Assessing Intracardiac Vortices with High Frame-Rate Echocardiography-Derived Blood Speckle Imaging in Newborns

Summary

The present protocol uses echocardiography-derived blood speckle imaging technology to visualize intracardiac hemodynamics in newborns. The clinical utility of this technology is explored, the rotational body of fluid within the left ventricle (known as a vortex) is accessed, and its significance in understanding diastology is determined.

Abstract

The left ventricle (LV) has a unique pattern of hemodynamic filling. During diastole, a rotational body or ring of fluid known as a vortex is formed due to the chiral geometry of the heart. A vortex is reported to have a role in conserving the kinetic energy of blood flow entering into the LV. Recent studies have shown that LV vortices may have prognostic value in describing diastolic function at rest in neonatal, pediatric, and adult populations, and may help with earlier subclinical intervention. However, the visualization and characterization of the vortex remain minimally explored. A number of imaging modalities have been utilized for visualizing and describing intracardiac blood flow patterns and vortex rings. In this article, a technique known as blood speckle imaging (BSI) is of particular interest. BSI is derived from high-frame rate color Doppler echocardiography and provides several advantages over other modalities. Namely, BSI is an inexpensive and noninvasive bedside tool that does not rely on contrast agents or extensive mathematical assumptions. This work presents a detailed step-by-step application of the BSI methodology used in our laboratory. The clinical utility of BSI is still in its early stages, but has shown promise within the pediatric and neonatal populations for describing diastolic function in volume-overloaded hearts. A secondary aim of this study is thus to discuss recent and future clinical work with this imaging technology.

Introduction

Intracardiac blood flow patterns play a key role in cardiac development, starting in fetal morphogenesis and continuing throughout the lifespan¹. Hemodynamic shear stress plays a pivotal role in the stimulation of cardiac chamber growth and architecture via the activation of specific genes^2,3. This occurs at both the intrauterine stage and in the early stages of life, thus highlighting the importance of hemodynamic influence on early cardiac development and the carry-over into adulthood³.

The laws of fluid dynamics state that blood passing along a vessel wall move slower when closest to the wall and faster when in the center of a vessel, where resistance is lower. This phenomenon can be demonstrated in any large vessel with pulse wave Doppler as the typical Doppler velocity time integral envelope⁴. When blood enters a larger cavity such as the heart, the blood farthest from the endocardial surface continues to increase its velocity relative to the blood closest to that surface and create a rotational body of fluid, known as a vortex. Once created, vortices are self-propelling flow structures that typically draw in surrounding fluid via negative pressure gradients. Thus, a vortex can move a greater volume of blood than an equivalent straight jet of fluid, promoting greater cardiac efficiency^4,5.

The literature suggests that the evolutionary purpose of vortices is to conserve kinetic energy, minimize shear stress, and maximize flow efficiency^4,5,6. Specifically for the heart, this includes storing hemodynamic energy in a rotary motion, facilitating valve closure, and the propagation of blood flow toward the outflow tract, as seen in Figure 1. Altered intracardiac blood flow patterns are expected in pathological situations such as volume-overloaded states and in cases with artificial valves^7,8. Thus, herein lies the true diagnostic potential of vortices as early predictors of cardiovascular outcomes in adults.

Intracardiac hemodynamics have gained increasing interest in the literature in both adult and pediatric populations. Several modalities are available for the qualitative and quantitative assessment of intracardiac hemodynamics and were comprehensively summarised in a recent review, with a specific emphasis on the intracardiac vortex⁹. One modality with great promise is echocardiography-derived blood speckle imaging (BSI), which offers the ability to noninvasively measure a number of qualitative and quantitative vortex characteristics, described below, at a relatively low cost and with excellent reproducibility¹⁰. BSI is currently commercially available using a high-end cardiac ultrasound system with an S12 or S6 MHz probe. The speckle-tracking features are analogous to those used in tissue speckle tracking to study myocardial deformation^11,12,13. Since red blood cells tend to move faster and with a higher Doppler frequency than the surrounding tissue, the two signals can be separated by applying a temporal filter. BSI uses a best-match algorithm to quantify the movement of blood speckles directly without using contrast agents. The blood velocity measurements can be visualized as arrows, streamlines, or path lines with or without underlying color Doppler images, and can highlight areas of complex flow¹⁰.

BSI has been shown to have good feasibility and accuracy for quantifying intracardiac blood flow patterns, with excellent validity compared to a reference phantom instrument and pulsed-Doppler^7,10,11. Whilst still very novel, BSI is a promising clinical tool for the early diagnosis of various cardiac pathophysiologies. The clinical application of vortex imaging has shown promise in newborn infants. Specifically, the behavior of a vortex in the left ventricle (LV) may have long-term implications on cardiac remodeling and predisposition toward heart failure.

The mechanism linking vortices to left ventricular remodeling is still relatively unexplored, but has been recently investigated in our laboratory and is the subject of ongoing work¹¹. This methodology article aims to describe the use of BSI in exploring intracardiac vortices and discuss the practical and clinical uses of vortices in assessing diastolic function in various populations. A secondary aim is to discuss the clinical relevance of BSI and present some of the work previously performed in neonates.

Protocol

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants' families included in the study. All images and video clips were de-identified following the acquisition.

1. Patient preparation

1. Set up the ultrasound machine adjacent to the patient's cot and connect a three-lead electrocardiogram (see Table of Materials).
2. Input the patient code and relevant details, such as body length and weight, and perform the echocardiogram according to the previously described standards¹².

2. Image acquisition

1. Specifically for BSI, obtain a shallow view of the LV in the apical four-chamber view with a narrow sector width, allowing an acquisition frame rate between 400-600 Hz.
2. Open a color box over the left ventricular cavity, narrow maximally to include only the region from the mitral valve to the endocardial apex, and from the septal endocardial border to the lateral wall endocardial border.
3. Increase the color gain to the point of speckling and reduce slightly. Set the color Doppler velocity scale limit to the appropriate diastolic velocity (20-30 cm/s in preterm infants) to maximally fill the color box with the slower-moving diastolic inflow.
4. On the touchscreen control panel of the equipment (see Table of Materials), tap BSI mode to reveal the intracardiac flow directions and vortices in RAW color format. Adjust the BSI box position and size to include the flow region of interest and record at least two cardiac cycles.
5. Repeat the procedure in the apical LV long-axis view or other views where intracardiac hemodynamic assessment is required (Figure 2 and Figure 3).

3. Image analyses

NOTE: The image analysis techniques for the LV vortex have been briefly described in previous work from our laboratory¹¹. The protocol used for assessing intracardiac vortices is as follows (Figure 3 and Figure 4).

1. Save two cardiac cycles from each respective patient to external media in their RAW DICOM format and transfer to a laboratory station with an image processing software (see Table of Materials) installed for detailed offline analyses.
2. Once offline, identify the most prominent or main vortex.
   NOTE: The main vortex is visualized as an elongated, oval-shaped, anti-clockwise rotating structure located in the upper left quadrant of the left ventricle near the septum, with the maximum vortex area found in late diastole (during the transmitral A-wave) in preterm infants (Video 1). The main vortex is usually found during the transmitral E-wave for older infants and children.
3. Record the number of independent, complete oval-shaped vortices forming throughout the cardiac cycle for each clip.
4. Measure the position of the main vortex relative to known landmarks within the LV. To determine the Vortex depth, using the "distance measurement" tool on the analysis software, measure the vertical distance from the vortex eye to the middle of the mitral valve annulus. For Vortex transverse position, measure the horizontal distance from the vortex eye to the endocardial border of the interventricular septum.
5. Measure the vertical and horizontal edge-to-edge distances of the main vortex relative to the LV length and width to obtain the vortex shape.
   NOTE: This also enables estimation of the vortex sphericity index as length divided by width.
6. Using the "tracing measurement" tool on the analysis software, click on and trace the outermost vortex ring at the point where the main vortex is most prominent to determine the main vortex area.
7. To assess Peak Vortex Formation Time (PVFT), record the cardiac frame when the vortex first appears (circular rings delineated) in the cardiac frame where the main vortex is most prominent and calculate the number of frames relative to the total number of frames in one cardiac cycle for the patient.
8. To assess vortex duration, measure the frames from which the vortex first appears when the vortex loses its circular ring formation. Vortex duration is then calculated as the number of frames relative to that patient's total number of frames in one cardiac cycle (Figure 5).

Representative Results

The acquisition of vortex clips is comparable to the standard methodology universally employed in obtaining color Doppler clips. Pioneering studies in adults have described vortices using the apical two-, three-, and four-chamber views¹⁴. The LV vortex is a ring-like structure that moves from base to apex. BSI visualizes the internal diameter of the ring (Figure 2). A vortex ring is usually not symmetrical in shape, hence alternative imaging planes can show variable vortex morphology or position. In a small analysis of 20 patients, it was found that the vortex position was comparable. In particular, the vortex sphericity index was higher in the four-chamber view when compared to the three-chamber view (Figure 3). The present study adopted the three-chamber view for vortex imaging, which yielded the most reproducible images in our experience.

Recent work from our laboratory has described the successful clinical application of echocardiography-derived BSI¹¹. A population of 50 preterm neonates received a comprehensive echocardiogram, which included BSI assessment, alongside traditional clinical data such as blood pressure and respiratory status. The feasibility and reliability of both the acquisition and interpretation of vortices in neonates were high and demonstrated that vortices could be described in detail based on the methodology discussed above. Specifically, a range of values for vortex area, position, morphology, number of apparent vortices, and timing characteristics, alongside traditional cardiac structure and function parameters, were identified. Furthermore, the population was subdivided into quartiles based on indexed LV volumes and demonstrated significant differences between the high and low quartile groups for various key vortex parameters (Table 1).

The analyses revealed several key associations between the novel BSI vortex parameters and traditional echocardiography-derived parameters of diastolic function and LV morphology. A strong positive correlation was seen between vortex area and LV end-diastolic dimension (r = 0.50, p < 0.01), and an inverse correlation was seen between vortex duration and the Ee' ratio - a surrogate measure of LV end-diastolic pressure¹² (r = -0.56, p < 0.01). These data suggest that vortices may provide unique insight into the diastolic function of a neonatal population and provide additional support to traditional well-established parameters.

The key associations described above between vortex area and LV morphology have prompted further ongoing work on the hypothesis that kinetic energy from intracardiac hemodynamics may influence early cardiac remodeling of the LV in preterm infants. A larger-scale prospective study has thus far revealed that at least one in four very preterm infants show signs of LV cardiac remodeling at the time of discharge. However, there is limited information available on the underlying mechanisms. Preliminary assessments found that vortices were less elongated at postnatal day 7 after preterm birth in infants who later developed cardiac remodeling, supporting the hypothesis that intracardiac blood flow patterns may play an important role in cardiac development after preterm birth¹⁵. Further studies are needed to validate these findings and explore whether early and short-term intervention can potentially prevent this pathway of abnormal cardiac development.

The application of BSI in characterizing intracardiac hemodynamics has also been explored in other cardiac landmarks where unique flow patterns are present (Figure 6). Preliminary assessments of bi-caval inflow patterns are conducted within the right atrium (Video 2) and right ventricular outflow tract hemodynamics during diastole (Video 3). These pilot studies aim to further describe patterns of venous return flow in neonates with various levels of respiratory support and gain further insight into the inter-relationships between respiratory changes and diastolic function.

Figure 1: Left ventricular intracardiac hemodynamics. This illustration visually shows intracardiac blood flow patterns and vortex formation within the LV. Please click here to view a larger version of this figure.

Figure 2: Left ventricular vortex ring. This schematic demonstrates the vortex ring seen from an apical three-chamber view using two-dimensional color Doppler and speckle tracking imaging. When using the apical three-chamber view, the main vortex (Smain 3 chamber) is smaller than the four-chamber view (Smain 4 chamber). The main vortex is usually larger when compared with any secondary vortices (Ssec). Please click here to view a larger version of this figure.

Figure 3: Blood speckle imaging-derived vortices in the apical views. This is a comparison of BSI derived vortices demonstrated using the apical four-chamber view (left) and the apical three-chamber view (right). The graphs represent the different shapes and locations of the vortices in the two apical windows. Please click here to view a larger version of this figure.

Figure 4: Assessing vortex morphology. This diagram demonstrates the manual methods used in our laboratory to obtain vortex morphology parameters from the apical four-chamber view. Please click here to view a larger version of this figure.

Figure 5: Assessing vortex timing characteristics. This figure demonstrates the methods used to obtain vortex timing characteristics such as vortex duration and peak vortex formation time. The vertical red line indicates at which stage of the cardiac cycle a vortex event occurs. Please click here to view a larger version of this figure.

Figure 6: Blood speckle tracking in other cardiac chambers. This figure shows intracardiac hemodynamics in other cardiac chambers. In the right ventricle (RV), the main vortex is a clockwise rotating structure that rolls along the septum with its maximum area just before the pulmonary valve and artery (PA). In the right atrium (RA), a main vortex is formed due to the mixing of inflow via the inferior vena cava (IVC) and superior vena cava (SVC) near the inferior border of the lateral wall and with anticlockwise rotation, and sometimes a second clockwise rotation near the RA appendage. The left atrium (LA) has limited areas where the flow of the four pulmonary veins is not directly mixing, and vortices can be difficult to capture. Please click here to view a larger version of this figure.

	LVEDVi Lowest quartile	LVEDVi Highest quartile
Ejection Fraction (%)	67(5)	69(5)
Longitudinal strain (%)	20.3(1.6)	23.5(2.7)*
MV VTI (cm)	6.4(1.9)	9.6(2.8)**
EA ratio	0.69(0.12)	0.84(0.10**
Ee' ratio	13.3(2.9)	19.7(8.0)*
IVRT (ms)	54(8)	44(8)**
Location
Vortex depth	0.58(0.10)	0.56(0.07)
Vortex transverse position	0.29(0.07)	0.37(0.15)**
Geometry
Vortex area (cm2)	0.44(0.28)	0.57(0.21)
Vortex area indexed to LV area	0.20(0.12	0.18(0.05)
Time properties
Vortex start time (% of RR)	88(5)	76(8)**
Peak vortex formation time (% of RR)	91(2)	82(8)**
Vortex duration (% of RR)	16(4)	11(2)**

Table 1: Comparison between infants with the lowest versus highest quartiles of indexed LV volumes. The data are presented as means + standard deviation (SD). **p < 0.01, *p < 0.05. Abbreviations: IVRT = isovolumic relaxation time; LVEDVi = left ventricular end-diastolic volume indexed to weight; MV = mitral valve. The table is reused from reference¹¹.

Video 1: Screen grabs from the LV vortex video. Please click here to download this Video.

Video 2: Screen grabs from the bi-caval inflow vortex video. Please click here to download this Video.

Video 3: Screen grabs from the right ventricular outflow tract vortex video. Please click here to download this Video.

Discussion

The importance of visualizing and understanding the intracardiac vortex
There are many possible clinical applications of high-frame rate echocardiography-derived vortex imaging. Their ability to provide valuable insight into intracardiac flow dynamics has been the interest of recent studies¹⁶. Moreover, vortex imaging may allow the detection of pre-symptomatic changes in LV architecture and function in neonates, which may have a bearing on long-term cardiac remodeling into adulthood¹⁵. This, in turn, may increase the accuracy and prognostic outcomes of follow-up treatments and surgeries. The use of BSI in visualizing intracardiac vortices has recently gained some traction in the literature, but remains largely unexplored. Aside from work in the present laboratory on prematurely born neonates, other clinical publications have demonstrated that echocardiography-derived BSI is feasible and clinically relevant in infants with congenital heart disease^7,17, valvular pathology¹⁸, and even in right ventricular pathology¹⁹.

The utility of vortex imaging in assessing diastolic function
The diastolic function of the LV describes its ability to fill up with blood and prepare a stroke volume for ejection. There have been major advancements in understanding diastolic function in patients with heart failure with preserved ejection fraction (HFpEF), especially in relation to pathophysiology, diagnosis, and prognosis with echocardiography^20,21. The diastolic function of the heart involves an active biochemical process of myocardial relaxation, where the actin and myosin cross-bridges detach and tension in the myocardial muscle fibers starts to reduce. When the mitral valve opens, blood enters the LV through suction created by the elastic recoil of myocardial fibers moving toward their original length (restoring forces). This lowers the LV cavity pressure and creates a pressure gradient between the atrium and ventricle. The final phase of diastolic function is generated by atrial contraction, which increases the LA pressure above the LV pressure and establishes the final LV end-diastolic pressure and volume before closing of the mitral valve and the beginning of the contraction²².

From a hemodynamic point of view, LV diastole involves the passage of a column of oxygenated blood from the atria into the ventricle in preparation for ejection. The position of the LV outflow tract adjacent to the mitral annulus means that blood enters the ventricle basally to apically and leaves the ventricle apically to basally. Recent advancements in the understanding of intracardiac hemodynamics suggest that this redirection of blood flow occurring in the transition from left ventricular filling to ejection follows a specific rotational direction to minimize shearing forces on the myocardium and conserve kinetic energy of the moving blood column, hence the formation of an intraventricular vortex^4,5(Figure 1).

Diastology guidelines have been summarised by the American Society of Echocardiography and the European Association of Cardiovascular Imaging¹². Several limitations exist with standard Doppler and two-dimensional derived assessment of diastolic function. These include, but are not limited to, heart rate, Doppler angle dependence, signal quality, and the difficulty of grading diastolic dysfunction using multiple parameters, which often do not align. Thus, the proposition of an angle and heart-rate independent parameter, with the potential for detailed insight into the relaxation and filling of the LV derived from a single primary measurement, is made possible with the introduction of the intracardiac vortex.

As revealed in the current results, visualization of the vortex yields several parameters, enabling some insight into diastolic cardiac function. Specifically, a significant association between vortex area/shape and LV morphology has been shown, as well as the relevance of vortex timing in predicting LV end-diastolic pressure. Additionally, variations in vortex position based on the imaging plane used are also seen (Figure 3), as well as positional differences in children with congenital heart disease in work from other authors (i.e., vortex positioned closer to the interventricular septum in volume-overloaded cases and patients with valvular pathology^7,12). The number of vortices seen in the LV may theoretically be related to LV architecture, but has yet to show statistical significance in this work and the work of others. Lastly, vortex imaging may give rise to more complex numerical measurements, such as vorticity, energy loss, and stored kinetic energy, which has shown some prognostic value in studying congenital valvular disease, such as bicuspid aortic valves¹⁸. The clinical application of BSI may provide feasible additional information to conventional color Doppler, helping to enhance the visualization of abnormal hemodynamic patterns in pathologies such as shunts, valvular regurgitation, and stenosis¹⁷.

Imaging and analysis of the intracardiac vortex: pros and cons
As previously described, intracardiac blood flow patterns can be visualized using cardiac magnetic resonance imaging (MRI), as well as echocardiography-derived particle imaging velocimetry, vector flow mapping, and BSI⁶. In neonates, BSI holds the greatest advantages due to its noninvasive nature and bedside application. Additionally, since image resolution and ultrasound beam penetration are inversely related, the very small body surface area of a neonate allows a high resolution to be utilized without sacrificing it for penetration depth. Conversely, as BSI requires high frame rates and resolution in order to feasibly capture intracardiac vortices, this technology is currently not able to be performed in larger patients, such as adults, where the greater penetration requirements compromise the resolution. Thus far, the largest number of patients in whom BSI was successfully applied was in a population of children with a median age of 7 years and a body surface area of up to 1.22 m²⁷.

Another limitation of BSI imaging is its dependence on high-quality two-dimensional images to estimate vortices accurately. Currently, BSI is not available in three-dimensional echocardiography, which limits the visualization of this complex three-dimensional structure. Furthermore, BSI incurs a significant loss of signal-to-noise ratio due to its limited penetration depth. In practice, this means that an unsettled neonate who is moving during the time of examination and a body structure that precludes an optimized and defined four-chamber view of the LV can form significant obstacles with this technology. Methods to calm the neonate during examination (for example, using sucrose) and other techniques for optimizing LV image quality in the four-chamber view (for example, positioning the neonate and operator techniques) should be readily implemented.

Lastly, this study was commercially limited to the vortex characteristics of the technology of choice (i.e., echocardiography-derived BSI). While the clinical relevance and reproducibility of these measurements are gaining traction in the literature, there is still a need to validate further what these markers mean in various pathologies and how they compare to other imaging modalities. For example, vortex architecture, positioning, and timing may be very useful in congenital heart disease, whilst the kinetic energy parameters, yet to be available with BSI, may serve well in long-term serial studies of cardiac remodeling.

Future directions
In summary, BSI is gaining rapid recognition as a low-cost, noninvasive, and valuable tool for assessing intracardiac hemodynamics and, more specifically, vortices. Work from the present laboratory has verified its reproducibility and demonstrated its clinical and practical utility as a supplementary tool for assessing cardiac function and remodeling after preterm birth⁸. Moving forward, the hypothesized link between intracardiac shearing forces on the myocardium and the subsequent cardiac remodeling seen at different points of early-life development requires further attention. Thus far, only architectural and temporal characteristics of vortices have been explored. However, as alluded to earlier, acquiring energetic parameters such as rotational kinetic energy and vorticity may provide further insight into the mechanism linking flow patterns and adverse cardiac remodeling. Clinically, this may subsequently allow more timely interventions to be implemented in at-risk patients.

Materials

Name	Company	Catalog Number	Comments
Tomtec Imaging Systems GmbH	Phillips	GmbH Corporation	Offline ultrasound image processing tool, used for calculating all vortex measurements
Vivid E95	General Electrics	NA	Cardiac Ultrasound device used to capture Echocardiography-derived Blood Speckle Imaging