How to Test for Carotid Artery Blockage

Carotid artery blockage is most commonly tested with a painless duplex ultrasound of the neck, which combines standard ultrasound imaging with Doppler flow measurement to estimate how narrowed the arteries have become. When more detail is needed, CT angiography or MR angiography can map the vessels with high precision. Which test you get depends on whether you have symptoms, what your doctor suspects, and how urgently a treatment decision needs to be made. The testing landscape is broader than most people realize, and the choice of method genuinely affects what gets found.

Why Carotid Blockage Matters Enough to Test For

The carotid arteries run up each side of your neck and supply most of the blood to your brain. When fatty plaque builds up in those arteries and narrows them, the condition is called carotid stenosis. That narrowing accounts for roughly 10 to 20 percent of all ischemic strokes, which are the kind caused by a blocked blood supply to brain tissue.¹ What makes carotid disease tricky is that the most common mechanism is not a slow choking off of blood flow. Instead, a piece of unstable plaque breaks loose and travels upstream into the brain as an embolus. That can happen even when the narrowing itself is moderate.² This dual nature of the threat, both the degree of narrowing and the stability of the plaque, shapes how doctors approach testing.

The Physical Exam Starting Point

Before any imaging, a doctor might listen to your neck with a stethoscope. A whooshing sound called a carotid bruit can indicate turbulent blood flow through a narrowed artery. It is a quick, free, zero-risk check, but it has real limitations. In the Northern Manhattan Study, listening for a bruit picked up only about 56 percent of people who actually had significant stenosis.³ A larger observational study found sensitivity around 71 percent for blockages of 50 percent or more, with the best detection at the most severe narrowings.⁴

Counterintuitively, a completely blocked carotid artery is the hardest to catch by ear: one study found a bruit in only about a quarter of total occlusions, because there is so little flow left to make noise.⁴ A pooled analysis confirmed that the likelihood of hearing a bruit does not actually increase as the stenosis gets worse.⁵ So a bruit is a useful clue that should prompt further testing, but the absence of one does not rule anything out.

Symptoms That Trigger Testing

Many people with carotid stenosis have no symptoms at all until a stroke occurs. When symptoms do appear before a full stroke, they often take the form of a transient ischemic attack, or TIA, sometimes called a “mini-stroke.” You might experience sudden weakness or numbness on one side, trouble speaking, confusion, or dizziness that resolves within minutes to hours. A less well-known warning sign is amaurosis fugax, a temporary loss of vision in one eye, sometimes described as a curtain falling over the visual field. In one review of patients undergoing carotid evaluation, about one in five had presented with this eye symptom.⁶ If you experience any of these, testing should be fast-tracked.

Duplex Ultrasound, the First-Line Test

Carotid duplex ultrasound is the most widely used first test. A technologist places a probe on your neck and images the artery in real time while measuring how fast blood flows through it. The key measurement is peak systolic velocity, the top speed of blood during each heartbeat. When an artery narrows, blood accelerates through the pinch point, so higher velocities mean more blockage. No needles, no radiation, no contrast dye, and the scan typically takes 20 to 30 minutes.

However, what counts as a “significant” velocity reading is less standardized than you might expect. A nationwide study found 60 distinct velocity thresholds in use across accredited vascular labs in the United States. The cutoff for moderate stenosis (50 percent or more) ranged from 110 to 245 cm/s across centers, and for severe stenosis (70 percent or more) it ranged from 175 to 340 cm/s.⁷ That variation matters in practice: the most liberal threshold for moderate stenosis would diagnose twice as many people as the most conservative one. About one in ten surgical candidates had velocity values that would be called severe at some labs but not at others.⁷ If your ultrasound result falls near a decision-making threshold, asking which cutoff values your lab uses is a reasonable question.

CT Angiography

When ultrasound results are borderline, inconclusive, or suggest significant disease, the next step is often CT angiography. You receive an intravenous injection of iodine-based contrast dye while a CT scanner takes rapid cross-sectional images of your neck and head. The scan itself takes seconds. Software reconstructs a three-dimensional view of your carotid arteries, showing the lumen, the vessel wall, and any calcified plaque.

CTA is highly accurate for severe stenosis. A 2024 meta-analysis found that across multiple studies, CTA had a combined sensitivity of 93 percent and specificity of 99 percent for detecting severe narrowing of the internal carotid artery.⁸ A head-to-head comparison of CTA, MRA, and Doppler ultrasound against catheter angiography rated CTA as the most accurate overall, outperforming both MRA and ultrasound.⁹ Even heavily calcified plaques, which can make ultrasound readings unreliable, are assessable on CTA source images.¹⁰

The downsides are radiation exposure and contrast dye. A standard carotid CTA delivers a radiation dose in the range of about 1 to 4 millisieverts, depending on the scanner settings. One trial showed that using a lower tube voltage (80 kV instead of the standard) cut the effective dose from roughly 4 mSv to about 1.2 mSv, though image quality dipped slightly in certain neck areas.¹¹ The iodine contrast can pose problems for people with poor kidney function or contrast allergies, so your kidney labs are usually checked beforehand.

MR Angiography

MR angiography uses a strong magnetic field instead of X-rays, so there is no radiation. It comes in two main flavors: time-of-flight (TOF) MRA, which relies on the motion of flowing blood to generate an image without any injected dye, and contrast-enhanced (CE) MRA, which uses a gadolinium-based contrast agent for sharper detail.

A large systematic review and meta-analysis found that for detecting severe stenosis (70 to 99 percent), TOF MRA had a sensitivity of about 91 percent and specificity of 88 percent, while CE MRA performed better at roughly 95 percent sensitivity and 92 percent specificity.¹² For detecting complete occlusion, both techniques were excellent, with CE MRA reaching nearly 100 percent sensitivity.¹²

A known weakness of MRA, particularly the non-contrast TOF version, is that it tends to overestimate how severe a blockage is. Turbulent or very slow flow can cause signal loss that mimics a tighter narrowing or even a total occlusion. One comparison study found TOF MRA misinterpreted severe stenosis as complete occlusion in two cases where blood was still getting through.¹³ A separate study confirmed that TOF MRA produced significantly higher stenosis grades than contrast-enhanced MRA at the same location.¹⁴ For moderate narrowings (50 to 69 percent), TOF MRA’s sensitivity drops to under 40 percent, meaning it misses most of them.¹²

MRA is a good option for people who cannot tolerate iodine contrast or who should avoid radiation, such as younger patients needing repeated monitoring. But if the clinical question is whether you need surgery, and the MRA suggests borderline or severe disease, a follow-up CTA or catheter angiography may still be needed to confirm the grade.

Catheter Angiography, the Traditional Gold Standard

Digital subtraction angiography, performed by threading a catheter from the groin or wrist up into the carotid arteries and injecting contrast while taking X-ray images, was the only method available for evaluating carotid stenosis for roughly half a century. It still offers the most direct, real-time view of blood flow and remains the reference against which all non-invasive tests are judged. But it is an invasive procedure that carries a small risk of stroke, arterial injury, and other complications. A study specifically examining the safety of arch aortography (a less targeted catheter approach) found that it had a lower neurological complication rate than selective catheter placement directly into the carotid.¹⁵

Because CTA and MRA have become so accurate, catheter angiography is now reserved for cases where non-invasive imaging gives conflicting results, when intervention is planned and the surgeon wants a road map in the same sitting, or when plaque characteristics need to be assessed in ways that a CT or MRI cannot provide. Most people being evaluated for carotid disease will never need a catheter study.

How Stenosis Is Measured, and Why It Varies

A number like “70 percent stenosis” sounds precise, but the percentage depends entirely on which measurement method is used. Two landmark clinical trials, one North American and one European, used different reference points to calculate stenosis, and the methods can disagree sharply. A comparison of three approaches on over a thousand angiograms found that the methods disagreed on whether a blockage should be classified as mild, moderate, or severe in 51 percent of measurements. The European and common-carotid methods categorized twice as many narrowings as severe compared with the North American method.¹⁶

This is not just an academic curiosity. Treatment guidelines hinge on specific percentage thresholds, so the method used to calculate stenosis can shift a patient from “watch and wait” to “surgical candidate.” A recent study demonstrated this concretely: when CTA stenosis was calculated using an area-based method instead of the standard diameter-based method, 76 percent of lesions that would have been called below 50 percent were reclassified into the 50-to-69 percent range, and the pool of patients meeting the 70-percent surgical threshold ballooned more than fourfold.¹⁷ If you are told you have a specific percentage of stenosis, it is worth asking which measurement standard your report used.

Should You Be Screened If You Have No Symptoms?

Given how dangerous a carotid stroke can be, you might think routine screening would make sense for people with risk factors like high blood pressure, smoking, or diabetes. The U.S. Preventive Services Task Force looked at this question and came down firmly against it. Their recommendation, reaffirmed in 2021, is that screening the general adult population for asymptomatic carotid stenosis does more harm than good.¹⁸

The reasoning is not that testing itself is dangerous, but that the cascade of follow-up can be. False positives lead to unnecessary invasive imaging or surgery. Even true positives in people with no symptoms often do better with medication (statins and blood-pressure control) than with surgery, and the surgical procedures themselves carry a small but real stroke risk. Modern medical therapy has improved so much that the net benefit of finding and fixing an asymptomatic blockage has shrunk considerably. Testing makes sense when you have had symptoms, specifically a TIA, a stroke, or the visual episodes mentioned earlier. It also makes sense when another test, such as a cardiac procedure, incidentally detects a bruit or other sign of possible carotid disease.

Looking Beyond Narrowing at Plaque Vulnerability

Measuring how tight a stenosis is tells only part of the story. Two people with 60-percent blockage can have very different risks depending on the composition of their plaque. A soft, lipid-rich plaque with a thin cap and hemorrhage inside it is far more likely to rupture and cause a stroke than a dense, calcified, stable plaque of the same size. This concept, called plaque vulnerability, has pushed imaging beyond just measuring the lumen.

High-resolution MRI can now visualize what is inside the plaque wall itself. In one study comparing MRI findings with tissue from surgery, MRI detected bleeding within the plaque with 90 percent sensitivity, though specificity was more moderate at 74 percent.¹⁹ Subsequent work confirmed that MRI can accurately identify intraplaque hemorrhage and track how plaques grow over time, with hemorrhage appearing to accelerate plaque progression.²⁰ These plaque-imaging techniques are still mostly used in specialized centers and research settings, but they represent a shift toward assessing risk based on what the plaque looks like rather than just how much it blocks the artery.

Point-of-Care Ultrasound in the Emergency Room

When someone arrives at an emergency department with stroke-like symptoms, every minute counts. Formal duplex ultrasound requires a trained vascular technologist and dedicated equipment, which is not always available at 2 a.m. This has led to growing interest in point-of-care ultrasound, performed by emergency physicians at the bedside with portable machines.

A pilot study found that emergency physicians using bedside carotid ultrasound achieved a sensitivity of 70 percent and specificity of about 87 percent for detecting significant stenosis.²¹ Those numbers are not precise enough to replace a formal vascular study, but they can help triage: quickly flagging a patient who likely has a severe carotid blockage so that definitive imaging and stroke-team activation happen faster. Bedside carotid ultrasound can also provide information about the cause of a stroke within minutes, which aids acute decision-making about treatments like clot-busting drugs.²²

A more recent study explored using common carotid artery flow-wave patterns to detect large-vessel occlusion stroke in the acute setting, finding that differences in end-diastolic velocity between the left and right sides could identify these emergencies with impressive accuracy.²³ This is still an evolving field, but the trajectory is clear: quick, low-cost bedside assessment is becoming a useful first filter in emergencies.

After Treatment, What Monitoring Looks Like

If you undergo a carotid endarterectomy (surgical plaque removal) or carotid stenting, the testing does not stop. Restenosis, a re-narrowing at the treatment site, can happen as the artery heals. Current surveillance recommendations call for an ultrasound within 30 days of the procedure, then every six months for two years, and annually after that, watching for stenosis that recurs above 70 percent.²⁴ The monitoring schedule is most intensive early on because restenosis is most likely in the first couple of years. If results stay stable for several years, some physicians extend the surveillance interval, though the evidence on exactly when to stop monitoring is still being refined.

Artificial Intelligence and Automated Analysis

The newest frontier in carotid testing is not a new scanning machine but software that reads the images. Deep-learning algorithms trained on thousands of CTA scans can now segment carotid plaque and classify its composition faster than a radiologist can, reducing analysis time from minutes to seconds. Hybrid deep-learning models have reached around 80 percent precision in plaque segmentation, and newer architectures achieve correlation coefficients near 0.89 for plaque volume measurements compared with expert readings.²⁵ On the ultrasound side, AI-assisted image segmentation combined with radiomic analysis (extracting measurable features from images that the human eye cannot easily quantify) has shown the ability to distinguish stable from unstable plaques, which could help predict which patients are most at risk of stroke.²⁶

One particularly promising application is motion analysis: tracking how a plaque moves during the heartbeat cycle. Systolic retractive motion biomarkers identified about 80 percent of vulnerable plaques in one study, offering a way to assess risk that goes beyond static snapshots.²⁵ These tools are not yet standard clinical practice, but they are moving from research into commercial products. Within the next few years, the ultrasound or CTA you get for carotid disease may come with an AI-generated risk report alongside the radiologist’s interpretation.

1
PubMed Central. A narrative review of the pathophysiology of ischemic stroke in carotid plaques
2
PubMed. Pathophysiological mechanisms of carotid plaque vulnerability: impact on ischemic stroke
3
PubMed Central. Carotid bruit for detection of hemodynamically significant carotid stenosis: the Northern Manhattan Study
4
PubMed Central. Carotid bruits as predictor for carotid stenoses detected by ultrasonography: an observational study
5
QJM: An International Journal of Medicine. Evaluation of the clinical utility of a carotid bruit
6
PubMed. Ophthalmologic findings as predictors of carotid artery disease
7
PubMed. Variation in Ultrasound Diagnostic Thresholds for Carotid Stenosis in the United States
8
PubMed Central. Systematic review and meta-analysis of the diagnostic value of computed tomography angiography for severe internal carotid artery stenosis
9
PubMed. Diagnostic accuracy of colour Doppler ultrasonography, CT angiography and blood-pool-enhanced MR angiography in assessing carotid stenosis
10
PubMed Central. Quantification of carotid stenosis on CT angiography
11
PubMed. Computed tomography angiography of the carotid arteries at low kV settings: a prospective randomised trial assessing radiation dose and diagnostic confidence
12
PubMed. Diagnostic accuracy of magnetic resonance angiography for internal carotid artery disease: a systematic review and meta-analysis
13
American Journal of Neuroradiology. Detection of Carotid Artery Stenosis: A Comparison between 2 Unenhanced MRAs and Dual-Source CTA
14
PubMed Central. Carotid Artery Stenosis: Comparison of 3D Time-of-Flight MR Angiography and Contrast-Enhanced MR Angiography at 3T
15
PubMed. Safety of arch aortography for assessment of carotid arteries
16
PubMed. Equivalence of measurements of carotid stenosis. A comparison of three methods on 1001 angiograms
17
Advances in Interventional Cardiology. Misclassification of carotid stenosis severity with area stenosis-based evaluation by computed tomography angiography
18
PubMed. Screening for Asymptomatic Carotid Artery Stenosis: US Preventive Services Task Force Recommendation Statement
19
PubMed. Hemorrhage in the atherosclerotic carotid plaque: a high-resolution MRI study
20
PubMed. Presence of intraplaque hemorrhage stimulates progression of carotid atherosclerotic plaques: a high-resolution magnetic resonance imaging study
21
PubMed Central. Can Emergency Physicians Perform Carotid Artery Point-of-Care Ultrasound to Detect Stenosis in Patients with TIA and Stroke? A Pilot Study
22
PubMed. Point-of-care ultrasound for stroke patients in the emergency room
23
PubMed Central. Point-of-care ultrasound of the common carotid arteries for detection of large vessel occlusion stroke: Results of the POCUS-LVO study
24
PubMed. Follow-up after carotid endarterectomy and stenting: What to look for and why
25
PubMed Central. Artificial intelligence in carotid computed tomography angiography plaque detection: Decade of progress and future perspectives
26
PubMed Central. Combining artificial intelligence assisted image segmentation and ultrasound based radiomics for the prediction of carotid plaque stability