CTA and cardiac cath a usually more accurate than MRI. Cardiac cath has been around for many decades, and is the only modality that affords us the ability to both diagnose and to treat. It comes with the downsides associated with fluoroscopy: radiation exposure, the downsides associated with iodinated contrast: renal failure if contrast nephropathy occurs, and the downsides associated with getting a catheter to the coronary artery: namely MI, stroke, arrhythmia, and bleeding. With cardiac catheterization we can assess... the patency of the coronary arteries... ...measure right sided heart pressures... And very importantly, we can sometimes directly intervene on problems revealed in the coronary arteries. Chest radiography is probably the cardiac imaging modality that's been around the longest. Unlike most imaging modalities we use to study the heart, chest radiography does not allow us to study the heart in 3 dimensions, as chest radiographs are basically a summation shadowgram where 3 dimensions get collapsed into two. This make it difficult to resolve individual cardiac structures reliably, though we can still infer some information for this cheap, fast, and widely available modality - a modality that I should add can be helpful in situations when what we thought was heart disease in a patient is actually being caused by something else. The chest radiograph may clue us in that a cardiac issue is present when the... Density of the cardiac shadow is abnormal, like in the present of cardiac calcification. Old MIs, valvular disease, coronary artery disease, and remote inflammatory pericardial disease can present with characteristic calcification patterns that can visible on a chest radiograph. Changes in the overall size of the cardiac shadow, especially when it looks bigger than usual, can be an indication that the patient's cardiac function is poor, that they have mitral valvular disease, or that a pericardial effusion has developed. Changes in the shape of the cardiac silhouette, can prompt us to more closely investigate the possibility of left atrial enlargement, left ventricular enlargement, or congenital anatomic variants. Changes in the position of the heart in the chest may serve as indicators for the presence of an emergency like a tension pneumothorax, or important issues like a large tumor or a large pleural effusion. Finally, radiopaque structures visible on a patient's chest radiograph can tell us a story about their cardiac surgical history in much less time that it takes to log into the EMR. Now let's take about... Cardiac CT. There are few objects in the body that are as challenging to do a great CT of than the heart... not only do we have to image something that moving quickly and constantly - which the patient has practically no control of, by the way - but we also have to do it in a way that allows us see very small structures well too. If we're hoping to study the heart in detail, we either have to complete a CT scan in less time than a heart beat or we have to somehow overcome cardiac motion artifact. Overcoming cardiac motion artifact is the more realistic path, and it takes a 3 part strategy. We'll (1) scan as fast as we can, (2) employ a technique called cardiac gating, and (3) slow the patients cardiac motion down a little. There are two things that permit us to physically do a CT scan faster. One is multidetector CT technology... and the second is cutting down the gantry rotation time of our CT machine as much as possible. Together both technologies help, but they're not enough. So we also must rely heavily on the second part of our strategy... Cardiac gating. With cardiac gating, we're going to record the patient's EKG at the same time we're acquiring their CT scan. The electrical signals that directly correspond to how the heart contracts over time are documented... in the patient's EKG. During each heartbeat... there's a P wave corresponding to the contraction of both atria, followed by a QRS complex corresponding to the contraction of both ventricles, followed by a T wave corresponding to the relaxation of both ventricles. The time between successive QRS complexes, or more specifically the R-R interval, corresponds to the duration of one complete cardiac cycle. Knowing all of this allows us to know when during our entire CT scan acquisition... the heart was is diastole and when it was in systole. The parts of our CT scan acquisition occurring during systole are when cardiac motion is the worst. On the positive side, the duration of systole is relatively fixed regardless of how quickly or slowly the patient's heart rate is. The parts of our CT scan acquisition occurring during diastole are when the heart is relatively still, and the best time to visualize the heart. On the negative side, the duration of diastole - unlike systole - is variable and changes as the patient's heart rate changes. If this is what a patient's EKG looks like at a leisurely heart rate... ...this is what it'll look like at a faster heart rate. Notice how the systolic durations remain relatively fixed, but the diastolic durations change. For the purposes of cardiac gating, we divide each cardiac cycle... into tenths. An ideal time to study the anatomy of the heart is usually from around... 55% of the way into the cardiac cycle through 75% of the way into the cardiac cycle. If the CT machine is continuously scanning the patient's heart through several consecutive cardiac cycles - as indicated by this continuous yellow bar... We could construct nice motion-free CT images of the heart if we were restrict the raw data we use to create our CT images to only the raw data acquired during the relatively motion free periods from 55% through 75% into each cardiac cycle. We refer to this as retrospective cardiac gating. Unless we're doing something fancy, all the data we acquired during the other 80% of the cardiac cycle indicated by the dark yellow regions is sort of wasted, in addition to the radiation dose incurred by the patient during that time. That's why folks came up with the concept of... prospective triggering... where we don't radiate the patient during the parts of the cardiac cycle we're not interested in. We turn the x-ray tube for a short duration at a point into each cardiac cycle that's prospectively determined, and hope that nothing unusual happens with the patient's rhythm after you hit the scan button. This works great when the patient's cardiac cycle is regular, since we can save a lot of dose, but can be problematic if the patient throws a PVC... in the middle of the scan, because that could cause us to collect raw data at the wrong part of the cardiac cycle and give us some really awful data to create our cardiac images with. If we had done retrospective cardiac gating... ...we'd have the ability to change where in the cardiac cycle we'd like to create our images from... So here's a summary of how prospective triggering compares to retrospective gating. Prospective trigger generally works best when heart rates do not exceed 70 beats a minute. One big advantage of prospective triggering is that radiation doses are a fraction of what patients receive with retrospective gating. However, prospective triggering is brittle. An arrhythmia could lead to a loss of decent data you need to create your images with. Retrospective gating is much more robust to heart rhythm and heart rate issues, though comes at the cost of much higher radiation dose. The CT acquisition mode with prospective triggering is an axial step-and-shoot technique, while the CT acquisition mode with retrospective gating is a helical technique. Because you're scanning the patient throughout their entire cardiac cycle with retrospective gating, you could conceivably make a movie of the entire cardiac cycle and derive functional information about the patient's heart. So how do you pick which type of cardiac gating to use? Prospective triggering tends to be a decent option for patients with low heart rates and stable sinus rhythm, while retrospective gating is better for patient with higher heart rates and with irregular rhythms. Here's an illustration of how gating effectively works... We're imaging the entire heart from data acquired throughout several successive cardiac cycles by reconstructing slabs of the cardiac anatomy, one slab at a time, using data from a very particularly fraction of each cardiac cycle. Besides doing CT scanning as fast as possible and cardiac gating, the third arm of our strategy is to keep the patent's heart rate as low as possible. In addition to asking patients to abstain from caffeine and nicotine before their scan, we can administer beta blockers which not only bring heart rate down but also decrease the likelihood of arrhythmia. There are three other things that can help make for a cardiac CT study successful... The heart moves and squashes a lot... when a patient breathes. So it's in our interests to finish the entire cardiac CT scan during a single breath hold before the patient has to breathe out. Realistically, we're talking about needing to keep cardiac CT scans under 15 seconds - which means we have to... keep the craniocaudal coverage of our scan... to a confined a territory as possible. We'll also want to keep our reconstruction and display... field of views as small as possible so that we can maximize the use of the pixels in each DICOM image to just the cardiac anatomy, and get just a little more spatial resolution for small objects like the coronary arteries. Finally, the way we handle the injection of intravenous contrast is important. If you look at routine enhanced chest CTs, you'll often see really dense contrast in the right heart chambers. The dense contrast can cause a lot of artifact that can make the nearby coronary arteries hard to see. For coronary CTAs, we often do a dual injection, where we inject a slug of intravenous contrast at a very high injection rate followed by a saline chaser bolus... push the dense contrast out of the right heart chambers. There are 5 major applications of cardiac CT: non-contrast imaging for coronary calcium scoring and contrast-enhanced imaging for studying coronary arteries, mapping pulmonary venous anatomy, assessing cardiac masses, and making measurements for transcatheter aortic valve replacement. Coronary calcium scoring is intended to be a... ...cardiovascular risk assessment tool. When we refer to cardiovascular risk assessment, we're typically referring to the likelihood a patient may die from coronary heart disease, have a non-fatal MI, or have a fatal or non-fatal stroke in the next 10 years. Folks usually like to stratify this risk in patients... we consider the risk high if the possibility of any of these events happening in the next 10 years is over 20%, while we consider the risk low if the possibility if under 10%. There's obviously a lot of interest in trying to accurately estimate this risk in patients, and models have been built that try to estimate this risk using the factors appearing in the left side on this slide. While folks agree that high risk patients need aggressive therapy to reduce the likelihood of any of those major adverse cardiovascular events from occurring, it's unclear what to do with folks who are at intermediate risk, that is at a 10-20% chance, for a major adverse cardiovascular event from occurring in the next 10 years... since it's in this group that the majority of cardiovascular events actually occur. How do we decide which patients in this risk group we pursue aggressive preventive therapy on? For some folks, coronary artery calcium might be a factor to influence this decision. Coronary artery calcium is a relatively sensitive predictor for coronary artery disease, however it's not specific for obstructive coronary arterial disease. While most folks agree that knowing if and how much coronary artery calcium is present seldom alters management decisions in patient stratified as low risk or high risk for a major adverse cardiovascular event in the next 10 years, some folks believe that coronary artery calcium might be helpful in their management of intermediate risk patients... perhaps as a means of motivating compliance with preventive therapies. The technique for coronary artery calcium scoring CT requires no intravenous contrast, and can be done with or without cardiac gating. Axial CT images are reconstructed at 2.5 to 3 mm for this technique and the amount of coronary artery calcium is semi-automatically quantified using the Agatston scoring system. Scores for the amount of calcium are assigned numerically, which correspond to very low, mildly increased, moderately increased, and moderately to severely increased coronary artery calcium. The measuring technique for coronary artery calcium is semi-automated, and here's how it works. Software will take all the axial CT images of the heart and nearby chest and color every pixel over 130 Hounsfield units pink. Some of these pixels will correspond to coronary artery calcium, and some will correspond to other calcified objects like calcified lymph nodes, calcified granulomas, or bone. That's where the manual part comes in... A radiologist will look at all the blobs of calcification on the entire CT image stack in a special software package and decide which ones are coronary artery calcium, and specifically in which coronary artery the calcification is in. They'll use a tool to tag which blobs are left main calcification, which blobs are LAD calcification, and so on. The tags are usually color coded... you can see that the one calcification on this image tagged as LAD calcification appears in yellow. Once the radiologist is finished tagging coronary artery calcium on the CT image stack... The software tallies up the volume of all the coronary artery calcium in square mm and their estimated mass in milligrams by vessel, and calculates a numerical score for the amount of coronary artery calcium the patient has using the Agatston scoring system. The coronary artery calcium score is then compared to people of matched age, gender, and race and a percentile provided. Finally the CT images are read as a normal CT would, looking for incidental findings like lung nodules, lymphadenopathy, for example. Coronary CTA, unlike coronary artery calcium scoring, is a contrast-enhanced CT study that not only permits us to get a much higher resolution look at the coronary arteries, but also see how patent the lumen of the vessels are. Coronary CTA is a sensitive test for obstructive coronary artery disease... though it's not quite not quite as specific... so basically, it's a test with a nice negative predictive value - but can occasionally result in some false positives with regards to the severity of coronary vessel occlusion. Because its negative predictive value is great, but its specificity is so-so, coronary CTA can be a good test for patient with a low or intermediate pre-test likelihood for coronary artery stenosis. On the other hand, coronary CTA is not an ideal test for folks with a very low pre-test likelihood of coronary artery disease because of its false positive rate is too high. Coronary CTA is also not an ideal test for folks with high pre-test likelihood of coronary artery disease because the sensitivity may not necessarily be high enough and many of these people will probably need catheter angiography in the end. So the applications of coronary CTA are to rule out significant coronary artery stenoses in patients with chest pain & either a low or intermediate likelihood of stenosis, in patients with non-acute chest pain & equivocal stress tests, in patients with acute chest pain & negative cardiac enzymes, in patients requiring clearance for non-coronary cardiac surgery. Coronary CTA can also be an option in patients for whom cardiac cath is contraindicated and in patients with coronary artery anomalies. Coronary CTA can be a very nice way of assessing the patency of coronary artery bypass grafts. Compared to native coronary arteries, coronary bypass grafts tend to be easier to assess because they're often of larger caliber and subject to less cardiac motion, in fact that sensitivity and specificity of CTA for coronary bypass graft occlusions and significant stenosis is very high. On the other hand, coronary CTA is pretty poor at assessing the native coronary arteries downstream from grafts and ungrafted segments due to those vessels' very small diameters and typically high preponderance of calcified plaques. From a technique perspective, coronary CTA is a contrast-enhanced cardiac CT study that requires cardiac gating and typically reconstructed at 0.625 mm axial slice thickness. Beta blockers are typically used in patients whose heart rates are above 60 and nitroglycerine is usually given immediately before the CTA if there are no contraindications. The review of a coronary CTA often begins with a typical chest CT read, looking for lung nodules, lymphadenopathy, and other potentially important findings. Then we move on to examine the heart in detail... ...typically in a 4D post-processing workstation that permits us to the heart in 3 planes, and to step forwards and back within 10 frames of a complete cardiac cycle. Unlike pulmonary arteries, which generally run in relatively straight radially oriented courses from hila to periphery, coronary arteries have a much more curvilinear course - and are best evaluated in multiple planes simultaneously instead of just one... Occasionally, linearized MPRs can linearize coronary arteries for further assessment. The ability to step back and forwards throughout frames of a single cardiac cycle is important too - sometimes a coronary artery segment may be obscured by cardiac motion blur transiently at one point in the cardiac cycle. In retrospectively gated coronary CTA studies, for which images of the entire heart have been captured throughout the entire cardiac cycle... we're able to re-create standard moving echocardiography views... that provide us with a familiar way of looking at the motion of the heart and its valves... Mapping pulmonary vein anatomy is another application of cardiac CT. The electrical impulses that trigger atrial fibrillation often arise from pulmonary vein segment just before they enter the left atrium... ...and one method of treating atrial fibrillation... to electrically isolated the pulmonary veins from the rest of the heart... burning lines... around the pulmonary venous ostia from inside the heart. This can by done by RF ablation or cryoablation, and has been reported to have success rates between 70 and 85% in folks with refractory atrial fibrillation. Since pulmonary venous anatomy can vary a lot from patient to patient, understanding each patient's pulmonary venous anatomy is crucial, and cardiac CT can provide a precise anatomic map to guide their pulmonary vein ablation procedure. Cardiac CT is also useful to assess for complications after pulmonary vein ablation, like perforation or pulmonary venous stenosis. The review of a pulmonary venous mapping CT typically begins with a typical chest CT read, looking for lung nodules, lymphadenopathy, and other potentially important findings. Then... Images of the left atrium and pulmonary veins emptying into the left atrium are reviewed on a 3D post-processing workstation. We typically will create a 3D shade volume rendering of the left atrium to help illustrate the anatomy of the left atrium and pulmonary veins. We'll usually also create endoscopic views within the left atrium chamber looking out into the pulmonary venous ostia. These views allow us to qualitatively describe the anatomy of the pulmonary veins entering the heart... ...while we'll then review the left atrium and pulmonary veins in a conventional 3D post-processing view to measure the diameter and angulation of each pulmonary vein. Finally a volume of the left atrial chamber is measured. Cardiac CT is also a good imaging modality for studying cardiac masses. Some cardiac masses are non-neoplastic, like intracardiac thrombi, vegetations on infected valves and lipomatous hypertrophy of the interatrial septum... while some cardiac masses are neoplastic. Two important facts to know about cardiac tumors is that... one: 3/4 of cardiac tumors are benign, and two: cardiac metastases are much, much more common that primary cardiac tumors. The need to characterize a cardiac mass usually occurs when either a cardiac mass has been incidentally discovered on a different imaging study or in the midst of a workup of a symptomatic patient. Now, you have a couple imaging modality choices for studying cardiac masses... Echocardiography is probably your first-like imaging study, however you may not be able to see the cardiac mass well enough and your ability to characterize what the mass is made of is often limited. For example, it's not always possibly to distinguish if a cardiac mass is thrombus or tumor on echo. Cardiac MR can be much more effective at characterizing cardiac masses, however it can sometimes be hamstrung by the weaknesses of MRI, such as lower spatial resolution, impact of imaging artifacts, and operator dependence. Cardiac MRI can also be tough on patients who can't hold their breath very long or who have claustrophobia. Finally, implanted cardiac devices may often disqualify a patient from being able to enter the MR magnet. Because of the limitations of echo and cardiac MRI, cardiac CT is often valuable tool for studying cardiac masses. Not only can it allow us to distinguish if a cardiac mass is fat, soft tissue, or calcification, the presence or absence of enhancement during cardiac CT can help suggest if a cardiac mass is thrombus, tumor, or fibrosis. In addition, cardiac CT also allows us to study anatomy immediately around the heart too, like other blood vessels and the lungs. The possibility of contrast-induced nephropathy, however, is one downside of cardiac CT. The review of a cardiac mass CT typically begins with a typical chest CT read, looking for lung nodules, lymphadenopathy, and other potentially important findings. Then... The remainder of the cardiac review is usually done on a conventional 3D post-processing workstation. Finally, cardiac CT is an important part of the planning that precedes the implantation of a percutaneous inserted transcatheter aortic valve replacement or TAVR. 1 in 8 elderly people have moderate or severe aortic stenosis. A diagnosis of severe aortic stenosis is associated with a 50% rate of mortality at 2-years, and TAVR - the percutaneous insertion of an aortic valve replacement mounted on a stent - is a very important therapeutic option since at least a third of patients with severe aortic stenosis are not candidates for surgical repair of their aortic valve. Here's a brief video of how a TAVR insertion and placement is performed. Accurately sizing the TAVR is absolutely critical in this procedure. If the TAVR is too large, you can cause a catastrophic aortic annular rupture. If the TAVR is too small, you could end up with a leak around the valve or even have the valve could dangerously float away. Positioning a correctly sized TAVR is also critical - you could precipitate a fatal MI if the TAVR accidentally covers a coronary artery origin. There's another critical factor with TAVR placement... ...the TAVR - like any large vascular stent - must be delivered through a vascular sheath. If the sheath is too large, you could substantially injure the artery it's in. There was a time when folks sizing TAVRs used echocardiography to measure the size of the aortic annulus. However, cardiac CT has replaced echo in this role, not only because CTA is more accurate, but because it allows us to do other things such as assessing the location and anatomy of the coronary arteries and also measuring the caliber of the artery through which the collapsed TAVR and sheath must first pass. A TAVR CT typically consists of an enhanced chest, abdomen, and pelvic CTA. The review of a TAVR CT begins with a typical chest, abdomen, pelvis CT read... followed by an extensive review of the CT volume on a 3D post-processing workstation. The review of the 3D post-processing workstation involves a lot of image manipulation and a lot of measurements. Folks may start this interpretation... first creating a linearized MPR of the entire vascular access route of the TAVR, starting from either common femoral artery at the groin all the way to the heart. This linearized MPR permits the radiologist to assess... ...the internal vessel diameter along the entire course the collapsed TAVR needs to travel, in addition to sites where an atherosclerotic plaque could potentially be dislodged. Many carefully measurements of the aortic annulus and its vicinity are then made on the 3D workstation using gated CT images of the heart. The major diameter, minor diameter, mean diameter, perimeter, and area of the aortic annulus during systole are measured. Other measurements must also be made, including the... angle of the aortic root... The major diameter, minor diameter, perimeter, and area of the left ventricular outflow tract 4 mm below the aortic annulus at systole... The major diameter, minor diameter, and area of the sinuses of Valsalva during diastole... And the cusp-to-commissure diameters of the sinuses of Valsalva during diastole. The sinus of Valsalva heights are measured at diastole for the right coronary cusp... and the left coronary cusp... and the major and minor diameters of the ascending thoracic aorta are measured 4 cm above the aortic annulus at diastole. Finally, a cine view of the aortic valve is created to high light the amount and location of aortic valvular calcifications, in addition to the amount of aortic valve excursion during a complete cardiac cycle. And that completes our review of the 5 applications of cardiac CT... which is one of the 6 modalities we commonly can use to image the heart. And since you stayed around to the end, I want to share one last slide that's useful when discussing cardiac imaging with patients. It's a nice visualization of the radiation dose a patient might get from each kind of cardiac imaging study... Using a low-dose lung cancer screening chest CT in blue as a reference, you can develop an appreciation for how coronary CTAs and cardiac nucs studies can result in approximately 7 times more radiation dose than a lung cancer LDCT, while the radiation dose of a chest radiograph is barely visible at this scale.