Indications.

Results from two separate guidelines endorsed by the Society for Vascular Surgery support the use of DUS as the initial imaging modality for both asymptomatic and symptomatic patients.^1,19 Initial carotid evaluation with DUS is acceptable for patients with a carotid bruit and to observe known asymptomatic stenosis. DUS is also frequently used for screening carotid evaluation in patients with known peripheral arterial disease, coronary artery disease, or abdominal aortic aneurysms. For symptomatic patients, initial carotid evaluation with DUS is appropriate for patients presenting with amaurosis fugax or hemispheric symptoms attributable to carotid territories. Secondary imaging, such as computed tomographic arteriography (CTA) or magnetic resonance arteriography (MRA), is necessary when sonography cannot be obtained, DUS results are equivocal in symptomatic patients, or confirmation of DUS findings is necessary for quality assurance.

Technique.

A linear array transducer (5 to 10 MHz) is sufficient to image the arterial anatomy. Carotid artery imaging in gray scale is performed in transverse and sagittal planes to assess the intima-media thickness (normal, <0.8 mm), to assess for presence of atherosclerotic plaque (intima-media thickness >1.5 mm), and to investigate the site or sites of more advanced plaque stenosis. The examination should include complete DUS imaging of the extracranial common carotid artery (CCA), ICA, and external carotid artery (ECA) as well as assessment of flow in the vertebral and subclavian arteries including bilateral brachial artery systolic pressure measurement. Accurate differentiation of the ICA from the ECA is paramount. The two can be differentiated on the basis of location (the ICA is posterior and lateral to the ECA), Doppler flow resistance (the ICA has low-resistance monophasic spectra, whereas the ECA has higher resistance multiphasic spectra), and the presence of branches (the ECA has branches) (see Fig. 16-8).

Atherosclerotic plaque features, such as homogeneity or heterogeneity, are evaluated in gray scale by B-mode ultrasound imaging. GSM assessment of each plaque can provide additional information. Multiple images can be obtained in transverse views and a single longitudinal view. Two points of reference are used, with 0 designated for black and 255 for white. Excessive calcification can limit GSM assessment by blocking the ultrasound. Multiple points of assessment allow a maximum and minimum GSM, with the difference in measurements determining the heterogeneity of the plaque.

Notation of the carotid bifurcation relative to the angle of the mandible provides clinically important information in determining difficulty of surgical exposure for lesions at or above the angle of the mandible. Color Doppler imaging is useful to interrogate tortuous CCA or ICA segments and to assist in correct assignment of the Doppler angle for measurements of velocity spectra. The CCA is examined as far proximally as possible to detect flow turbulence caused by CCA origin stenosis. Velocity spectra are recorded (≤60-degree Doppler angle) at multiple sites in the CCA, ICA (proximal, mid, distal), and ECA. Color or power Doppler imaging is used to identify regions of maximal stenosis for pulsed Doppler interrogation.

Interpretation

Accuracy.

In a validation study of the consensus criteria by AbuRahma et al,²² a PSV of 125 to 230 cm/s for detection of angiographic stenosis of 50% to 69% had a sensitivity of 93%, specificity of 68%, and overall accuracy of 85%. The consensus criteria for diagnosis of 50% to 69% stenosis can be significantly improved by use of an ICA PSV of 140 to 230 cm/s, with a sensitivity of 94%, specificity of 92%, and overall accuracy of 92%. A PSV of 230 cm/s or more for a stenosis of 70% or greater had a sensitivity of 99%, specificity of 86%, and overall accuracy of 95%. PSV was more accurate than EDV or ratio.²² These results are comparable to those of a meta-analysis of the literature, in which a PSV of 130 cm/s had a sensitivity and a specificity of 98% and 88% in detecting 50% stenosis. A PSV of 200 cm/s or more for detection of a greater than 70% stenosis was 90% sensitive and 94% specific.²⁰

When ultrasound is compared with computed tomography by the NASCET method of measurement for carotid stenosis, B-mode imaging with high resolution can provide impressive similarities to measurements with computed tomography imaging at some centers. A PSV of 370 cm/s to determine a greater than 70% stenosis was found to have a sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of 87%, 90%, 82%, 93%, and 93% compared with CTA.²³ An EDV of 140 cm/s had an overall accuracy of 90% in detecting a greater than 80% stenosis, and the same accuracy was detected with an ICA/CCA ratio of 6.

Limitations.

Significant tortuosity can result in higher incidence of sampling errors, and recent surgery can limit imaging. A contralateral severe stenosis or occlusion can falsely elevate ipsilateral velocities. Proximal or distal (tandem) lesions can prohibit accurate assessment of extracranial artery stenosis. Overestimation of the degree of stenosis with DUS may be explained by erroneously used angle corrections (i.e., angles >60 degrees). Nearly total occlusions substantially decrease velocities by diminished flow. Severe calcification may prevent accurate interrogation of the entire lesion, potentially missing the greatest point of stenosis.

Interpretation.

No flow should be detected by color or power Doppler at the diseased carotid bulb and beyond in the distal ICA. Resistance to flow in the ipsilateral CCA should be increased, with flow at or approaching zero in early diastole. Low-resistance flow may be seen in the ECA because of the development of collateral flow. A characteristic flow “thump” may be recorded in the proximal ICA as flow abuts the occlusion.

Accuracy.

The sensitivity, specificity, PPV, and NPV of DUS in detecting a carotid artery occlusion are 91%, 99%, 92% to 96%, and 98% compared with conventional arteriography.^28,29

Technique.

A 10- to 15-MHz specialized transducer provides high-resolution tissue imaging because the examination depth is less than 5 cm. The duplex probe is placed in a sterile plastic cover, and with saline used for acoustic coupling, the surgeon slowly scans the carotid repair in multiple sagittal and transverse planes and notes wall, lumen, and color Doppler flow features. The entire endarterectomy region is imaged, with special attention paid to repair site endpoints, transected plaque ends, and sites of vessel clamping. The ICA should be imaged as far distally as possible.

Interpretation.

The finding of wall irregularity or a small (<2 mm) flap in association with undisturbed flow (PSV < 125 cm/s) indicates a minor residual abnormality that is not found to be associated with postoperative stroke or thrombosis. When scanning identifies larger (≥3 mm) flaps, stenosis with a PSV higher than 150 cm/s and turbulent flow spectra, or lumen thrombus, immediate repair of the abnormality is recommended on the basis of clinical outcome data documenting a low (<1%) perioperative incidence of stroke and thrombosis and late (<3%) recurrent stenosis. If repair site imaging is normal and only elevated PSV is found, confirmation of an anatomic abnormality with arteriography is recommended before proceeding with operative revision. In most instances, lesions undergoing revision had both a B-mode and color Doppler abnormality and velocities higher than 200 cm/s.

Limitations.

Routine intraoperative imaging is not universally performed; in addition, no randomized studies evaluating observation of small technical defects and their natural history have been performed. Clamp time as well as status of stenosis of the contralateral carotid artery may influence velocities.

Technique.

Imaging techniques are similar to those used for carotid artery imaging.

Interpretation.

Innominate artery stenosis can often be demonstrated with noninvasive imaging by a combination of waveform analysis in both the right vertebral artery and the right CCA. The right CCA may show a squared-off waveform appearance, with systolic spikes and midsystolic deceleration during the cardiac cycles. Retrograde or bidirectional flow in the right vertebral artery may also be present. CCA disease demonstrates abnormal acceleration times and damped waveforms, and a critical proximal CCA stenosis or occlusion can cause retrograde flow from the ECA, providing antegrade flow to the ICA. A PSV lower than 30 cm/s in the ICA is suggestive of a proximal lesion. In addition, a PSV in the proximal CCA higher than 180 cm/s and an EDV higher than 30 cm/s also are relatively specific but poorly sensitive findings for significant stenosis.³⁶

Accuracy.

Few studies have looked specifically at the sensitivity and specificity of detecting proximal intrathoracic lesions. These lesions occur in 4% of patients undergoing carotid DUS, and accuracies above 99% and sensitivity and specificity of 96% and 100%, respectively, have been reported.³⁷ More recently, a study compared computed tomography with duplex velocity criteria in evaluating CCA stenosis greater than 50%. A PSV of 182 cm/s had a sensitivity and specificity of 64% and 88% in detecting a greater than 50% stenosis, whereas an EDV higher than 30 cm/s was inferior, with a sensitivity of 54% and a specificity of 74%.³⁸

Limitations.

Limitations of duplex detection of proximal CCA stenosis include congestive heart failure and cardiac valvular dysfunction as well as concomitant distal ICA stenosis. Because of the inability to directly insonate the vessels of interest, indirect criteria are sparsely reported in the literature, and no convincing universal criteria are currently recommended.

Indications.

TCD can be used to assess stroke patients for the adequacy of collateral circulation, intracranial stenosis, and arterial spasm. Less common indications include assessment of vasoreactivity; detection of emboli during or after carotid intervention; monitoring of carotid shunt patency; evaluation of the influence of arteriovenous malformations on cerebral flow; and monitoring of flow changes associated with cerebral trauma, hyperperfusion syndrome, or the diagnosis of “brain death.” An increased risk for stroke in young black children with sickle cell disease can also be determined by detecting MV values of the MCA higher than 120 cm/s produced by the anemia or the development of intracranial stenosis. The Stroke Prevention Trial in Sickle Cell Anemia (STOP) demonstrated a stroke reduction benefit of transfusion for anemia on the basis of TCD findings.⁴⁰

Technique.

For accurate comparison with established standards, vital signs during the examination should include systolic blood pressure higher than 90 mm Hg, heart rate above 60 beats/minute, and SpO₂ above 95%. A 1.5- to 5-MHz phased array transducer can image the intracranial arteries through the following windows: transtemporal, suboccipital, and transorbital. Blood flow direction with pulsed Doppler and velocity spectra, including MV, PSV, and RI, should be documented. Identification of arteries is based on color Doppler imaging, flow direction, and vessel depth (Table 16-3). Normal flow direction in the MCA (sample volume depth of 30 to 60 mm) is toward the transducer, whereas flow in the proximal ACA (depth of 60 to 85 mm) is away from it. Basilar artery and PCA flow is imaged through the foramen magnum scan window. Basilar artery (depth of 80 to 110 mm) flow is away from the transducer, with terminal PCA branches and P1 and P2 segments normally demonstrating bidirectional flow toward and away from the transducer. The orbital window is used to evaluate the ophthalmic artery (normal flow toward the transducer) and the ICA within the carotid siphon. Sample volume should be between 6 and 10 mm with maximum gain.

Interpretation.

Normal intracranial and basilar artery flow is characterized by a low-resistance spectral waveform, RI of less than 0.7, and PI of less than 4. Interpretation of duplex-recorded velocity spectra (0-degree Doppler angle) is based primarily on MV measurement, with normal values being in the range of 40 to 60 cm/s and similar in each hemisphere (see Table 16-3).⁴¹ Detection of emboli is based on recording of a distinctive low-velocity, high-amplitude transient produced as the embolic particle (air, platelet thrombus) passes through the pulsed Doppler sample volume (Fig. 16-14).^42–46

Elevations in mean flow velocity (MV) can indicate either a focal stenosis or a spasm, with the extent of artery involvement used to distinguish the two conditions. Arterial spasm is graded into three categories on the basis of the MV value: mild (80-120 cm/s), moderate (120-180 cm/s), and severe (>180 cm/s). The ratio of MV or PSV to velocity in the distal ICA should be higher than 6.^45–47 Hyperperfusion syndrome after carotid intervention typically demonstrates MV in the moderate elevation range, increased in comparison to the contralateral hemisphere, and a spectral waveform with a low (<0.6) RI, which indicates a loss of vasoreactivity.⁴⁷ To detect intracranial stenosis of greater than 50%, different criteria should be used: MCA mean flow velocity higher than 100 cm/s; systolic to prestenotic ratio of more than 2; and vertebral/basilar artery mean flow velocity higher than 90 cm/s.

Downstream from a high-grade (>70% to 80%) ICA stenosis or occlusion, the MCA and ACA velocity spectra may be normal or abnormal, depending on the functional patency of the circle of Willis. If the anterior or posterior communicating arteries are hypoplastic or absent, the TCD spectra will demonstrate a delayed acceleration time and decreases in MV and pulsatility compared with the contralateral hemisphere. If MV values are normal, the compensatory crossover flow from the contralateral ACA may be demonstrated by MV values higher than 150% of the ipsilateral MCA.

Accuracy.

If rigorous criteria are followed, the sensitivity and specificity of TCD for brain death confirmation, compared with cerebral angiography, are 88% and 100%, respectively.⁴⁸ Regarding the MCA stenosis criteria of higher 100 cm/s for a greater than 50% stenosis, sensitivity is 77% to 100%, specificity is 93% to 97%, PPV is 73% to 88%, and NPV is 94% to 100%; the overall accuracy is 90%. For a systolic to prestenotic ratio above 2 in detecting an intracranial stenosis greater than 50%, sensitivity is 80%, specificity is 94%, PPV is 76%, and NPV is 95%, with an overall accuracy of 91%. Finally, for a vertebral/basilar mean flow velocity higher than 90 cm/s in detecting an intracranial stenosis greater than 50%, the sensitivity is 69%, specificity is 99%, PPV is 95%, and NPV is 93%, with an overall accuracy of 94%.^49,50

Limitations.

Visualization of a poor temporal window is the most common limitation (5% to 30%). Proximal lesions in the extracranial artery can limit the value of intracranial velocities. False-positive findings are due to collateralization of flow in the presence of proximal ICA stenosis.

Technique.

High-frequency probes (5-11 MHz) are generally used in both longitudinal and transverse views, evaluating the temporal artery between the two layers of temporalis fascia.

Interpretation.

Stenosis is defined as a doubling of the PSV within the temporal artery with associated decreasing distal velocities and spectral broadening in the region of stenosis. Temporal artery occlusion is diagnosed when no color flow is detected within the temporal artery; a “halo” sign, which is a hypoechoic ringed region around the vessel secondary to edema, denotes temporal arteritis. In a prospective trial of 55 patients with suspected temporal arteritis, patients treated with corticosteroids demonstrated resolution of the hypoechoic edema by 4 weeks.⁵³

Accuracy.

A meta-analysis in 2010 reported a sensitivity of 68% and a specificity of 91% for DUS evaluation of the temporal artery. Bilateral halo signs were less sensitive at 43% but more specific at 100%.⁵⁴

Limitations.

Anatomy can limit examinations secondary to early branching of the frontal and parietal rami. Tortuosity and atherosclerosis may produce elevated velocities, and misdiagnosis of an adjacent vein as a halo has also been reported.

Indications.

DUS evaluation is beneficial for the following: patients with suspected symptomatic, chronic peripheral arterial disease^55–60; patients with suspected acute limb ischemia; patients with pedal infection without a palpable pulse; patients with atheroembolic or thromboembolic disease states; and patients who require surveillance after revascularization procedures.

Technique.

The extent of duplex arterial mapping should be individualized to the indication for peripheral arterial testing. Screening for aneurysm and imaging to identify tibial artery calcification are appropriate indications for testing in selected patients. Evaluation of patients with symptoms of claudication or other signs of peripheral arterial disease can localize sites of occlusive disease. Artery scanning proceeds from proximal to distal, with notation of artery diameter, wall morphology, plaque or thrombus accumulation, and velocity spectral waveform; it can include both upper and lower extremities.

Abdominal imaging is optimally performed after fasting to minimize intestinal gas artifact. Lower limb testing should include imaging (3- to 5-MHz phased, curvilinear, or linear array transducer) of the abdominal aorta and iliac arteries for aneurysm and occlusive disease and the common femoral arteries for the presence of normal (i.e., triphasic) velocity spectra. Multiple scan windows are used to image the aorta at the level of the renal arteries, proceeding to the tibial arteries at the ankle. From the inguinal ligament distally, a 5- to 7-MHz linear array transducer is used to map the arterial tree. PSV and EDV should be reported in the arteries evaluated, with ratios of PSV compared with a more proximal segment of designated healthy vessel.

The findings on duplex mapping should be recorded in a schematic of the extremity arterial tree, analogous to an arteriogram, with notation of the site or sites of aneurysm or occlusive disease and measurements of velocity spectra at nondiseased arterial segments (common femoral, superficial femoral, popliteal, and tibial arteries) and at sites of stenosis (Fig. 16-15). Standardized values for arterial diameter and PSV in the lower limb arteries have been reported for healthy subjects under resting conditions (Table 16-4).^55–59 Duplex mapping allows classification of atherosclerotic occlusive disease in the aortoiliac, femoral-popliteal, and popliteal-tibial arterial segments based on Trans-Atlantic Inter-Society Consensus (TASC) guidelines for grading of lesions from A through D according to lesion length and morphology.⁶⁰

Interpretation.

Duplex-derived data to estimate severity of stenosis of the lower extremity have not undergone a consensus recommendation process. Velocity-based criteria have been recommended by Cossman et al⁶¹ since 1989 (Table 16-5).

The PSV/EDV ratio should be calculated from an adjacent segment of a nonstenotic artery. Although the PI can aid in grading severity of stenosis, this number is not routinely independently used.

Figure 16-16 depicts the classic progression of deterioration of ankle-brachial index (ABI), decreasing PSV, and decreasing PI.

Classification of peripheral arterial stenosis is based on duplex-derived criteria, similar to carotid duplex testing (e.g., PSV, EDV, and PSV ratio across a stenosis identified by color Doppler imaging) (Fig. 16-17; see Table 16-5). A doubling or step-up in PSV to more than 150 cm/s (V_r > 2) indicates 50% or greater diameter reduction stenosis. Duplex criteria for a critical flow-limiting stenosis include loss of the triphasic waveform, spectral broadening with an increase in velocity (PSV > 200 cm/s, EDV > 0 cm/s), and V_r greater than 3 across the stenosis. Higher degrees of stenosis (>75%) are associated with an EDV higher than 100 cm/s and a V_r above 4.⁶²

Accuracy.

A prospective comparison of DUS to contrast-enhanced MRA was performed with 152 patients, with conventional digital subtraction angiography as the “gold standard.” A PSV ratio of more than 2.5 was used to define a significant stenosis. Duplex had sensitivity, specificity, and overall accuracy of 76%, 93%, and 89%, respectively, whereas MRA had a sensitivity of 84%, a specificity of 97%, and an overall accuracy of 94%, with statistically significant improvements in MRA sensitivity (P = .002) and specificity (P = .03).⁶³ Collins et al⁶⁴ performed a systematic review of the literature comparing the three different imaging modalities. In detecting a greater than 50% stenosis, contrast-enhanced MRA demonstrated a median sensitivity and specificity of 95% and 97%; CTA showed a sensitivity and specificity of 91% and 91%; and DUS showed a sensitivity and specificity of 88% and 96%. For detection of occlusions, contrast-enhanced MRA was more sensitive and specific than CTA or DUS.

Interpretation

Accuracy.

The accuracy of DUS for detection of iliac artery disease is as follows.

Limitations.

For the aortoiliac disease segment, body habitus is the main limitation. For infrainguinal stenosis, the major limitation of accuracy is the presence of a proximal stenosis, which can falsely decrease distal velocities. Consequently, the velocity ratio may more accurately identify focal areas of stenosis. DUS may signify a high-grade stenosis instead of occlusion if a collateral vessel is visualized therefore falsely demonstrating flow adjacent to an occluded vessel. Incomplete vessel imaging may occur as a result of poor patient cooperation or acoustic shadowing caused by plaque calcification. Arterial segments difficult to image include the proximal external iliac artery, internal iliac origin, and tibioperoneal trunk. Calcific changes in the tibial arteries can decrease the accuracy of velocities.

Technique.

Similar to native infrainguinal arterial examination, B-mode, color-flow, and pulsed Doppler imaging should be performed of the inflow, conduit, and outflow vessels. B-mode imaging can often identify early technical issues (i.e., retained valves or sclerotic vein segments). Pulsed Doppler with waveform analysis should be mapped of the lower limb with PSV, EDV, and V_r recorded. Specifics of the stenosis should also be recorded: diameter of the vein in transverse imaging above and below the stenosis, including length of stenosis in centimeters and anatomic location.

Interpretation.

A combination of ABI, PSV, and V_r has been used to stratify patients into risk of subsequent vein graft failure. A PSV higher than 180 cm/s with associated V_r of 2.0 and no change in ABI correlates well with a 50% stenosis. A critical vein graft stenosis with impending failure has a PSV higher than 300 cm/s, a V_r of more than 3.5, and a decrease in ABI by 0.15 or more. Refer to Table 16-6 for risk stratification of failure.

Vein graft lesions with the duplex-derived velocity spectra of a high-grade stenosis (PSV > 300 cm/s, EDV > 20 cm/s, V_r across the stenosis >3.5) correlate with a greater than 70% diameter reduction stenosis and should be repaired (see Fig. 16-6). In a prospective study, application of these threshold criteria identified all grafts at risk for thrombosis, and only one lesion with high-velocity criteria regressed. Multiple investigators have observed an approximately 25% incidence of graft thrombosis in stenotic bypasses when a policy of no intervention was followed.^70–73

The risk for graft thrombosis is predicted by using the combination of high- and low-velocity duplex criteria and decreases in ABI (see Table 16-6). In the highest risk group (category I), the development of a pressure-reducing stenosis produces low flow in the graft, which will result in thrombosis if it is decreased below the “thrombotic threshold velocity.” Prompt repair of category I lesions is recommended, whereas category II lesions can be scheduled for elective repair within 1 to 2 weeks. A category III stenosis (PSV of 180 to 300 cm/s, V_r < 3.5) does not reduce pressure or flow in the resting limb. Serial scans at 4- to 6-week intervals are recommended to determine hemodynamic progression of these lesions. An important feature of a “graft-threatening” stenosis is its propensity to progress in severity, to reduce graft flow, and to form surface thrombus—events that can precipitate thrombosis. By use of serial duplex scans, a category III stenosis that does not progress can be distinguished from a progressive lesion that needs to be repaired. The majority (approximately 80%) of bypass grafts will have no stenosis identified (i.e., category I scan). For these patients, surveillance at 6-month intervals is recommended. In patients with category I scans, a graft flow velocity lower than 40 cm/s indicates a “low-flow” bypass that is at increased risk for thrombosis by the concept of the thrombotic threshold velocity, which is lower in autologous vein than in prosthetic bypasses. Prescribing an anticoagulation regimen of warfarin to maintain the prothrombin time at an international normalized ratio of 1.6 to 2 and antiplatelet therapy (aspirin, 81 mg/day, or clopidogrel bisulfate, 75 mg/day) may reduce the incidence of low-flow vein bypasses.⁷⁰ Patients who are at high risk for development of vein graft stenosis⁷⁰ include those with abnormal results on baseline testing (PSV, 180-300 cm/s; V_r, 2-3.5), multisegment venous conduit, repeated bypass grafting, and need for warfarin therapy (Fig. 16-19).

Indications.

Indications for DUS include screening of patients at high risk for aneurysm and evaluation of patients with symptoms potentially attributable to aneurysms.¹⁹ Patients with pulsatile masses in the carotid or peripheral artery should be evaluated for the presence of aneurysms.

Technique.

A curved sector transducer with 1- to 5-MHz imaging head should be used to evaluate the aortoiliac segment in transverse and longitudinal planes with B-mode. The suprarenal, infrarenal, bifurcation, and iliac arteries should be imaged perpendicular to their centerline. Diameters in anteroposterior and transverse projections are recorded with assessment of presence of mural thrombus. Color flow is then used to determine the degree of mural thrombus present. Cross-section image should be captured during maximal pulsation with the cardiac cycle to determine calibrated measurement of maximum anterior-posterior and transverse diameter. Carotid and peripheral artery aneurysms can be assessed with higher frequency probes, as described for patients with occlusive disease. Because of mural thrombus, noninvasive imaging, not arteriography, is the modality of choice for screening and surveillance, and computed tomography is considered the gold standard for comparison.

Interpretation.

Maximum aneurysm diameter should be reviewed with confirmation of adventia-adventia measurements perpendicular to the centerline axis. Aneurysm reporting (abdominal aorta, iliac, femoral, popliteal) should include a description of its morphology (saccular versus fusiform) and extent and of the presence of mural thrombus or dissection. During evaluation of peripheral artery aneurysms, multiple sacs should be ascertained as these are frequently seen in patients with femoral pseudoaneurysms.

Accuracy.

Compared with computed tomography, the frequency of a 5-mm discrepancy in measurement varies from 8% to 33%.^75,76 Significant discrepancy between ultrasound and computed tomography has been reported in aneurysms between 5.0 and 5.4 cm, with 70 on computed tomography scan demonstrating a measurement of more than 5.5 cm.⁷⁶ Many authors have recommended computed tomography examination when the aneurysm size exceeds 5.0 cm. Data are limited on the accuracy of ultrasound-based imaging in the carotid and peripheral arteries.

Limitations.

Tortuous anatomy can potentially lead to inaccurate dimensional measurements. As in other regions of the abdomen, the presence of bowel gas and obesity can hamper image quality. Interobservational differences with ultrasound or computed tomography imaging are generally less than 2 mm but can be up to 0.5 cm. Inflammatory aneurysms may also have poorly defined tissue definition, and adventitia-to-adventitia measurements can be obscured.

Technique.

A curved sector transducer with 1- to 5-MHz imaging should be used with B-mode to assess the aneurysm sac size, similar to a native aneurysm examination. Duplex instrumentation should be set up for detection of low flow, and both color and power Doppler imaging are performed to identify attachment site (type I), patent side branch (type II), or stent-graft component separation (type III) endoleaks. Velocity spectra are recorded at sites of flow within the aneurysm sac outside of the stent graft. Flow in patent lumbar or inferior mesenteric artery (IMA) side branches may exhibit to-and-fro velocity spectra or a monophasic waveform indicating flow into the abdominal aortic aneurysm sac from one side branch and out a different side branch. Velocity in type II endoleaks is recorded because low-velocity (<80 cm/s) type II aneurysms are more likely to thrombose.^77–79

The use of intravenous ultrasound contrast agents may improve the sensitivity and specificity for detection of an endoleak. The ultrasound beam causes the microbubble agent to resonate at a certain frequency that increases all blood flow signals when harmonic duplex imaging is performed. After intravenous administration, the contrast agent circulates in the blood pool for 10 to 15 minutes, which allows imaging of the abdominal aortic aneurysm sac for flow outside the stent-graft or patent sac side branches.⁷⁸ Duplex imaging is not reliable for the detection of device migration.

Interpretation.

Maximum aneurysm sac size should be determined and compared with previous examination findings to determine changes in aneurysm size with time after endovascular exclusion. Assessment of the aneurysm sac for flow and categorization of endoleak type should be performed. The risk of limb-related complications has been shown to be potentially increased when PSV within the graft limbs exceeds 300 cm/s.

Accuracy.

Duplex scanning for detection of endoleaks is on the order of 80% to 85%, and it may be more sensitive than contrast-enhanced computed tomography unless delayed imaging is included in the scan protocol.⁸⁰ Wolf and associates reported a sensitivity of 81% and a specificity of 95% (PPV, 94%; NPV, 90%) for DUS in comparison to the gold standard computed tomography for detection of endoleaks.⁸⁰ A meta-analysis of DUS and contrast-enhanced ultrasound versus computed tomography to detect endoleaks after endovascular aneurysm repair found a sensitivity of 77% and a specificity of 94% for DUS. Improved results were seen in a smaller sample of 288 patients with a 98% sensitivity and an 88% specificity with the contrast-enhanced duplex scanning.⁸¹

Technique.

A 3- to 9-MHz linear array transducer is used in most patients; however, in morbidly obese patients, a curved sector transducer with 1 to 5 MHz may be necessary. B-mode imaging should evaluate pseudoaneurysm size in anteroposterior and transverse dimensions. Maximum diameter and number of aneurysm sacs as well as the dimensions of residual flow within the aneurysm sac should be recorded. The neck diameter, length, and vessel of origin should be assessed. Imaging should also include color flow in longitudinal and transverse imaging planes.

Interpretation.

Hematomas lack flow and cannot be differentiated from a thrombosed pseudoaneurysm. Duplex characteristics of pseudoaneurysms include flow outside the artery, the presence of a neck from the puncture site to the aneurysm sac, and a classic to-and-fro flow pattern in the neck corresponding to blood flow into the aneurysm during systole and sac emptying during diastole (see Fig. 16-19). The proximal native arterial segment often has a low-resistance pattern compared with the more distal segment. This is in contrast to an arteriovenous fistula, which will have arterialization of the draining vein and a low-resistance pattern in the artery proximal to the fistula. On color-flow imaging, there is often perivascular speckling due to vibrating soft tissues from high flow rates.

Accuracy.

Sensitivity and specificity for detection of pseudoaneurysm are high at 95% and 95%.⁸⁵ A prospective study of 53 patients with suspected groin-related complications compared physical examination with DUS evaluation. All patients underwent operative exploration. The results of these evaluations were compared with the findings at operation. Physical examination was 100% sensitive and 100% specific in the diagnosis of arteriovenous fistula, whereas duplex scan achieved a sensitivity of only 58% and a specificity of 100%. Physical examination was 92% sensitive and 93% specific in the detection of femoral pseudoaneurysm compared with a sensitivity of 83% and a specificity of 100% for duplex scanning.⁸⁶

Limitations.

Large overlying hematomas can obscure imaging, as can large body habitus. Uncommonly, a branch with adjacent hematoma may be misinterpreted as a pseudo­aneurysm. Lymph nodes can also be vascularized and provide a false-positive examination finding; they can usually be differentiated by a radial flow pattern. Branch vessels can be confused with needle tracks or a pseudoaneurysm neck.

Duplex-Guided Thrombin Injection.

This ultrasound-guided examination can be both diagnostic and therapeutic for iatrogenic pseudoaneurysms. With gray-scale imaging, a spinal needle is positioned in the periphery of the aneurysm sac with the tip of the needle verified. Color Doppler should be used to confirm the needle tip is in the aneurysm sac within the flow region before changing to B-mode and injection performed. Injections are done with small aliquots of 0.1 mL of thrombin, and the aneurysm is reimaged in color flow to assess for successful closure. After the injection is completed, Doppler examination of the extremity is performed to assess for successful aneurysm thrombosis and for distal embolization.

Indications.

Indications for renal artery DUS evaluation include early-onset hypertension (<35 years of age), acute renal failure with aortic dissection, unknown etiology of azotemia, and unexplained size discrepancy between kidneys (>1.5 cm difference). In addition, patients with malignant or resistive hypertension or worsening chronic hypertension should be considered for renally mediated hypertension.¹⁹

Technique.

Duplex scanning of the renal arteries is best performed in the morning after an overnight fast to minimize intestinal gas and to provide optimal conditions for imaging of the visceral aortic segment, its branches, and adjacent veins. The examination begins with the patient in the supine position with the head slightly elevated 20 degrees, which allows the viscera to descend in the abdomen. Through a midline scanning window, 3- to 5-MHz abdominal transducers with focal regions adjusted to the appropriate depth are used to identify the aorta, superior mesenteric artery (SMA), and proximal renal arteries. Allowing patients to raise their arms over the head elevates the ribs and improves transducer access for imaging of the renal arteries. The aorta is scanned from the level of the diaphragm to the origins of the iliac artery. Identification of the celiac artery and SMA arising anteriorly helps in locating the renal artery origins. Transverse color Doppler imaging of the aorta at the level of the SMA will locate the anteriorly positioned left renal vein and the proximal right and left renal arteries arising below the SMA takeoff and slightly posterior to the aorta. The course of the right renal artery is posterior to the inferior vena cava. Use of color or power Doppler helps trace the renal artery course for velocity spectral recording along its length to the extent possible. The examiner should be aware of anatomic variations (duplicate or multiple renal arteries, pelvic and horseshoe kidneys, retroaortic renal vein, aneurysmal dilatation of the aorta or renal or visceral arteries) and the color and power Doppler features of fibromuscular hyperplasia. Care should be taken to record velocity spectra at angles of insonation of less than 60 degrees. When velocity spectra are recorded from the renal parenchyma (interlobar and arcuate arteries, three recording sites), a 0-degree Doppler angle of insonation should be assigned. Kidney length, which correlates with renal mass and function, is measured pole to pole during maximum inspiration by use of a flank approach to optimize visualization of the renal cortex margins. The PSV and EDV should be sampled from the origin and the proximal, mid, and distal renal arteries. The RI is calculated by examining the interlobar or segmental arteries according to the following formula:

Interpretation.

Renal artery stenosis has historically been and continues to be categorized as less than or greater than 60% stenosis. Interpretation is based on the maximum PSV within the most stenotic region of the main renal artery, which typically occurs at the ostium in atherosclerotic disease. Normal renal PSV ranges from 70 to 127 cm/s. Diagnostic criteria for greater than 60% stenosis include a color-flow jet with post-stenotic turbulence and a PSV higher than 180 to 200 cm/s (Table 16-7).⁹⁰ However the largest series to assess the optimal criteria suggests a higher threshold of PSV of 285 cm/s.⁹¹

A renal-aortic velocity ratio (RAR) above 3.5 has also been used to determine a greater than 60% renal artery stenosis. The recent publication recommending a higher PSV by AbuRahma et al⁹¹ recommended an RAR of 3.7 for detection of significant renal artery stenosis. Other duplex parameters that suggest a significant renal artery stenosis include damping of the intrarenal artery signal, acceleration time of more than 100 milliseconds, and pulsus tardus waveform (prolonged systolic acceleration evident in delayed systolic upstroke and rounded systolic peak). The finding of a pulsus tardus waveform is a useful secondary criterion for critical renal artery stenosis.⁹² The RAR cannot be used for interpretation when the suprarenal aorta is aneurysmal or the PSV is lower than 45 cm/s. In this instance, a PSV higher than 200 cm/s is the criterion for greater than 60% diameter reduction stenosis and has a diagnostic accuracy similar to that with the use of RAR alone. Interpretation of renal artery occlusion requires the absence of renal artery flow, a small (<9 cm) kidney length, and a low (PSV < 20 cm/s) hilar artery flow. The renal RI appears to be the best predictor of transplant failure, with some series using higher than 0.7 and some higher than 0.8 as abnormal levels.

Accuracy.

Renal duplex for grading of atherosclerotic renal artery stenosis (<60%, >60%, occluded) has a sensitivity of 88% to 85% and a specificity of 92%. Overall agreement with angiography is in the range of 80% to 90%.^92–95 Accuracy is influenced by the presence of diseased accessory renal arteries but not by concomitant aortoiliac disease or renal insufficiency.

In a recent series comparing the best parameters and classical criteria, AbuRahma et al⁹¹ reported a PSV higher than 285 cm/s for detection of a greater than 60% stenosis; the sensitivity, specificity, and overall accuracy were 67%, 91%, and 81%, respectively. In the same series, a PSV higher than 180 cm/s and an RAR of 3.5 had a sensitivity, specificity, and overall accuracy of 72%, 81%, and 78%. Changing to a threshold criterion of a PSV higher than 200 cm/s changed only the specificity to 83%.

The only prospective comparison of DUS, captopril renography, MRA, and CTA in assessing renal artery stenosis found that MRA and CTA were more sensitive and specific than ultrasound or captopril renography (P < .001), with a reported sensitivity and specificity for DUS of 73% and 71% versus 94% and 62% for CTA and 93% and 91% for MRA.⁹⁵

Limitations.

As with other abdominal vascular examinations, body habitus and bowel gas can compromise imaging quality and therefore examination accuracy. Multiple renal arteries can exist, and missing the main renal artery can provide misleading results. Overall, the percentage of inadequate examinations is approximately 10%, much higher than other imaged territories.

Interpretation.

Currently, there is consideration for a higher PSV to determine significant in-stent stenosis. A more accurate PSV for determining a significant in-stent restenosis appears to be 250 cm/s.⁹⁶ After surgical bypass, the anastomosis should be assessed for velocities exceeding 200 cm/s and for damped waveforms in the hilar arteries.⁹⁷

Accuracy.

The sensitivity, specificity, and accuracy of using a PSV of 180 cm/s to determine a greater than 60% stenosis are 73%, 80%, and 77%, whereas the sensitivity, specificity, and accuracy of using a PSV of 250 cm/s are 59%, 95%, and 83%.⁹⁶

Technique.

Previous operative notes should be reviewed to document the origin of the transplanted renal artery. A thorough examination should include the aorta and iliac arteries. The renal vein should also be assessed for patency and any evidence of thrombus. Extrarenal evaluation should include PSV and EDV measurements in the iliac and renal (proximal, mid, and distal) arteries. The classic parvus-tardus pattern of intrarenal signal is a loss of the early systolic peak with a prolongation of the acceleration time. In addition, the calculation of the renal RI should be recorded.

Interpretation.

A PSV exceeding 250 cm/s and a PSV ratio compared with distal external iliac artery of more than 1.8 are suggestive of significant disease. An RI exceeding 0.7 has corresponded to transplant failure. A prolongation of the acceleration time of more than 0.10 second has also been associated with proximal renal artery stenosis.

Accuracy.

A PSV higher than 250 cm/s for detection of a greater than 50% stenosis has a sensitivity of 100% and a specificity of 95%.⁹⁸

Indications.

Patients with abdominal pain with suspected acute or chronic mesenteric ischemia or ischemic colitis should be assessed for a hemodynamically significant visceral lesion. Patients are also often evaluated preoperatively before liver transplantation.

Technique.

A low-frequency 2- to 5-MHz curvilinear phased array transducer with a Doppler angle of 60 degrees or less to maximize consistent measurements is used. The supine position in a fasting patient is typically recommended, but an oblique window may have to be used to profile the visceral vessels. Transverse and sagittal planes are used while the probe is placed just inferior to the xiphoid process. After visualization of the celiac artery, attention is focused on the SMA. The origins of both vessels are best seen in the sagittal view, with the major celiac branches (hepatic, splenic, and left gastric arteries) viewed from a transverse imaging plane where the classic “seagull sign” is best demonstrated. Assessment should be made in both gray scale and color Doppler followed by Doppler analysis with a sample volume of 1.5 mm in the orifice of the proximal region of both celiac artery and SMA. The PSV and EDV as well as post-stenotic turbulence should be recorded.

Interpretation.

Velocity criteria developed by the University of Washington vascular laboratory group are shown in Box 16-1. Grading of stenosis of the visceral arteries should include assessment of the PSV, the EDV, and the mesenteric-aortic ratio. The most accurate of these ultrasound-detected abnormalities is the PSV. Additional findings include direction of flow within the visceral artery (i.e., a reversal of flow in the hepatic artery is highly suggestive of critical stenosis or occlusion of the celiac artery).^102,104,106 Occlusion or high-grade stenosis of the celiac artery can result in falsely elevated PSV in the SMA and alter the fasting signal to a low-resistance waveform and vice versa.

If the right hepatic artery arises from the SMA, the fasting SMA spectral waveform is monophasic with an RI similar to that of the celiac artery, but the PSV should be less than 150 cm/s in the absence of stenosis.^99,104

Accuracy.

A study by AbuRahma et al^102,103 reported the following. For criteria defining a lesion greater than 70% stenosis of the celiac artery, a PSV higher than 320 cm/s had a sensitivity of 80%, a specificity of 89%, and an overall accuracy of 85%. An EDV higher than 100 cm/s had a sensitivity of 58%, a specificity of 91%, and an overall accuracy of 77%. For criteria defining a greater than 70% lesion in the SMA, a PSV higher than 400 cm/s had a sensitivity of 72%, a specificity of 93%, and an overall accuracy of 85%. An EDV higher than 70 cm/s had a sensitivity of 65%, a specificity of 95%, and an overall accuracy of 84%.

Limitations.

Anatomic variants of the mesenteric circulation, present in up to 20% of patients, can contribute to interpretation errors. Body habitus and overlying bowel gas can obstruct view of the vessels and underlying structures, limiting the evaluation.

Interpretation.

A normal visceral stent examination finding is associated with a widely patent color-flow lumen and nondisturbed velocity spectra, whereas an elevation of the PSV to more than 400 cm/s or an EDV higher than 105 cm/s has correlated with a greater than 70% stenosis for both the celiac artery and the SMA. When borderline velocity criteria for stenosis are recorded, testing after a meal should be performed.

Accuracy.

For the celiac artery, a PSV higher than 363 cm/s used for determination of a greater than 70% in-stent stenosis has a sensitivity of 88%, a specificity of 92%, and an overall accuracy of 90%. For the SMA, a PSV higher than 412 cm/s used for determination of a greater than 70% in-stent stenosis has a sensitivity of 100%, a specificity of 95%, and an overall accuracy of 97%. The EDV for both celiac artery and SMA in-stent restenosis was statistically inferior to the PSV.¹⁰³

Technique.

The examination uses an ultrasound duplex imager with pulsed wave and color-flow Doppler capability and Doppler spectral analysis. Curved and phased arrays are used to examine deeper structures, such as arteries and central veins, or during examination of obese patients. Linear arrays are used for more superficial imaging. Transducer frequencies include the C5-2/2.5 MHz, L7-4/4.0 MHz, L12-5/6.0 MHz, P4-2/2.0 MHz, and P4-1/2.0 MHz. High frequencies give better sensitivity to low flow, whereas low frequencies have better penetration.¹⁰⁹

The patient is placed supine for brachial artery pressures, with the head turned away from the upper limb being examined. The central veins are assessed with low-frequency transducers, examining the distal innominate veins and the subclavian veins. The infraclavicular segments of the subclavian veins and the axillary veins are easier to image. Doppler findings suggestive of abnormality or stenosis include focal increases in flow velocity with post-stenotic turbulence proximally and a continuous flow pattern distally.¹¹⁰ Superficial veins are then examined. B-mode imaging in transverse orientation allows diameter measurement and visualization of tributary locations. The vein is compressed to ensure no evidence of stenosis, thrombosis, or occlusion. The longitudinal view and addition of color flow can provide additional anatomic and physiologic information as needed. The entire cephalic vein is scanned from the wrist to the confluence with the subclavian vein. The basilic vein is scanned from the wrist to the confluence with the brachial vein. Many laboratories use a tourniquet to aid with vein distention; if a tourniquet is used, a vein diameter of 2.5 mm is generally acceptable, whereas a diameter of more than 3 mm is preferred when no tourniquet is used.¹¹⁰ Arterial mapping is then performed. The subclavian, axillary, brachial, radial, and ulnar arteries are examined for atherosclerosis, tortuosity, or unusual anatomy; normal findings are triphasic waveforms with no significant focal velocity increases. Radial and ulnar artery diameters are measured at the wrist, and brachial diameters are also sometimes recorded. Brachial artery pressures are obtained from both arms for comparison, and discrepancies between sides can indicate inflow stenosis from the proximal arteries.

Interpretation.

The artery and vein diameters are analyzed. Arteries with diameters smaller than 1.5 to 2.0 mm have been associated with increased nonmaturation rates in arteriovenous fistulae.¹¹¹ Venous diameter should be 2.5 mm or more for creation of an arteriovenous fistula and 4.0 mm or more for placement of a graft.¹¹¹ Absence of segmental stenosis or occluded segments, continuity with the deep venous system, and contiguous length of nondiseased vein of more than 10 cm have also been predictive of improved outcomes with access creation.¹¹² The presence of large tributaries or multiple branches should also be noted as these could affect maturation rates or ultimate function of the access. Vein wall thickening may indicate prior phlebitis, which along with current thrombus should be noted.

Accuracy.

In detecting relevant stenoses, DUS has a sensitivity and specificity of 90.9% and 100% for the subclavian artery, 93.3% and 100% for upper arm arteries, 88.6% and 98.7% for forearm arteries, and 70% and 100% for arteries of the hand.¹¹³ DUS has sensitivity, specificity, PPV, and NPV of 81%, 90%, 90%, and 78% for detection of venous stenosis, thrombosis, and occlusion.¹¹⁴ Sensitivities decrease for more proximal veins; sensitivity for detection of abnormalities is 79% for the subclavian vein, 75% for the innominate vein, and 33% for the superior vena cava.¹¹⁴

Limitations.

Patients with previous access attempts or revisions can present with distorted anatomy, areas of occlusion or thrombosis, and varying amounts of subcutaneous conduit, all of which can impede or confuse typical preoperative mapping procedures. Morbidly obese patients can present with vessels that are difficult to visualize completely. In addition, aberrant anatomy, multiple branches, or significant tortuosity can make single-vessel identification and complete tracing challenging.

Technique.

The instrumentation is similar to that described in the preoperative assessment. The examination should begin with evaluation of the inflow artery in transverse view to delineate the anatomy. The artery is then examined in longitudinal view; the area proximal to the access anastomosis is examined, measuring the PSV and EDV. Duplex images are then obtained at predetermined locations along the fistula or graft: inflow artery proximal to the fistula or graft; inflow artery distal to the fistula or graft; anastomotic sites; puncture sites; proximal, mid, and distal outflow vein or graft; and axillary and subclavian veins.¹⁰⁹ Peak systolic velocities and waveforms should be measured in any area where velocity increase or turbulence is noted. Finally, B-mode imaging should be used to evaluate the depth of the access from the skin surface and conduit diameters along various points of the access. Diameter measurement is an important variable in the calculation of flow rates along various portions of the access, as seen by the following equation:

Interpretation.

DUS imaging performed after construction of dialysis access can accurately characterize the hemodynamics by measurement of volume flow and identification of stenosis.¹¹⁶ Testing enables recognition of a low-flow nonmaturing conduit, identifies lesions suitable for remedial intervention, documents adequate volume flow for effective dialysis before the first cannulation, and predicts access patency. The mean volume flow rate of forearm autogenous vein fistulae (800 ± 200 mL/min) is less than that of arm autogenous vein fistulae (1400 ± 400 mL/min) or prosthetic bridge grafts (1200 ± 200 mL/min). For a normally functioning dialysis access, the midconduit PSV is in the range of 200 cm/s and the arterial anastomotic PSV is in the range of 400 cm/s. A focal increase in a threshold PSV higher than 400 cm/s and a PSV ratio above 3 are the criteria for a greater than 50% diameter reduction stenosis. Robbin et al¹¹⁷ demonstrated that flow rates of 500 mL/min or higher are associated with maturity in nearly 90% of fistulae. Duplex-measured volume flow rates of less than 500 mL/min predict poor access function during hemodialysis and failure. A measured volume rate of less than 500 to 800 mL/min can be used as a threshold to recommend contrast-enhanced fistulography to define dialysis access anatomy and to image the central veins for stenosis or occlusion.

Accuracy.

Several studies have investigated the accuracy of duplex scanning to identify hemodynamically significant stenosis that develops in dialysis accesses in hopes of allowing potential interventions to diminish the chance of access failure. Grogan et al¹¹⁸ defined a hemodynamically significant stenosis of 50% or greater in either the arterial inflow or venous outflow using a systolic V_r of more than 2 as well as a systolic V_r of more than 3 and a minimum PSV of 400 cm/s for an anastomotic stenosis, with a sensitivity of 93% and a specificity of 94% confirmed by fistulography. With regard to flow rates and prediction of potential failing accesses, Back et al¹¹⁹ determined that a threshold conduit flow rate of 800 mL/min better discriminated failing and functional fistulae and bridge grafts (accuracy 77%) than a flow rate greater or less than 500 mL/min (accuracy 67%).

Limitations.

DUS measurements in dialysis accesses may be affected by pathologic change that alters flow measurements, such as the presence of a well-collateralized occlusion, low systemic pressure, or central venous stenosis or occlusion, as well as by technician error, such as poor Doppler angle.

Technique.

Ultrasound examination of the penis is conducted with the patient supine or in the lithotomy position. High-frequency (7.5-14 MHz) linear array transducers are typically used. The penis is examined in transverse and longitudinal planes starting at the level of the glans and moving down to the base of the penis.¹²¹ Peak systolic velocities of the cavernosal arteries are then recorded, measured at the junction of the proximal third and distal two thirds of the penile shaft. When pharmacologic testing is necessary, prostaglandin E₁ (10 to 20 µg) is injected into one of the corpora cavernosa laterally in the distal part of the penis.¹²¹ Other vasoactive agents used are papaverine (30 to 60 mg) and trimix (papaverine, phentolamine, and prostaglandin E₁). Tumescence and rigidity as well as velocities are recorded in both cavernosal arteries at 5, 10, 15, 20, 25, and 30 minutes after injection. Velocities in the deep dorsal vein are also recorded.¹²¹

Interpretation.

Arterial insufficiency is demonstrated in the flaccid penis when the PSV is less than 10 cm/s in the cavernosal artery.¹²² The average PSV after injection of vasoactive agents in the normal patient has been found to be 30 to 40 cm/s; arterial insufficiency is diagnosed with PSVs lower than 25 cm/s.¹²³ PSVs between 25 and 35 cm/s are equivocal, and secondary criteria for erectile dysfunction in this range include asymmetry of more than 10 cm/s in PSV, increase in the diameter of the cavernosal artery by 75% after intracavernosal injection, focal stenosis in the cavernosal artery, and retrograde flow.¹²³ Acceleration time has also been used to detect cavernous atherosclerosis, with a cutoff point of less than 100 milliseconds indicating the presence of atherosclerotic disease.¹²⁴ Adequate arterial inflow, short-duration erection, and persistent antegrade flow of 5 cm/s throughout all phases of erection suggest venous leak.¹²⁵

Accuracy.

A PSV lower than 10 cm/s in the cavernosal artery of a flaccid penis predicts arterial insufficiency with an accuracy of 93%.¹²² A velocity threshold of 25 cm/s with cavernosal injection of vasoactive agents has a 92% accuracy in diagnosis of arterial insufficiency.¹²³ An acceleration time cutoff point of 100 milliseconds has a sensitivity of 66% and a specificity of 71%.¹²⁴

Limitations.

Anxiety in patients undergoing penile duplex examination as well as reproducibility of the examination can affect outcomes. Patient response to vasoactive agents or intolerance to agents due to various comorbidities can also limit the scope of duplex examination. PSVs after vasoactive injection that fall between 25 and 35 cm/s are equivocal and rely on evaluation of additional criteria.