The Dean Effect: An Aortic Arch Flow Artifact Mimicking Dissection

Abstract

The unique hemodynamics of the aortic arch create conditions for potential formation of a flow-related artifact that mimics disease on CT angiographic images. The hemodynamic basis for this artifact can be explained by fluid mechanics incorporating a mathematical principle known as the Dean number. Therefore, in this review, the artifact is referred to as the Dean effect. It is important for radiologists and other clinicians to recognize this artifact when encountered. It is also helpful for the interpreting radiologist to have a basic understanding of the relevant hemodynamic principles. This review provides an example of the artifact, reviews the basic underlying hemodynamics, and presents methods of how to prevent this artifact and distinguish it from pathologic mimics in clinical practice.

Keywords: CT Angiography, Vascular, Thorax, Aorta, Artifacts, Blood, Dissection, Hemodynamics/Flow Dynamics

© RSNA, 2022

Summary

The case presented in this report depicts an important artifact that should be recognized when encountered, in order to prevent misdiagnosis.

Introduction

The unique structure of the aorta and its related hemodynamics create the potential for a flow-related artifact at CT angiography (CTA) that may be mistaken for disease. The relatively large-diameter curved tubular structure of the aorta forms a nearly U-shaped, greater than 180° turn in the thorax. This extreme curvature is unique in human vascular anatomy. Under certain conditions, these characteristics can result in incomplete mixing of intravenous contrast material with two distinct flow laminae of contrasting CT attenuation visible on CTA images of the aorta. It is important for imaging interpreters to recognize this artifact and its causes to avoid misdiagnosis as aortic disease. The purpose of this article is to provide a case report example of the artifact at CTA, review the basic hemodynamics responsible, discuss underlying clinical pathologic and technical imaging factors, and present methods of how to prevent this artifact and distinguish it from pathologic conditions.

Case Presentation

A 71-year-old woman with history of hypertension, atrial fibrillation, and congestive heart failure with an ejection fraction of 20% was initially hospitalized for 3 weeks due to severe flu-like symptoms. After discharge, the patient experienced continued symptoms of shortness of breath, dyspnea on exertion, and orthopnea and presented again to the emergency department. She was found to be in atrial fibrillation, and chest radiographs revealed findings of pulmonary edema. Transthoracic echocardiography and chest CTA were performed. Echocardiography helped confirm a dilated and globally hypokinetic left ventricle with an ejection fraction of 20%–25%. Chest CTA revealed a curvilinear hypoattenuation along the anterior inferior descending thoracic aortic arch and aorta (Fig 1), initially interpreted as dissection versus flow-related artifact. The patient's condition deteriorated, requiring transfer to a tertiary care center for advanced critical care. CTA was repeated after transfer and revealed no evidence of dissection (Fig 2).

Figure 1:

The “Dean effect” artifact in a 71-year-old woman with increasing dyspnea and a history of hypertension, atrial fibrillation, mitral regurgitation, and congestive heart failure (New York Heart Association grade IV) with an estimated ejection fraction of 20%. One sagittal (left) maximum intension projection and three axial (right; 3.0-mm section thickness) contrast-enhanced (Omnipaque 350; GE Healthcare) CT angiographic images of the thorax through the level of the aortic arch in a narrowed and decentered soft-tissue window (convolutional kernel l40) demonstrate a flow-related artifact with distinct flow separation. This results in a heterogeneous hypoattenuation along the inner curvature of the arch due to incomplete mixing of unopacified blood with contrast material. The sagittal plane best demonstrates that the finding is artifactual.

Figure 2:

Sagittal contrast-enhanced CT angiographic (CTA) image through the level of the aortic arch from the repeat chest CTA examination performed in the same 71-year-old patient the next day. This image demonstrates resolution of the artifact and confirms a normal aorta.

Imaging Artifact: The Dean Effect

This unique flow artifact (Fig 1) occurs most often in the aortic arch and proximal descending aorta on thoracic CTA images obtained in the arterial phase. Non–contrast-enhanced blood appears to eddy along the inner curvature of the arch, while faster-flowing contrast-enhanced blood flows along the outer curvature of the arch, into the descending aorta and arch vessels. These two fluid columns create distinct laminae of contrasting CT attenuation. The consequent artifact may be mistaken for aortic dissection, intramural hematoma, aortic intimal thrombus, or neoplasia (aortic intimal sarcoma). Another example of this artifact has been described in a previous publication by Henry et al (1). Similar artifacts have also been described in the pulmonary circulation, in cases of dilated and tortuous aortic aneurysm, and in association with extracorporeal membrane oxygenation cannulation (1–4).

Hemodynamics

The dynamics of fluid flowing through a curved tube were first described using a mathematical principle known as the Dean number (5–7). This principle was named for British scientist W.R. Dean, who published the theoretical solution for laminar flow of fluid through a curved pipe in 1927 (5,6). Although this principle does not account for pulsatility, to our knowledge it is the closest experimental model to the hemodynamic conditions in the aorta. Applying these somewhat simplified principles provides a plausible explanation for the cause of this artifact and allows for the identification of methods to reduce or eliminate it.

According to Poiseuille law, when fluid flows through a tubular structure, such as a blood vessel, it develops concentric laminar flow due to physical interactions with the vessel wall (8). In a curved tube, this process becomes more complicated. At the proximal aspect of the curvature, differential flow velocities begin to develop through the cross section of the tube. The greatest fluid velocity occurs near the outer radius of the curvature, while the slower flow occurs near the inner radius of the curvature (7). Similarly, in the case of the aorta, the fastest blood flow occurs near the outer, cephalad aspect of the curvature, and the slowest occurs near the inner, caudal aspect (9–13) (Fig 3). Notably, these unique hemodynamics have also been implicated in the process of asymmetric atherosclerosis at the inner curvature of the aortic arch (14).

Figure 3:

Figure reproduced from Janiczek et al demonstrates differences in through-plane velocities at the inner and outer curves of the murine aortic arch at three-dimensional phase-contrast MRI. The four cross sections (i-iv) are denoted by the blue rectangles transecting the aortic arch. The outer radius of the aorta is at the superior aspect of the lumen in all sections (i-iv). This demonstrates that higher velocities occur along the outer radius of the arch curvature, and lower velocities occur along the inner radius of the arch curvature, contributing to the development of this artifact. (Reproduced, with permission, from reference 9.)

As flow through a curved tube continues, these concentric layers of laminar flow with differential velocities form two distinct counter-rotating channels of flow laminae, known as Dean vortices (5–7) (Fig 4). Subsequently, secondary effects develop that disrupt laminar flow and result in mixing of the fluid (5–7). The peripheral flow laminae within each vortex are swept toward the outer edge of the lumen and then toward the inner curvature, before rapidly mixing in the center of the lumen. This mixing process, as depicted by artist illustration in Figure 5, is known as a perturbation of laminar flow and is the primary cause of contrast material–enhanced blood mixing with unopacified blood in the mid to distal aortic arch (7–13). As the work of Dean explained, this perturbation has a direct relationship with flow velocity (5–14).

Figure 4:

Illustration of “Dean vortices” in a cross section of fluid flowing through a curved tube, with added velocity color coding detailing the superimposed process of laminar flow perturbation. The highest-velocity fluid is depicted in red (nearest the outer curvature), and the lowest-velocity fluid is depicted in blue (nearest the inner curvature). Again, the arrows indicate the vector of rotation within the flow laminae comprising each counter-rotating Dean vortex. In these images, the vector of net flow is into the plane of the image. (Image courtesy of A. Kalpakli, R. Örlü, and P. H. Alfredsson, KTH Mechanics, Royal Institute of Technology, Stockholm, Sweden.)

Figure 5:

Artist's conceptual illustration of the proposed process of flow perturbation and mixing of blood as it flows from the proximal aorta (superiormost cross section) to the descending aorta (inferiormost cross section). The highest-velocity flow (red) occurs at the outer curvature, while the lowest-velocity flow (blue) occurs at the inner curvature. This unique separation of differential flow velocities is the hypothesized origin of the artifact. As the Dean vortices form (second cross section) and begin to experience flow perturbation and mixing (third cross section), there is eventual flow velocity homogenization and a return to more normal laminar flow as the blood enters the straight tube of the descending aorta (fourth cross section). (Reprinted, with permission, from Aletta Ann Frazier, MD)

Under normal physiologic conditions, this process would occur very quickly, with early homogenization of blood and intravenous contrast material, resulting in uniform contrast opacification of the vessel lumen at imaging. However, in some cases, particularly in patients with very low cardiac output, the mixing process would be prolonged and exaggerated (7–13). This is the presumptive cause of an artifact at thoracic CTA of the aorta (Fig 1). This apparent eddying effect within the aortic arch and proximal descending aorta becomes visible at early and arterial phase imaging, a phenomenon that can be described as the Dean effect. The relatively rapid acquisition rate of modern scanners may increase the likelihood that this artifact is captured at imaging.

Other factors that may alter the timing of contrast material mixing due to their effect on the Dean number include vessel diameter, the radius of the curvature of the aortic arch, and blood viscosity (7–13). For a given cardiac output, the Dean effect would be most pronounced in conditions that create a low Dean number, such as when vessels are relatively small in diameter and large in curvature radius (7–13). This would also be the case if the blood viscosity was increased, such as in the setting of dehydration or polycythemia (7–13). Conversely, if a patient is anemic or has recently received very-high-volume fluid resuscitation (thus reducing blood viscosity), the Dean number will be higher and mixing will occur more rapidly, making it less likely the artifact will become visible (7–13). Although these factors are much less likely to be clinically relevant than flow velocity, which is the dominant variable, they may still be contributory (5–13,15).

Distinguishing Artifact from Disease

The differential diagnosis for the Dean effect may include aortic dissection, intramural hematoma, aortic intimal thrombus, or neoplasia (aortic intimal sarcoma). The finding becomes especially difficult to distinguish from pathologic conditions in victims of high-velocity trauma, when the index of suspicion for intimal injury and aortic dissection are high. There are several available options to help distinguish this artifact from true disease. First, it is important to evaluate the finding in all planes, especially the sagittal plane (Fig 1). The sagittal plane will best depict the layering of contrast material and is usually sufficient to reliably distinguish the artifact from pathologic conditions.

In cases where uncertainty persists despite evaluating the abnormality in all planes, a repeat CTA with a slightly longer delay will usually demonstrate resolution of the artifact (Fig 2). To ensure that the CTA is performed after mixing is complete, consider utilization of delayed bolus tracking with the region of interest centered in the descending, rather than ascending, thoracic aorta. If this is unsuccessful, a repeat scan when the patient is more hemodynamically stable and cardiac output has improved should increase the likelihood of artifact resolution. Alternatively, cine images obtained with retrospectively electrocardiographically gated CTA could potentially show the transience of the artifact during the cardiac cycle. If the patient's condition is suitable, there is a potential role for MR angiography or steady-state free precession cine of the aorta, either of which would likely demonstrate absence or resolution of the artifact.

Conclusion

The Dean effect is an uncommon aortic arch flow artifact at CTA of the thoracic aorta that may be mistaken for aortic pathologic conditions. This artifact occurred in a patient with a very low cardiac output, which is presumably the cause, either due to a severely reduced cardiac ejection fraction and/or very low heart rate. Imagers should be familiar with the appearance and cause of this artifact, the differential diagnosis, and methods of reliable distinction to prevent misdiagnosis and unnecessary workup.