Proximal Isovelocity Surface Area and Flow Convergence Methods

Published on 21/06/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2359 times

16 Proximal Isovelocity Surface Area and Flow Convergence Methods

The proximal isovelocity surface area (PISA)/flow convergence technique is an accepted quantitative measure of both valvular regurgitation and stenosis. Although it can be applied to any valve, subvalvular lesion,¹ valve prosthesis,² or any other structure with an orifice (e.g., a ventriculoseptal defect³), the PISA technique is used principally to assist in determining the severity of mitral regurgitation (MR), mitral stenosis (MS), and aortic insufficiency (AI) when other methods are less concordant and appear less sound.

The shortcomings of color Doppler flow mapping to determine the severity of valvular insufficiency are numerous and have been repeatedly characterized.⁴^–⁶ Although MR color Doppler jet size (area and length) predict angiographic grade, they exhibit a weak correlation with regurgitant volume (R_Vol) and do not predict hemodynamic dysfunction.⁷ In some lesions, such as functional/ischemic MR, color Doppler flow mapping tends to systematically overestimate the severity of mitral insufficiency; in fact, most jets larger than 8 cm² do not correspond to severe MR, advancing the concept of the need for quantitative determination of the severity of mitral insufficiency.⁸ Eccentric jets of MR correlate much less well with severity of MR⁹ due to complex spatial redistribution and loss from frictional forces.¹⁰ The effect of general anesthesia on the severity of mitral insufficiency is profound: more than half (51%) of patients with moderate to severe MR improved by at least one severity grade when assessed by transesophageal echocardiography under general anesthesia.¹¹ In the postoperative state, PISA determination of grade of MR correlates far better with angiographic grade of MR (r = 0.89 and 0.92, P < 0.001) than does color Doppler flow mapping determination of severity (r = 0.44, P < 0.1).¹² Given essentially perfect specificity (100%, positive predictive value: 100%),¹³ the finding of upper vein pulmonary venous flow reversal is the single most useful parameter to determine that MR is severe, but is limited by imperfect transthoracic sampling (reducing sensitivity: 82%),^13,¹⁴ and occasionally by the effect of highly eccentric jets or massive atrial compliance. The single most common scenario in which the PISA technique is applied is in describing the severity of MR when color flow mapping is confounded by severe jet eccentricity and the pulmonary venous spectral tracings are confounded by poor quality.

The PISA method arises from the suitability of color Doppler flow mapping to depict the hemodynamic phenomenon of flow convergence as fluid is pushed toward and through a restrictive orifice (one that imparts a pressure gradient). As blood is forced toward, and then through, a restrictive stenotic or regurgitant orifice it accelerates progressively toward its maximal velocity within its tightest stream—the vena contracta. The phenomenon of flow convergence, coupled with the versatility of color Doppler flow mapping, lends itself to the depiction of volumetric flow across a restrictive orifice, because by color Doppler flow mapping, a series of concentric “isovelocity” rings or hemispheres are depicted over the area of convergence. The greater the flow rate/volume and the smaller the orifice, the larger the flow convergence and acceleration.

Flow acceleration occurs within a hemisphere before the orifice, largely independently of the shape of the orifice, which eliminates one of the most common variables encountered in valve disease. The greater the flow volume, the larger the hemisphere of flow acceleration and the greater dimension of the concentric isovelocity rings. Hence, the dimension of the isovelocity rings depicts the flow rate: a large PISA is consistent with a large flow rate. Optimal hemispheric depiction by color Doppler occurs when the contour velocity is approximately 5% to 10% of the orifice velocity.¹⁵

The hemisphere of flow acceleration is oriented in line with the orifice; the base of the hemisphere sits on the orifice. As many orifices are oblique to the valve structure, the hemisphere may be oblique or very oblique to the angle of imaging, which engenders difficulty in recording accurate peak velocity and velocity time integral (VTI), which are needed for subsequent calculations.

In many cases, the full hemisphere of flow acceleration cannot form because physical structures are so close to the orifice that they deny (“constrain”) the formation of a geometric hemisphere. Isovelocity mapping constraint occurs commonly: in organic MR, as with mitral valve prolapse and flail leaflets; in mitral stenosis, should the diastolic shape of the valve leaflets yield a cone, as invariably happens when subvalvar disease predominates; or in aortic stenosis, as the walls of the left ventricular outflow tract confine the isovelocity rings. In such cases, applying the usual PISA method yields less accurate or inaccurate results. “Angle correction” has been proposed as a remedy for cone-shaped orifices, which are common in mitral stenosis (the orifice area calculation is multiplied by the oblique angle of the orifice [in degrees] divided by 180). The correction often is feasible for MR and mitral stenosis, but less so for aortic stenosis. Without angle correction, in the presence of constraining walls, there is significant overestimation of flow when a hemispheric model is used.¹⁶

The extent of convergent flow is readily depicted and described using color Doppler flow mapping, by measuring the dimension of the hemisphere formed from the blood flow. As the blood accelerates toward the orifice, velocity aliasing occurs and a distinct two-color (mostly red-blue) interface occurs at the boundary of the shell. At this interface the velocity is equivalent to the aliasing velocity, which is represented by the color scale. The ability to select color Doppler flow mapping parameters, such as the aliasing velocity, affords the ability to optimize the depiction of the hemisphere of flow convergence, and thereby the parameters needed to calculate aliasing flow velocity and the dimension at which flow velocity aliases. The ready means to adjust the baseline aliasing velocities makes it possible to optimize the velocities of the PISA concentric isovelocity rings (usually by lowering the aliasing velocity) by shifting the baseline down, but also to allow somewhat of a constant over the aliasing limit for the mathematical calculation of the descriptors of severity of regurgitation—the effective regurgitant orifice (ERO) and the R_Vol. The same technique can be used to determine the orifice of a stenotic lesion.

PISA method parameters needed for the equations that determine the oriface area and R_Vol include the following:

The radius of the first aliasing hemisphere (r)

The aliasing velocity (V_a)

The peak velocity across the orifice (V_max)

The VTI of flow across the orifice

Measuring the radius of the first aliasing hemisphere is the single most difficult aspect of the PISA method, and should be the focus of attention and time. As the measurement is squared, error compounds rapidly; hence, optimizing the image and measurement is critical. Identification of valve plane (by two-dimensional echocardiography) is critical because the PISA measurement is from the aliasing velocity to the valve orifice.

Proximal Isovelocity Surface Area Scanning Parameters

The parameters used in PISA calculations to describe flow through a restrictive orifice are discussed in the following sections.

Two-Dimensional and Grayscale Imaging

The level of the orifice is one means by which it is possible to delineate the inner aspect of the PISA hemispheric shell and its radius. Neither the structure with the orifice nor the orifice itself is readily seen on the color Doppler flow-mapped image. A separate grayscale image is needed. Some software allows the color Doppler flow mapping to be toggled off and on from the image, or side-by-side display of the image with and without color flow mapping, which is very useful. Zoom views are needed.

Often the plane of tissue in which the orifice resides has a dimension of depth. Usually, the orifice is at the deeper rather than the most superficial aspect of the tissue.

Color Doppler Measurements

Continuous Wave Doppler Measurements

It is critical that the angle of insonation be within 20 degrees of parallel to ensure that measurement error is within 10%:

where f_D is the Doppler shift of reflected ultrasound, f₀ is the transmitted frequency, v is the blood or tissue velocity, c is the sound velocity in tissue, and α is the isonation angle between the ultrasound beam and the direction of motion (velocity vector).

The peak velocity of the continuous wave spectral Doppler flow profile is obtained by measuring with calipers. Determ ination of the peak velocity requires sufficient signal and adequate signal-to-noise ratio to have the actual peak depicted. Often the true peak is unclear or not depicted unless the amount of signal is adequate. Hence, for mild and moderate grades of insufficiency, continuous wave (CW) parameters may be weak.

The VTI of the continuous wave spectral Doppler flow profile is obtained by edge-tracing the spectral “envelope.”

Incomplete spectral profiles are a common problem for mild and moderate grades of insufficiency.

Proximal Isovelocity Surface Area Equations

Derivations of the Proximal Isovelocity Surface Area

Aliasing velocity is calclulated in the direction of flow at the radial distance r:

VFR, or simply flow (cc/sec, mL/sec) = 2π ×

× V_a(cm/sec)

R_Vol(cc) =

where V_a = distance from the orifice level or vena contracta; V_max = peak velocity of the regurgitant jet; VTI = velocity time integral of the mitral regurgitant jet; VFR = volumetric flow rate; ERO = effective regurgitant orifice; and R_Vol = regurgitant volume. (When measuring peak mitral regurgitation velocity in m/sec, convert it to cm/sec.)

Regurgitant volume (R_Vol; cm³, mL):

where ERO is in cm² and VTI is in cm.

Although the equations are straightforward, unless the PISA technique is carefully and consistently applied, and applied with awareness of limitations, PISA determinations of MR severity may be discordant with other determinations, and add little.

The second best application of PISA methods is in the assessment of mitral stenosis, where, as with mitral insufficiency, the alignment for imaging and Doppler sampling generally is better than they are for aortic valve disease.

Transesophageal echocardiography often provides optimal depiction of flow convergence of mitral stenosis and insufficiency jets. However, due to the limited means of transesophageal echocardiography to align sampling with the flow, sampling may be suboptimal or inadequate; this may apply to all left-sided flow disturbances (MR, mitral stenosis, aortic insufficiency, and aortic stenosis). Heavily calcified mitral leaflets may diminish the depiction of MR flow convergence patterns on the far side of the mitral leaflets due to shadowing.

Simplified Methods

Adapted/simplified methods have been proposed and validated for use in situations where some needed parameters, such as the peak velocity and VTI of the CW spectral display, are poorly determined due to either insufficient signal or sampling malalignment. The technique of Rossi et al.¹⁷ uses the generally tight relationship of the peak velocity and VTI, which they observed among 272 patients, to average 3.25 (mean ± SD, 3.25 ± 0.47), allowing an estimated flow rate calculation:

This technique was demonstrated to have excellent correlation with Doppler and volumetric reference standards for estimation of R_Vol (r = 0.96 and 0.97, SEE = 11 mL [for both]; P < 0.001). This technique was noted to overestimate very large R_Vol and to be less suitable for MR due to mitral valve prolapse.¹⁷

The simplest and fastest method to use for the assessment of mitral insufficiency assumes a peak mitral insufficiency velocity of 500 cm/sec and requires the aliasing velocity to be set at 40 cm/sec. An isovelocity radius of greater than 1 cm establishes a regurgitant orifice of 0.5 cm² (severe).

Technical points on proximal isovelocity surface area method

By convention, although not without recognized limitations, PISA radius is measured at mid-systole. The convention is based on the rationale that consistently making all measurements (i.e., radius, peak velocity) at mid-systole provides the best correlation for instantaneous assessment. In most pathologies, regurgitant flow increases rapidly in early systole and is maximal by mid-systole; therefore, maximal flow is identified. However, given the differing pathologies responsible for regurgitation, some of which provide dynamic orifices, mid-systolic flow rate may or may not be maximal and representative of average flow. For example, in mitral valve prolapse, mid-systole may not actually be the time of peak regurgitation—progressive prolapse through systole may increase the ERO area progressively until it is maximal in late systole. Similarly, functional MR, as may happen with cardiomyopathy, may lessen progressively through systole as the ventricular volume diminishes, allowing better approximation of the mitral leaflets and a smaller regurgitant orifice. From the point of view of orifice stability, the prototypic lesions of MR are rheumatic MR and MR due to perforations, where the orifice is essentially constant throughout systole. However, in many adult patient populations, functional MR and MVP greatly outnumber the rheumatic and endocarditic cases.^18,¹⁹ With holosystolic and central jet MR the PISA technique is appropriate, whereas with late systolic and eccentric jet PISA is problematic.

Ideally, the PISA would be sampled at numerous times through the flow interval to account for differing flow rates that may occur.¹⁹ The flow convergence radius should be measured on three different cardiac cycles.

Most studies of PISA technique in valvular disease have compared PISA determinations of ERO area to reference standards of quantitative Doppler and quantitative 2D techniques.

The PISA technique tends to overestimate ERO area, especially when it is large.²⁰ Other factors that may incite error are, not surprisingly, poor-quality PISA hemispheres, which greatly lessen the accuracy of the calculations. MVP often affords a dynamic orifice that may reduce the accuracy of calculations, as mid-systolic flow rate may not be the maximal flow rate. Among patients with optimal depiction of flow convergence, the correlation of PISA determination of ERO area with those obtained by qualitative Doppler and quantitative two-dimensional methods is excellent.²⁰ Table 16-1 presents a grading scheme for judging the severity of valvular insufficiency.

Proximal Isovelocity Surface Area: a Summary

The PISA technique enables quantification of regurgitant or stenotic flow rate, orifice area, and R_Vol.

Scrupulous technique is critical to optimize acquisition of the variables needed to make calculations:

• Zoom views of PISA with narrow sectors

• Optimized aliasing velocities

• Scrolling among views to detect optimal images

• Maximizing the completeness of CW spectral tracings

• The aliasing radius should not be measured in highly eccentric/oblique PISA along the lateral margins, because this will underrepresent flow.

• 3 to 5 measurements of the radius are taken, from the average of the outer rim to the underside/base of the hemisphere

Several assumptions are implicit in the use of the technique, such as hemispheric flow acceleration across a planar orifice.

Overestimation of ERO and R_Vol by the PISA technique may occur for several reasons:

• Overestimation is inherent in the technique, especially for severe MR.

• The effect of wall constraint ¹⁶

• Dynamic ERO

Constraint by nearby structures devalidates the assumption of hemispheric flow model, and confers a “constraint angle” that should prompt the use of angle correction. Complex constraint geometry renders application of the technique unwise.

In the setting of multiple regurgitant orifices, assessments of the multiple PISAs may add up to the cumulative effect, but provide a tiresome challenge.

Highly eccentric MR may afford the parasternal long-axis view the means by which to assess both aliasing radius and peak velocity.

Atrial fibrillation, common in mitral valve disease, renders PISA determinations of MR volume less accurate. In atrial fibrillation, more than 5 cardiac cycles (7–10) are needed to ascertain average MR.

Simplified versions of the equations may be used:

• R_Vol calculation using 3.25 when the CW spectral recording of flow is inadequate

• A PISA radius >1 cm at an aliasing velocity of 40 cm/sec (assuming standard MR velocity) yields severe MR, as the ERO is 0.5 cm².

The PISA technique has contributed to the understanding of the variable and often complex nature of some orifices (Table 16-2).

TABLE 16-1 Severity of MR by PISA Method

TABLE 16-2 Proximal Intervelocity Surface Area Method Summary: Pros and Cons

	Pros	Cons
Mitral Regurgitation	• Heavily validated ^4,²⁰^–²² • The principal application of PISA techniques is in the resolution of severity of mitral insufficiency (especially distinguishing ² moderate from severe), particularly when color Doppler flow mapping techniques are unreliable (e.g., eccentric wall impacting jets or multilobulated jets, and pulmonary venous sampling is poor quality). • Uncertainty about MR severity is common, and PISA is a useful adjunctive method. • Often as good by TTE as by TEE

• Technical factors may render the technique inapplicable.

• The PISA technique for MR severity determination suffers from

• Technically poor color Doppler depiction of converging hemispheres

• Difficulty in assessing multiple regurgitant orifices

• Very oblique orifices (which are common in myxomatous disease—flail and prolapse lesions). Even if the PISA hemisphere is well depicted, CW interrogation may be significantly misaligned, leading to undersampling of true peak velocity and VTI.

• Aliasing velocities of <20 cm/sec are less reliable, and would require use of regression equations.²³

• There is a systematic tendency to overestimate regurgitant flow when it is more severe.²²

• The need to measure several radii

Mitral Stenosis

• Validated ²⁶^–²⁹

• Often very clear PISA depiction by TEE; thickened rheumatic mitral leaflets may diminish the transthoracic depiction of the PISA.

• PISA estimates of MVA are not influenced by the presence of aortic regurgitation, mitral regurgitation or atrial fibrillation.²⁸

• Standard techniques to determine MVA are relatively robust, and TTE and TEE calculations of mitral valve gradient are very good, leaving a lesser role for adjunctive methods.

• Oblique orientation of the leaflets and the orifice may diminish feasibility.

• Cone-like orifices are a problem, as the standard PISA equation assumes a planar orifice. “Angle correction” should be considered (orifice angle/180 degrees × ERO), although not all studies demonstrated its benefit.²⁸

• Most MS orifices are long and slit-like, and are less suitable for the PISA method.

• Less accurate in the presence of atrial fibrillation

• Less accurate in the presence of MR

Aortic Insufficiency

• Validated ³⁰

• Feasible by both TTE and TEE; often TEE provides better color Doppler flow mapping.

• The aortic valve is relatively deep for transthoracic imaging, and PISA depiction often, but not always, suffers from poor image quality.

• TEE does not provide adequate alignment for CW assessment.

• The PISA technique fares poorly with eccentric SI jets

Aortic Stenosis

Minimally validated

• Standard techniques to determine AVA are relatively robust, and TTE calculations of aortic valve gradient are very good, leaving a lesser role for adjunctive methods.

• PISA hemispheres within the LVOT generally become confined/constrained by the walls of the LVOT and distorted into nonhemispheres, devalidating the use of hemispheric models.

• Concurrent subvalvar, LVOT, or intracavitary gradients will contaminate the isovelocity maps.

AVA, aortic valve area; CW, continuous wave (Doppler); ERO, effective regurgitant orifice; LVOT, left ventricular outflow tract; MR, mitral regurgitation; MVA, mitral valve area; PISA, proximal isovelocity surface area; SI, stroke index; TEE, transesophageal echocardiography; TTE, transthoracic echocardiography; VTI, velocity time integral.

Figure 16-1 Schematic rendering of the convergence of flow (downward in the example) through an idealized planar restrictive orifice. Unless the orifice is severely asymmetric or there is impingement by nearby structures, flow accelerates symmetrically to cross through the orifice. At a given radius from the orifice, the flow velocity is constant (i.e., there are “isovelocity” lines at a given radius). The graded color mapping display lends itself to depicting an isovelocity line at the first aliasing of the flow in the direction of flow, yielding a critical variable for calculation of flow rate, orifice area, and flow volume. The actual occurrence of adequate alignment, a fully planar orifice, and proximal isovelocity surface area shells with little variation to the radius is a matter of chance, even with optimization of technique.

Figure 16-2 Continuous wave Doppler recording of flow across the orifice of interest is a needed variable for orifice area and volume calculations. It is often challenging, however, to achieve optimal alignment and a complete and well-defined spectral profile that is a complete parabola.

Figure 16-3 An uncommonly well-formed proximal isovelocity surface area of aortic insufficiency obtained from the apical three-chamber view and by baseline shift.

Figure 16-4 Well-formed proximal isovelocity surface area (PISA) of mitral insufficiency. The images display the shell (left) more uniformly and the base (right) of the PISA in a better defined way.

Figure 16-5 An optimized proximal isovelocity surface area view of mitral insufficiency, which at first glance appears sufficiently well-defined to measure, but upon scrutiny has at least 30% to 40% variation in the apparent radius.

Figure 16-6 Continuous wave Doppler spectral display of sampling of mitral insufficiency. It is only with the last image that the actual peak of the profile is clearly displayed.

Figure 16-7 Proximal isovelocity surface area (PISA) image of mitral insufficiency in early (left), mid- (middle) and late (right) systole. By (imperfect) convention, PISA is measured at mid-systole, which is not always readily established by the electrocardiographic tracing and is better established by scrolling through the cycle. Note also the wide variation of the dissipation jet area within the left atrium.

Figure 16-8 Rheumatic mixed mitral stenosis/insufficiency. From the parasternal long-axis view, the proximal isovelocity surface area (PISA) of the mitral stenosis (MS) is superbly defined, even without baseline shift, whereas the PISA of the mitral insufficiency (MI) is poorly defined. From a zoomed apical 4-chamber view with baseline shift (left middle image), the PISA of the PR is well-defined. The continuous wave spectral tracing of the MR is disappointing, never clearly yielding the peak of the profile. By calculation, the ERO is 60 mm², which qualifies as severe. The spectral display of pulsed Doppler sampling of the pulmonary veins does not yield systolic flow reversal, a standard sign of severe MI; the size of the left atrium may have rendered it a sufficiency capacitor to have obviated pulmonary venous flow reversal. The atrium and appendage are clearly enlarged, which, in the context of mitral insufficiency, supports the finding that the grade is severe; however, other parameters also must be considered, such as MS and atrial fibrillation, both of which were present and were contributory to the left atrial dilation. Although parameters are discrepant the PISA calculation augurs that the MR is severe, even if the pulmonary venous profile does not.

Figure 16-9 Side-by-side displays of optimized (zoom view and baseline shift) proximal isovelocity surface area (PISA) of mitral insufficiency with the underlying 2D grayscale views beside the color Doppler flow-mapping views. The upper views are representative of mid-systole and the lower views of late systole. Determination of the outer margin of the “shell” of the PISA is established from visual analysis of the color Doppler display. The inner aspect of the radius is best determined from scrutiny of both the color Doppler display and the 2D grayscale views, where, it is hoped, at least one of the two will convincingly depict the level of the orifice.

Figure 16-10 Constraint and angle correction, which can be more conceptually feasible than practical. The upper images depict mitral stenosis: the ideal, planar orifice affording 180 degrees of hemispheric convergence (left); the ideal planar conical model of constraint affording a lesser angle for convergence (simple, readily measured constraint angle; middle left); dome-like constraint yielding varying angles of convergence at different radii away from the orifice (middle right); and complex curved constraint, as with subvalvar disease and thickened leaflet levels—the curves of the leaflets and the subvalvar apparatus inflex in opposite directions, yielding varying angles of constraint at different radii (right). The first three of the lower images depict mitral insufficiency. The conceptually ideal planar orifice of mitral insufficiency yielding 180 degrees of hemispheric convergence (left); reverse conical constraint, as with functional mitral regurgitation and tenting of leaflets—there is more than 180 degrees of convergence—a situation for which there is no available model to correct for (middle); complex constraint due to a convergent orifice and nearby constraint from the lateral wall of the left ventricle (right). Determining an angle of constraint giving the opposing curves is difficult. The final (far right) lower image depicts aortic stenosis with constraint from the interventricular septum and the anterior mitral leaflet. The zone of convergence in this case is not well modeled by a hemisphere or partial hemisphere.

Figure 16-11 Correlations of estimated (y-axis) and calculated (x-axis) regurgitant volume (R_Vol) using quantitative Doppler (left) and quantitative two-dimensional echocardiography (2-D echo; right).

(From Rossi A, Dujardin KS, Bailey KR, et al. Rapid estimation of regurgitant volume by the proximal isovelocity surface area method in mitral regurgitation: Can continuous-wave Doppler echocardiography be omitted? J Am Soc Echocardiogr. 1998;11:138–148. Used with permission.)

Figure 16-12 Correlations between the effective regurgitant orifice (ERO) area obtained by the proximal isovelocity surface area (PISA) method and by quantitative Doppler echocardiography (left) and quantitative two-dimensional echocardiography (2 DE; right). Circles, patients with optimal flow convergence; dashed line, identity line; solid line and gray zone, regression line and 95% confidence interval in patients with optimal flow convergence, respectively; squares, patients with nonoptimal flow convergence.

(From Enriquez-Sarano M., Miller F.A Jr.., Hayes S.N., et al. Effective mitral regurgitant orifice area: clinical use and pitfalls of the proximal isovelocity surface area method. J Am Coll Cardiol. 1995;25:703–709. Used with permission.)