What Do Cancerous Lymph Nodes Look Like on Ultrasound?

Cancerous lymph nodes on ultrasound tend to look rounder than normal, darker (more hypoechoic), and they lose the bright central stripe that healthy nodes display. Radiologists assess a combination of features rather than relying on any single sign, because no one characteristic is perfectly reliable on its own. Understanding what those features are and how they fit together can help you make sense of an ultrasound report or a conversation with your doctor.

What a Normal Lymph Node Looks Like First

To spot what’s wrong, it helps to know what’s right. A healthy lymph node on ultrasound typically appears as a small, oval or kidney-bean-shaped structure. Its outer rim, called the cortex, shows up as a thin, dark (hypoechoic) layer, while the center contains a bright (hyperechoic) stripe known as the fatty hilum. Think of it like an almond: a thin darker shell around a lighter core. Blood flows into the node through vessels at the hilum and fans outward in an orderly pattern. When any of these features change, the sonographer starts paying closer attention.

The Core Signs of Malignancy on Grey-Scale Ultrasound

Grey-scale, or B-mode, ultrasound is the standard black-and-white image most people picture when they think of an ultrasound scan. Several features on this image raise suspicion for cancer.

Round shape: Normal nodes are elongated, with the long axis roughly twice the short axis or more. Cancerous nodes become rounder as tumor cells expand them from within. A long-to-short-axis ratio approaching 1 (essentially a circle in cross-section) is one of the most consistent red flags.
Loss of the fatty hilum: That bright central stripe disappears when tumor cells replace the normal fatty tissue. An absent hilum is one of the strongest single indicators of malignancy on ultrasound.
Intranodal necrosis: Pockets of dead tissue inside the node can appear as irregular dark areas, sometimes giving the node a cystic or partially liquid-filled look. This is distinct from the smooth, central brightness of a normal hilum.
Calcifications: Tiny bright dots with shadowing behind them sometimes appear inside cancerous nodes, particularly with certain tumor types like papillary thyroid cancer.
Cortical thickening: Instead of a thin, even dark rim, the cortex becomes abnormally thick, or it thickens unevenly in one spot, forming a focal bulge. This asymmetric thickening can be an early sign of tumor deposits before the whole node is overtaken.

These features were described in a widely referenced classification system for axillary lymph nodes, where nodes with focal cortical lobulation or a completely dark appearance with an absent hilum were classified as likely metastatic, while nodes with thinner, more uniform cortices were considered benign.¹ A review of cervical node imaging confirmed that this package of signs, including shape, absent hilus, necrosis, and calcification, helps distinguish metastatic and lymphomatous nodes from benign ones.²

Blood Flow Patterns on Doppler

When the sonographer switches to color or power Doppler mode, you see a map of blood flow inside the node. This is where things get especially telling. Normal, reactive lymph nodes receive blood through the hilum, so their flow pattern looks orderly and central, fanning outward like the branches of a tree.

Cancerous nodes disrupt that tidy architecture. Tumor cells recruit their own blood vessels (a process called angiogenesis), and those new vessels tend to grow in disorganized patterns around the edges of the node or in chaotic clusters scattered throughout. On Doppler, this shows up as peripheral flow (vessels around the rim rather than through the center) or mixed flow (a jumbled combination of both). A study of cervical lymph nodes found that central or hilar blood flow was characteristic of benign nodes, while peripheral or mixed flow was characteristic of malignant nodes, with the difference being highly statistically significant.³ Researchers have described several vascular patterns in detail: a hilar type with orderly branching from the center, a spotted type with chaotic vessel fragments scattered inside the node, a peripheral type with vessels rimming the outside in a “basket” pattern, and a mixed type combining more than one of these distributions.⁴

Beyond the visual pattern, spectral Doppler can measure how resistant the blood flow is inside those vessels. Two numbers come up frequently in reports: the resistive index and the pulsatility index. Malignant nodes tend to have much higher values for both. In one study, the average resistive index in metastatic nodes was about 0.92 compared to 0.59 in benign nodes, and the average pulsatility index was roughly 2.66 versus 0.90.⁵ Another study confirmed that volume, pulsatility index, and resistive index were all significantly higher in malignant nodes, while the long-to-short axis ratio was significantly lower.⁶ The high resistance likely reflects the abnormal, tangled vessels that tumors create: blood is being pushed through poorly formed channels that don’t dilate the way healthy vessels do.

Elastography Adds Stiffness Information

Elastography is a newer ultrasound technique that measures tissue stiffness. Cancerous tissue tends to be stiffer than normal tissue because tumor cells pack densely and often trigger fibrosis around them. Ultrasound elastography can display this stiffness as a color map overlaid on the standard image, or it can generate a numerical value called shear wave velocity.

One study measured shear wave velocity in cervical lymph nodes and found that malignant nodes averaged about 3.96 meters per second, compared to about 2.71 m/s for benign nodes. Using a cutoff of 2.93 m/s, the test caught roughly 93% of malignant nodes (sensitivity) with a specificity around 75%.⁷ A broader review noted that elastography can detect early, localized tumor deposits within a node before they distort the overall shape or hilum, which gives it an edge in catching cancer at an earlier stage.⁸

Elastography is not yet standard everywhere, and its accuracy depends on the operator and the equipment. But it is increasingly used as an add-on to conventional ultrasound, particularly when the grey-scale and Doppler findings are ambiguous.

How the Appearance Varies by Cancer Type

Not all cancers leave the same fingerprint on a lymph node. The specific ultrasound features that dominate depend heavily on the type of primary tumor.

Papillary Thyroid Cancer

Thyroid cancer metastases in neck lymph nodes have some of the most distinctive appearances on ultrasound. The nodes often become hyperechoic (brighter than the surrounding muscle) rather than darker, which is somewhat unusual for metastatic disease. Microcalcifications, those tiny bright dots that throw shadows, are a hallmark, reflecting the same calcification patterns seen in the primary thyroid tumor. Cystic change, where part of the node fills with fluid, is also characteristic. A study of papillary thyroid cancer patients found that absent hilum appeared in about 88% of metastatic nodes, hyperechogenicity in 86%, a round shape in 80%, calcifications in roughly half, and cystic necrosis in about 20%. In benign nodes from the same patients, calcifications and cystic necrosis were absent entirely.⁹ Because of these distinctive features, ultrasound is often the first-line imaging tool when thyroid cancer recurrence is suspected.¹⁰

Breast Cancer

In breast cancer, the axillary (armpit) lymph nodes are the main focus. Rather than looking at the whole node’s shape, radiologists pay particular attention to cortical thickness. A cortex measuring 3 mm or more is a widely used threshold. One study found that a cortex at or above 3 mm was present in about 83% of nodes with metastasis versus only about 23% of nodes without, and having that degree of thickening was associated with roughly 17 times greater odds of metastasis.¹¹ Another study confirmed that an abnormal hilum and diffuse cortical thickening were both independently associated with nodal metastasis in breast cancer patients.¹² Focal thickening, where just one part of the cortex bulges outward, can be even more suspicious than generalized thickening, because it may indicate a single metastatic deposit has started growing.

Melanoma

Melanoma metastases in regional lymph nodes can appear strikingly dark on ultrasound, sometimes almost black, because of the melanin content and the densely packed tumor cells. Focal, nodular, or asymmetric cortical thickening is an early sign, and hilum loss or displacement combined with chaotic vascular signals is considered virtually diagnostic regardless of node size.¹³ Surveillance protocols for melanoma patients often use simple criteria: a hypoechoic, round or oval node lacking a visible hilum is considered metastatic, while nodes showing some but not all of those features, or with irregular margins or focal dark spots, are considered suspicious and warrant further workup.¹⁴

Lymphoma

Lymphoma is the odd one out in many ways. Because lymphoma originates in the lymph node itself rather than spreading there from somewhere else, affected nodes often enlarge dramatically while initially keeping a more uniform internal appearance than metastatic nodes do. They tend to be round and markedly hypoechoic, sometimes so uniformly dark that they can mimic a cyst on high-resolution scanners. Studies have found that lymphomatous nodes are frequently heterogeneous and show a distinctive micronodular pattern inside, though true posterior acoustic enhancement, which would make them look cystic, is actually uncommon.

Head and Neck Squamous Cell Carcinoma

For cancers arising in the mouth, throat, or larynx, metastatic cervical nodes are diagnosed using size criteria combined with internal architecture. A node thickness greater than 6 mm along with intranodal lesions suspicious for a metastatic focus are key criteria.¹⁵ These nodes frequently show intranodal necrosis (dead tissue pockets), because squamous cell carcinoma metastases tend to outgrow their blood supply quickly.

Benign Conditions That Look Like Cancer

Ultrasound is good at raising suspicion, but it is not perfect at confirming cancer. Several benign conditions can produce nodes that look alarmingly similar to malignant ones. Granulomatous diseases like tuberculosis and sarcoidosis are notorious mimics: they can cause nodes to lose their hilum, develop internal necrosis, and even calcify, checking many of the same boxes that metastatic disease does.¹⁶

Reactive lymph nodes, those that enlarge in response to a nearby infection, generally maintain their oval shape and visible hilum. But sometimes a vigorous immune response can make a node quite large and slightly rounder, which creates a gray zone on ultrasound. Kikuchi disease, cat-scratch disease, and other unusual infections can push nodes into that ambiguous appearance. This is why imaging alone rarely gives a definitive cancer diagnosis. When ultrasound features raise enough suspicion, the next step is usually a biopsy.

From Suspicion to Confirmation Through Biopsy

Ultrasound’s real strength in lymph node assessment is not just identifying suspicious nodes but guiding a needle directly into them for tissue sampling. Ultrasound-guided fine-needle aspiration draws out a small number of cells through a thin needle, while core needle biopsy uses a slightly larger needle to extract a sliver of tissue that preserves the node’s architecture. Both are done in real time, with the radiologist watching the needle on the screen to ensure it reaches the right spot.

A large meta-analysis of axillary lymph node biopsies in breast cancer found that fine-needle aspiration had an overall sensitivity of about 79% and specificity of about 96%, while core needle biopsy had slightly better sensitivity at about 85% with specificity around 93%.¹⁷ Another meta-analysis similarly found core needle biopsy to be the more sensitive technique, catching about 88% of metastatic nodes compared to 73% for fine-needle aspiration.¹⁸ A head-to-head comparison at a single institution confirmed this pattern, with core needle biopsy achieving 88% sensitivity and fine-needle aspiration reaching about 73%, while both had perfect specificity when a positive result was returned.¹⁹

What this means practically is that if a biopsy comes back positive for cancer, you can trust it. A negative result is slightly less certain, especially with fine-needle aspiration, because the needle may have sampled a part of the node that was not yet involved. This is one reason clinicians sometimes recommend a repeat biopsy or surgical excision when the clinical suspicion remains high despite a negative needle result.

Contrast-Enhanced Ultrasound

Contrast-enhanced ultrasound uses tiny gas-filled microbubbles injected into a vein to make blood vessels inside the node far more visible than standard Doppler allows. The technique can reveal the microvasculature of the node in real time, showing whether the blood vessels branch in orderly patterns or grow in the chaotic, irregular ways that tumors encourage.

There are two approaches. Perfusion contrast-enhanced ultrasound involves standard intravenous injection to visualize the blood flowing through a node, while lymphatic contrast-enhanced ultrasound uses a direct injection into the tissue near the node to trace lymphatic drainage, which is useful for identifying sentinel nodes in breast cancer staging.²⁰ A pilot study using super-resolution contrast imaging found that benign nodes showed significantly higher microvascular density and flow measures than malignant nodes in the early seconds after a low-dose injection, with the gap widening over time.²¹ Contrast-enhanced ultrasound is still more of a research and specialty-center tool than an everyday clinical standard, but it is gaining ground in breast cancer staging and head-and-neck oncology.

Artificial Intelligence in Lymph Node Assessment

Machine learning models trained on thousands of ultrasound images are increasingly being tested as aids for radiologists. The idea is not to replace the human reader but to flag suspicious nodes that might otherwise be overlooked, or to offer a second opinion on borderline cases.

A large meta-analysis of ultrasound-based machine learning models found pooled sensitivity of about 84% and specificity of about 85% for distinguishing benign from malignant lymph nodes.²² A meta-analysis focused specifically on predicting lymph node metastasis in breast cancer reported a pooled sensitivity of about 88% and specificity of about 75%.²³ A multicenter study using AI with endoscopic ultrasound, where the probe is placed inside the gastrointestinal tract to image nearby lymph nodes, reported even higher numbers: sensitivity around 99% for malignant nodes with an overall accuracy above 98%.²⁴

Those numbers are promising, but some context matters. The endoscopic ultrasound study reported performance in a controlled research setting, and real-world accuracy tends to dip once models encounter a wider variety of patients, machines, and operators. The meta-analyses that pool many different studies tend to show more modest performance, which is probably closer to what you’d expect in everyday clinical use. Still, AI is moving from experimental to practical in radiology departments, and lymph node assessment is one of the areas where it could help most, since the features are often subtle and the stakes of missing a positive node are high.

Why Equipment and Technique Matter

Ultrasound is uniquely operator-dependent. The same node can look dramatically different depending on the transducer frequency, the angle of the probe, and the scanner’s settings. Higher-frequency transducers give better resolution for superficial nodes in the neck or armpit but cannot reach deeper nodes in the abdomen. Adjusting the dynamic range, beam steering, and speckle reduction settings can all change how clearly internal features are displayed.²⁵

This variability means that a lymph node ultrasound is only as good as the person performing it and the equipment they use. A subtle focal cortical thickening might be invisible at a low transducer frequency or with suboptimal gain settings. Experienced sonographers at high-volume cancer centers tend to catch more abnormalities because they see more abnormal nodes, their pattern recognition is sharper, and their machines are typically state-of-the-art. If you are having lymph nodes evaluated because of known or suspected cancer, the quality of the ultrasound center matters more than you might think. A specialist oncology imaging center is worth the trip if one is available.

For the same reason, comparing ultrasound images from different dates or different facilities can be tricky. A node that looks unchanged on paper may actually have developed early cortical thickening that the second scan missed because of a different technique or machine. Radiologists account for this by reading images in context, always comparing current findings to prior studies performed at the same facility when possible.