Current Clinical Concerns in CT: Results : Protocols
Pulmonary embolism
1. We have a volume zoom at our hospital and use bolus tracking for our PE CTAs of the chest. We have been setting our trigger at 160, requested by our old radiologists. Most of the time the scans are good, but once in a while the monitoring scans go on and on and we have to hit start, the scans then are not enhanced. I called applications and they suggested lower the trigger. Any other suggestions?
Answer: I would agree that lowering the trigger is the first step. With bolus tracking, you need to know when the curve stops rising, and you must take control and immediately trigger the scan. |
2. We are getting poor enhancement using a GE light speed 8 slice. Currently we are trying SMART PREP using a threshold of 150. Is this a correct threshold? Should it be lower? On some studies we are not peaking at 150, but lower. Also, we are scanning top down as opposed to bottom up. Is this correct?
Answer: I might push the target to 120 HU. I do like scanning from the bottom up. This may solve your issues. |
3. What threshold has to be used for bolus tracking in pulmonary angiography (Asked by 2 individuals)? We use 16-MDCT with a scanning time of PA (12 seconds) and CM of 300 mgI, volume 100 ml and rate of injection 4 cc/sec.
Answer: Most people would suggest a threshold of 100-120 HU as a good target. I would also use this target, but scan from diaphragm to lung apex. This will give a good contrast profile, and minimize artifact off the SVC as well. |
4. We have a 4 slice Siemens Volume Zoom and there is a difference of opinions among my staff as to where to place the marker for bolus tracking. Some use the pulmonary artery and others use the RV. We scan from bottom up, with factors of 5 mm slice, 1 mm collimation and feed of 5 mm.
| Answer: I place it in the pulmonary artery. The flow measurements are more predictable. |
5. On several occasions, with our multi-slice scanner, we have noted something unusual on our PE studies. We inject at a rate of 3.5 cc/sec and will see great opacification in the subclavian and the aorta, but only minimal or intermittent opacification in the pulmonary vessels. Can you explain why this occurs?
Answer: We too have noticed that in up to 5% of cases, despite doing everything correctly, the study is suboptimal. Possibilities include that the patient valsalved during the study or differential flow through the pulmonary vessels. For example, we have found that pregnant patients are particularly difficult to time. |
6. There was an article in March 2003 issue of Radiology that indicated a potential problem with PE scanning in many patients with clinically unsuspected patent ductus arteriosus. We have seen a large number of patients with a scan pattern that matches this article, i.e. bolus tracking shows a good peak in PA, full scan done in inspiration and is finished before the contrast bolus is over, but the PA opacification is significantly worse than the aorta. Have you had this experience? Do you believe this article? Any suggestions? Should PE scans be done in expiration on a 16 slice?
Answer: We have seen this, and it may be due to patent foramen ovale (PFO) or ASD as the article suggests, but other thoughts include that the patient valsalved during the study or differential flow through the pulmonary vessels. |
7. We have a problem with several PE studies we are doing. The scans are done on a GE 4 slice using 'smart-prep', but we are missing contrast enhancement in the PA's, yet the SVC and right heart are well enhanced, suggesting some sort of Valsalva during the breathold. Have you experienced this and how have you corrected the problem? Our timing seems to be perfect.
Answer: We have seen this, and it may be due to patient valsalva during the study or differential flow through the pulmonary vessels |
8. My PE study had contrast streaking in the pulmonary artery from the SVC. What can I do to minimize this? Isn't it normal to still have contrast going into the SVC as you finish the PE study?
Answer: This is a common problem, but can be reduced or eliminated by using a saline bolus following contrast injection. Also, you can minimize the problem by scanning from the diaphragm to the top of the chest, rather than from the apex downward. |
9. What do you suggest for lung perfusion study in cases of PE?
Answer: I am not certain exactly what you are asking; however, we do evaluate the lung parenchyma for regions of hypoperfusion. In the future with 64 slice scanner, we may be looking at lung perfusion. |
10. Does anyone have any experience or knowledge of performing a CTA for pulmonary embolism with gadolinium? Specifically, can gadolinium be give in sufficient volume to perform the study with a single-slice spiral scanner at .7 second per rotation?
Answer: I know of no article and would not be very confident that it would. I do not think it would be satisfactory, and we do not use it. |
11. When doing a PE study, every tech in our hospital has an opinion about how important it is to use an 18g angiocath. However, I noticed in your protocols for PE;s, you recommend 3 ml/sec. I personally believe that 3 ml/sec can be accomplished with a 20g. Is a 20g also an adequate gauge? Can you mention what the deal is with angiocath placed in the AC area vs. the hand?
Answer: A 20 g works fine. We prefer and use the AC fossa on our studies. When the only line is in the wrist, we use it, but very carefully. |
12. We just installed a 16 slice sensation in our ER. We are tweaking protocols and have some issues with our chest studies. We would like to be able to opacify the PAs and the aorta in order to obtain a diagnostic study of both. Currently we are using the bolus tracking feature with the ROI placed on the aortic root. The threshold is set to 50 HU. Our thinking is that with the low HU setting that as soon as contrast enters the root, it should trigger the scans initiation and we should get good opacification of both PAs and aorta. We have been having some success to date with this. We are using about 120 cc of Omnipaque 300 with a rate of 4 cc/sec. My question to you is does this sound feasible and if so, should we be using a higher concentration of contrast, say Omni 350 for example? It is my opinion that with the multislice we need the higher iodine concentrations for consistency. Would you agree?
Answer: Typically 50 HU is too low, and may not lead to a satisfactory pulmonary artery exam. Most people would use a threshold of 120 HU. |
13. We are having problems with about 10% of more of our PE scans on our Sensation 16. We are using the .75 collimator, 1 mm recon, bolus tracking with trigger (we have tried 120-160 HU, currently using 130 HU), 4 cc/sec injection, Isovue 300 (occasionally 370 on large patients), 150 cc loaded in case the bolus tracking triggers late so that we won't run out of contrast, and scan from lung bases up. We have 2 principle problems: (1) The center of the pulmonary arteries appear less dense than the periphery, which I assume is flow artifact (or beam hardening) OR (2) The monitoring image when the bolus tracking triggers looks beautiful with a dense PA, but then 5-6 seconds later when the scan starts, the aorta looks good but the PA has started to fade. This is very strange considering I still use 150 cc of contrast and you can usually see lots of contrast coming in from the arm at the end of the run. I have shown a case of this to Siemens apps and they are perplexed too.
- a. Could you address the issue of flow artifacts on PE studies in more detail and why Visipaque would help? Are there other ways to decrease these flow artifacts? Could beam hardening artifacts be a problem here?
Answer: I do not think it is beam hardening. Visipaque may be valuable since it is isosmolar, atleast in theory and in several published articles, it is less likely to have flow related or mixing artifacts. Perhaps increasing the concentration to 350 mgI/mL may be valuable. |
- b. Our patients tend to be pretty large but we are using techniques as recommended by Siemens apps and doing what they consider appropriate modifications of mA for large body habitus. Would I be better off using the 1.5 mm collimator for some patients?
Answer: If patients are large, the only way to get good studies are a combination of increasing the mAs and increasing the contrast volume. |
14. Is there a role for 370 or other higher strength contrast agents in PE imaging?
Answer: The answer is no. Although in theory one might think it would be better, in reality it has no advantages and potential disadvantages, as the increased density may hide small PE's. |
15. Is it possible for a bolus to peak and then quickly fade, even if the power injector is still pushing contrast in?
Answer: In theory, if you are still injecting contrast, although the peak may not increase, you should be able to maintain the peak. |
16. Noticed in your PE protocol that nothing was mentioned about CM concentration! We routinely use 80 cc of 320 mgI/ml with a Siemens Sensation 16, with satisfactory results. In patients with renal insufficiency, we may go even further down. The record so far is 40 cc, 160 mgI/ml iodixanol (6.4 g iodine) at 4 ml/sec in a 50 kg person with GFR of 22 ml/min and using 80 kV, 16 x 1.5 mm slices, pitch .75, 220 mAs, care bolus with 5 sec delay and 6.5 sec scanning time. Enhancement in the pulmonary arteries was 320 HU!! It appears that we could individualize CM dosage, especially in low-weight patients (with renal failure) where we also can use 80 kV without too much noise.
Answer: We usually use 120 ml of Visapaque-320 and in the past had used 120 ml of Omnipaque-350. I agree that the dose per patient to get a good study varies, especially in regard to cardiac output. With decreased output, low volumes are impressive. We are also trying to begin using saline flushes, and see how much advantage that provides. |
17. What is the protocol for CTPE study for pregnant patients? Has that been modified from your routine PE study? We have dose reduction software but should the kV and mAs be lowered also?
Answer: We lower the mAs a bit but be careful, as these patients often, because of fluid overload, can have very grainy scans. You want to make sure you can find the PE when present. |
References
Schoepf UJ and Costello P. CT angiography for diagnosis of pulmonary embolism: State of the art. Radiology 2004; 230: 329-337.
- Summary: This recent review of the state of the art addresses the advantages of multislice CT for pulmonary embolism, as well as the limitations. In addition, the authors discuss data management in the era of large volume MDCT datasets.
An algorithm is decribed, addressing the roles of d-dimer, ultrasound, CT and conventional pulmonary angiography.
- Summary: This single detector helical study compared thoracic CT examinations in 50 patients who received 60 mL of 370 mgI/ml contrast followed by a 30 mL saline flush. These 50 patients had undergone a previous thoracic CT with 75 mL of 370 mgI/ml contrast. A double power injector was used to administer the saline flush. There was no significant difference in attenuation levels in the pulmonary artery and ascending aorta, nor in the grading for depiction of mediastinal and hilar structures. However, the saline flush reduced the superior vena cava attenuation and associated perivenous artifact.
- Summary: A survey of 57 members of the Society of Thoracic Radiologists is presented in this study. Of those who participated, 43 (75%) reported that they perform CT angiography in pregnant patients suspected of having pulmonary embolism. Of these, 26 (60%) require informed consent from the patient. With respect to the protocol: 17 (40%) modify the protocol to reduce radiation dose. The majority (71%) accomplish this by decreasing the scanning area along the z-axis. Additional means of dose reduction reported include increasing the pitch, decreasing the field of view, eliminating the frontal and lateral scout images, reducing milliampere-seconds or peak kVp, and increasing detector collimation thickness.
- Summary: This study consisted of 289 patients with suspected PE who underwent contrast enhanced single detector helical CT. Section thickness was 3-5 mm and recontruction index 2-3 mm. Patients received 80-150 ml of 270 mgI/ml Visipaque or 300 mgI/ml Ultravist. Contrast was infused at 2-3 cc/second, and imaging initiated 15 seconds after the onset of contrast infusion in most patients. The scans were performed during deep inspiration. Forty-five studies demonstrated decreased pulmonary artery enhancement, with aortic enhancement equal to or greater than the pulmonary arteries. In 39 of these patients who underwent echocardiography either ASD (N=3) or patent foramen ovale (PFO) (N=36) was confirmed. The significantly lower pulmonary artery enhancement in these patients resulted in 27 of the 45 patients having studies with insufficient vessel contrast for the diagnosis of PE. The authors describe how deep inspiration and power injection of contrast could contribute to an increased right atrial pressure, and right-to-left shunting from right to left atrium. The authors suggest that echocardiography prior to CT could identify the patients with such shunts, and CT could be performed during shallow respiration or end-expiration. Alternatively, a longer scan delay might result in imaging during contrast recirculation, or the rapid acquisition times afforded by multislice CT could enable a short breath-hold acquisition during expiration. The authors also report the importance of recognizing the presence of a PFO in a patient with pulmonary embolism, because of the higher rate of complicated hospital course due to stroke or arterial embolism, as well as a higher rate of death.
- Summary: This is a case report describing a 77 year old woman with a history of iodinated contrast allergy and renal failure. She underwent gadolinium-enhanced spiral CT (Gadodiamide, 0.4 mmol/kg, 2 ml/s, delay 18 s) Correlative studies including lung scintigraphy and MRI were non-diagnostic. The CT study demonstrated a large thrombus located proximally and in a segmental artery of the right lower lobe.