PATIENT PREPARATION

The patient should be kept NPO for at least 4 hours prior to the scan as a precaution, as is standard for all contrast CT examinations, because of the potential for nausea and vomiting associated with contrast administration. It is extremely important that the patient not consume caffeine for at least 12 hours prior to the examination; this helps with heart rate control, which is critical for optimal image quality. With the newer generation of scanners having improved temporal resolution, high-quality images can be obtained with higher heart rates.

Electrocardiogram (ECG) leads need to be attached, which are used by the scanner to determine data reconstruction and electrocardiographic tube modulation, if used. In addition, this allows monitoring of the patient for absolute heart rate, heart rate variability, and arrhythmias, which are critical in determining the optimum protocol. A good electrocardiographic tracing (Fig. 9-1) is required, especially if using electrocardiographic tube modulation for overall dose reduction, because the x-ray tube modulation is synchronized with the cardiac cycle, using the R wave as a reference. It is important to ensure that the wire leads and ECG patches are outside the scanning field as much as possible, and in particular to ensure that the wires are not coiled on the patient’s chest, which can result in significant streak artifact. For someone who is very hairy, good adherence and electrical conductivity of the pads may require shaving a small area. In particular, noisy electrocardiographic tracings (Fig. 9-2) can lead to inappropriate modulation of the tube current, resulting in a poor signal-to-noise ratio at the specific phase of the cardiac cycle required for coronary evaluation. In addition, be sure that the R wave peak is of greater amplitude compared with the T wave to prevent tube current modulation synchronized to the wrong phase of the cardiac cycle. Some scanners allow switching of the lead at the scanner console; if not, manual switching of the lead attachments to the ECG patches on the patient can often correct a situation in which the T wave has a higher peak.

FIGURE 9-1 Good electrocardiographic tracing with clean R wave peaks and essentially no baseline noise.

FIGURE 9-2 Poor-quality ECG tracing, with very noisy baseline.

If the patient’s baseline heart rate is above an acceptable limit for optimal image quality, the use of β blockers is recommended, provided no contraindications (e.g., asthma, aortic stenosis) are present.¹ This acceptable limit is increasing with the latest generation of scanners. However, with the current generation of 64-slice scanners, a heart rate below 65 beats/min is highly recommended to increase the likelihood of an optimal high-quality scan with the least amount of motion artifact. There are multiple protocols for the administration of β blockers. The two basic protocols include the use of oral administration, both the night before and the morning of the scan, or just the morning of the scan, versus intravenous (IV) administration just prior to the scan.¹ In general, oral administration seems to result in a more predictable result, and it also requires less time and personnel commitment from the support staff. However, IV administration can be effective if the patient did not have prior administration of an oral β blocker or the target heart rate was not achieved after the administration of an oral dose. IV administration tends to have an all or none effect, meaning that usually if there is a noticeable decrease in heart rate with the first dose administered, a desired heart rate can usually be achieved. However, in many patients, the maximum dose (25 mg metoprolol at our institution) is administered without a noticeable change in heart rate. If the patient’s heart rate does not reach the target, then the scan can be obtained with the understanding that it will likely be suboptimal, with the potential for nonevaluable segments of the coronary arteries, or the patient can be rescheduled with more aggressive oral β blocker administration prior to the scan.

If the patient has severe arrhythmias, then the patient is generally not a good candidate for cardiac CT and requires management of the arrhythmia prior to a coronary CTA. Usually, this is the obligation of the referring physician unless the performing physician is comfortable with antiarrhythmic medications.

A perfect breath-hold is critical for coronary imaging in particular. This requires the patient to have a steady breath-hold throughout the entire scan and needs to be relaxed—that is, no Valsalva maneuver. Therefore, the patient is coached on the breath-hold technique, which is practiced at least twice with the patient prior to the scan. The patient is instructed to take a breath in and hold the breath, with additional instructions during the breath-hold to be relaxed, not to bear down, and not to let any air out or take any air in once the breath is held. The breath-hold is practiced for approximately 5 seconds longer than the time required for the actual scan being performed.

A baseline heart rate and blood pressure are documented. Then, a brief history is obtained from the patient, including contraindications to the medications that will be administered for the examination, which may include a beta blocker and nitroglycerin. The patient is questioned about history of asthma, heart failure, and heart murmur with respect to β blockers and erectile dysfunction medications, and sildenafil with respect to nitroglycerin. Nitroglycerin is administered to dilate the coronary arteries for optimal visualization.² If there are no contraindications, the medications to be administered for the examination include up to 25 mg of metoprolol IV, in 5-mg push increments, and two 0.4-µg sublingual tablets. The onset of action for a sublingual nitroglycerin tablet is 1 to 3 minutes; it reaches maximum effect at 5 minutes, lasting at least 25 minutes. Therefore, these will be administered just prior to the scout and coronary CTA will be performed at approximately 5 to 6 minute after administration.

Additional history, which is collected while administering the β blocker, includes any symptoms, as well as their character and length of duration. Also, risk factors, including smoking, diabetes, lipid profile, hypertension, and family history, are documented. If additional testing has already been performed, these data are collected. especially if the previous testing led to the performance of cardiac CT.

SCANNING TECHNIQUE

The scan can be performed using retrospective electrocardiographic gating or a prospective electrocardiographic triggering (step and shoot) technique. With the retrospective electrocardiographic gating technique, the scan is performed using a standard helical scan acquisition, as for routine chest or abdominal CT scans, except a much smaller pitch, usually around 0.2, is used to ensure adequate data acquisition at all anatomic locations throughout the entire cardiac cycle. At the same time that the helical CT scan is obtained, the electrocardiographic tracing from the patient is recorded. Unless electrocardiographic tube modulation is used to reduce radiation exposure, the electrocardiographic tracing has no effect on the helical CT scan at the time of data acquisition. Rather, the tracing will be used to reconstruct axial source images from the raw data during the predefined portion of the cardiac cycle chosen after the scan. The tracing is used as a reference to determine which data at a specific anatomic location were collected during the defined phase of the cardiac cycle.

Contrast Administration

The IV contrast is administered through an 18-gauge angiocatheter in the antecubital fossa. IV access is preferably via the right arm, which helps prevent significant streak artifact from the dense inflow of contrast, especially when scanning to the level of the aortic arch for evaluation of aortic pathology or the origin of coronary artery bypass grafts (in particular, the left internal mammary). To ensure adequate IV access, consider training CT technologists to place the angiocatheter under ultrasound guidance. We also use a trapeze, which helps keep the patient’s arms straight, thus helping to prevent complications of high-flow injections. We also find it more comfortable for our older patients, who frequently have shoulder disabilities that make it difficult for them to place their arms above their head. The trapeze has 10 pounds of weight at the other end, helping to hold the patient’s arms comfortably in an extended position, out of the way of the scan (Fig. 9-3).

FIGURE 9-3 Trapeze device used to elevate patient’s arms during the scan.

The contrast is administered using a test bolus or a track and trigger technique. Briefly, the test bolus system injects a 20-mL bolus of contrast followed by 20 to 40 mL of saline push at the same injection rate planned for the coronary CTA scan. Intermittent low-dose images are obtained at the level of the carina, beginning approximately 10 seconds after the start of injection and are acquired every 1 to 2 seconds, until the peak of contrast has been demonstrated in the ascending aorta.⁷ The images are loaded into an analysis software application on the scanner, which determines the time of peak contrast enhancement. The coronary CTA scan is then obtained with the full dose of contrast administered and the scan acquisition timing determined according to peak enhancement. The scan acquisition is usually planned to begin approximately 2 to 4 seconds after the peak enhancement determined from the test bolus to allow stabilization of the enhancement curve prior to scanning. This was an important technique in the past, especially with the slower scanners and longer breath-holds required for coronary CTA examinations (40 seconds on a four-slice scanner). With the faster scanners today having scan times as short as 4 to 5 seconds on a 256-slice scanner, a track and trigger technique can result in excellent opacification, without the need for the additional 20 mL of contrast of the test bolus. Therefore, coronary CTA with the track and trigger technique on the 256-slice scanner can be performed with as little as 50 mL of contrast.

For coronary CTA evaluation, the scan is performed in a craniocaudal direction to allow for washout of dense contrast in the right heart and allow better visualization of the right coronary artery. This scan direction also allows adequate opacification of the distal coronary vessels while minimizing contrast in the coronary veins. A patient who has undergone coronary artery bypass grafting will also be scanned in a craniocaudal direction, beginning at the level of the clavicles to include the origin of the left internal mammary artery. This is a longer total scan time and therefore a larger volume of contrast is required. When performing cardiac CTA for pulmonary vein evaluation, the scan is performed in a caudocranial direction. This allows for washout of dense contrast from the superior vena cava (SVC), which is usually in close approximation to the right superior pulmonary vein, resulting in difficulty evaluating the right superior pulmonary vein because of streak artifacts. It also allows for washout of the pulmonary arterial system, which is very helpful in the postprocessing portion of the examination. To create volume rendered images of the pulmonary veins and left atrium, which are used for an overview of the anatomy, the pulmonary arteries need to be removed. If the pulmonary arteries have been washed out, with a resulting difference of 100 HU (Hounsfield units) or more between the pulmonary arteries and veins, a threshold technique based on Hounsfield units can be used to render the pulmonary arteries invisible, rather than having to cut the vessels out.

The amount of contrast is tailored for each patient and each examination. For injection of contrast, a rate of 5 to 7 mL/sec is recommended, depending on the contrast medium’s viscosity and iodine content. It is very important to be sure that the contrast has been warmed to near body temperature to allow easy contrast injection without reaching the pressure limit of the power injector, which can fragment the contrast bolus. Two techniques are currently used for contrast administration—contrast followed immediately by a saline push (two-phase), or contrast followed by a contrast and saline mix followed by saline push (three-phase). The purpose of these techniques is to provide minimal opacification in the right heart to help identify any incidental findings of the right heart, such as a thrombus or mass. However, if there is too much contrast in the right heart, the right coronary artery (RCA) could be obscured by streak artifact or there could be the false appearance of a stenosis caused by a streak artifact. Intracardiac shunts such as patent foramen ovale (PFO), atrial septal defect (ASD), or ventricular septal defect (VSD) may be obscured if there is not enough difference in contrast enhancement between the left and right heart.

To calculate the volume of contrast required for a two-phase technique, the time required to perform the scan needs to be known. If using a track and trigger technique, the time from the point the scan is triggered to the time the scan actually begins (post–threshold trigger delay) needs to be added to the time of the actual scan acquisition. For residual mild enhancement in the right ventricle, an extra 1 to 2 seconds of contrast injection is added. Thus, the total time of contrast injection should be the scan time plus post–threshold trigger delay time plus a 1- to 2-second additional “fudge factor.” For example, for a new 256-slice scanner, the scan time is 5 seconds and the time from trigger to the beginning of the scan is 4 seconds, which results in a total of 10 to 11 seconds of contrast injection needed to opacify the right heart minimally. If we use 6 mL/sec, 60 to 66 mL of contrast is needed.

POSTPROCEDURAL CONSIDERATIONS

Once the scan is complete, if everything has gone well, there will be optimal quality images, resulting in easier interpretation. First, review the electrocardiographic tracing to be sure that no arrhythmias occurred during the scan that will require editing for optimization of the data set. Next, pan and zoom the images to optimize spatial resolution for viewing the coronary arteries. Keep in mind that with a standard 512 × 512 matrix, the smaller the field of view, the better the spatial resolution will be, which is paramount in coronary artery evaluation. In general, reconstructed axial images should be 0.6 to 0.9 mm thick, with a 50% increment (0.3 to 0.45 mm). If performing stent evaluations, the thinnest possible slice thickness with a 50% increment is recommended, combined with a sharp kernel or filter.^10,11 It is usually helpful to reconstruct three coronary data sets, including the phase of the R-R interval, which is usually optimal, as well as ±5% around this phase of the cardiac cycle. Depending on the scanner manufacturer, this may be 60% to 65% to 70% or 70% to 75% to 80%. It should be noted that there is a difference in manufacturer definition of the percentage of the R-R interval, either defining it based on the beginning of the temporal window or at the center of the temporal window. At a heart rate of 60 beats/min, this results in approximately a 10% shift, depending on how the phase is defined (i.e., 65% vs. 75%). In addition, if there is motion in the right coronary artery, a systolic reconstruction may be beneficial (e.g., 35% or 40%).⁹ A multiphase data set of 5% to 95% every 10% is also reconstructed, with a slice thickness of 1.5 mm and increment of 1.5 mm; this is used for valve evaluation and qualitative and quantitative cardiac functional analysis. Finally, a full field of view data set can be created, which includes all soft tissues from skin to skin along the z-axis coverage of the scan, to evaluate for incidental findings, which may be the source of the patient’s chest pain. These are reconstructed using a routine chest CT protocol such as 4-mm slice thickness with a 3-mm increment.

With the current generation of software on a dedicated workstation, it is fairly easy and accurate to determine cardiac function using the multiphase data set.¹² Curved multiplanar reconstructions (MPRs) of the coronary arteries are also created and stored on the picture archiving and communication system, which is very helpful in communicating the findings to referring physicians.

As for image interpretation, this is almost exclusively performed on a dedicated workstation. There are multiple software packages and techniques available for image interpretation, but those with the most experience tend to use the full axial data sets combined with interactive MPR and maximal intensity projection (MIP) visualization when a lesion is identified. This has been found by at least one study to be the most accurate method of image interpretation, as opposed to prerendered curved MPRs or vessel view technique.¹³ Creating a structured report is highly encouraged.¹⁴ The recommended method of reporting coronary narrowing and stenosis is based on visual inspection without noting specific diameter or area narrowing measurements because generally the CT technique overestimates the narrowing compared with coronary angiography.

Pitfalls and Solutions: Artifacts

Breathing artifacts are easily recognized by stair-step artifacts in the sternum (Fig. 9-4). There are essentially no techniques during postprocessing that will compensate for breathing artifacts. If there is gross movement of the patient during the scan, the sternum and spine will have stair-step artifacts.

FIGURE 9-4 Stair-step artifact of the sternum (green arrows) and soft tissues of the heart and mediastinum secondary to patient breathing during the scan. Notice the lack of artifact in the spine (yellow arrows), indicating that this is breathing and not gross patient motion.

If there are stair-step artifacts in the heart but not in the sternum or spine, this is consistent with a synchronization and data reconstruction issue, referred to as electrocardiographic misregistration (Fig. 9-5). In this situation, the images can be significantly improved by editing the tagging of the electrocardiographic tracing. This may be a result of an arrhythmic beat, which can be completely ignored (Fig. 9-6), or it may be caused by inappropriate R wave tagging by the scanner either tagging a T wave or putting in false R wave tags caused by a noisy electrocardiographic tracing (Fig. 9-7). Once the ECG has been corrected, the images often are at the least diagnostic and frequently are of high quality.

FIGURE 9-5 A, Stair-step artifact seen only in the heart (yellow arrows) but sparing the sternum (green arrows). B, Artifact is absent after editing, confirming electrocardiographic misregistration.

FIGURE 9-6 Electrocardiographic editing in which the premature beat (highlighted in blue) is simply ignored when the axial data set is constructed from the raw data.

FIGURE 9-7 Poor-quality electrocardiographic tracing, which resulted in extra R-wave tags being placed by the scanner secondary to a noisy baseline.

If there are no problems with the tagging, but there still remains a motion or stair-step artifact in the images, especially those of the RCA, a systolic reconstruction may be very helpful (see earlier).

KEY POINTS

There are many technical considerations when performing routine cardiac CT, even in a daily clinical practice.

The time spent to optimize the examination up front, however, will be well rewarded at the time of image interpretation because of the resulting high image quality.