Colon polyps come in a surprisingly wide range of shapes, sizes, and surface textures, and each visual feature gives doctors clues about whether a polyp is harmless or potentially dangerous. Most polyps found during colonoscopy fall into a few recognizable categories: raised bumps on a stalk, flat mounds that blend into the surrounding tissue, or subtle depressions that barely register on camera. Roughly one in four adults screened by colonoscopy has at least one adenoma, the type most likely to become cancerous over time. Understanding what different polyps look like and why their appearance matters can help you make sense of a colonoscopy report and the follow-up schedule your doctor recommends.
Why Shape Matters More Than You Might Think
When a gastroenterologist peers through the colonoscope, the first thing they assess is a polyp’s shape and how it sits on the colon wall. Shape is not just a cosmetic detail. It affects how easily the polyp can be removed, how likely it is to harbor abnormal cells, and even whether it gets spotted at all. Doctors generally group polyp shapes into three broad categories based on the Paris classification system, which was developed specifically to standardize how endoscopists describe what they see.
Pedunculated polyps are the classic image most people picture: a mushroom-like growth dangling from a narrow stalk. The stalk makes these polyps relatively easy to identify on camera, as they sway with the movement of fluid in the colon. They’re also usually straightforward to remove by looping a wire snare around the base. Sessile polyps, by contrast, sit flat against the colon wall with a broad base and no stalk. They look more like a dome or a small bump. Because they’re flush with the tissue, they can be trickier to snare cleanly, especially as they grow larger.
Flat and depressed polyps are the hardest to see. These lesions rise only slightly above the mucosal surface, or in the case of depressed types, actually dip below it. One imaging study found that fewer than half of flat lesions could be detected on CT colonography, and those that were visible measured at least 2 millimeters in height and 7 millimeters across. Lesions just 1 millimeter tall were invisible on the scan.1PubMed. Flat polyps of the colon: detection with 16-MDCT colonography–preliminary results Flat polyps are clinically significant because, despite their inconspicuous appearance, they can carry a disproportionately high risk of containing advanced or dysplastic tissue.
Adenomas and the Road to Cancer
Adenomas are the polyps that get the most attention because they sit on the pathway that can eventually lead to colorectal cancer. They account for the majority of polyps found during screening. A large meta-analysis covering more than 637,000 people estimated the overall prevalence of adenomas at about 24%, with men affected more often than women and prevalence climbing with age.2PubMed. Global Prevalence of Colorectal Neoplasia: A Systematic Review and Meta-Analysis Not every adenoma is equally worrying, though. What the polyp looks like under the microscope, and even during the procedure itself, tells your doctor a lot about the level of risk.
Under the microscope, adenomas are divided into tubular, villous, and tubulovillous types. Tubular adenomas have a relatively uniform, tube-like glandular structure and are the most common and least aggressive. Villous adenomas have finger-like projections that give them a shaggy or frond-like surface, and they carry higher malignant potential. A landmark long-term study found that patients with tubulovillous, villous, or large adenomas (1 centimeter or bigger) in the rectosigmoid had about 3.6 times the expected rate of future colon cancer. Those with multiple such adenomas faced a risk roughly 6.6 times higher. By contrast, patients with only small tubular adenomas actually had a cancer rate below the general population’s expected baseline.3PubMed. Long-term risk of colorectal cancer after excision of rectosigmoid adenomas
During a colonoscopy, adenomas tend to appear reddish compared to the paler surrounding mucosa, and their surface may look slightly rough or irregular. Larger adenomas sometimes have a lobulated or cauliflower-like surface. Size is one of the strongest visual risk predictors: polyps under 1 centimeter are less likely to contain villous tissue or high-grade changes, while those over 1 centimeter warrant closer attention and more aggressive follow-up.
Hyperplastic Polyps and the Serrated Family
Hyperplastic polyps are the most common non-adenomatous polyps you’ll encounter in a colonoscopy report. For decades they were considered completely harmless, and many small ones still are. They tend to be tiny, pale, and smooth, typically appearing in the rectum and lower left colon. On camera, they often look like small, flat, slightly glistening bumps that are hard to distinguish from normal mucosa without enhanced imaging.
The view that all hyperplastic polyps are innocent has shifted, however. Research has shown that some hyperplastic polyps may serve as precursors to colorectal cancer through a distinct molecular pathway involving DNA methylation and mismatch repair defects, rather than the classic adenoma-to-cancer sequence.4PubMed. Hyperplastic polyps and colorectal cancer: is there a link? This realization helped drive the recognition of a broader “serrated” family of polyps, which includes traditional hyperplastic polyps, sessile serrated lesions, and traditional serrated adenomas.
Sessile serrated lesions deserve special attention because they’re particularly easy to miss. They tend to be flat, pale, and located in the right colon, where the bowel wall is darker and folds can obscure them. Their color closely matches the surrounding tissue, and their margins are often indistinct. A characteristic visual clue is a mucous cap: a yellow or whitish film of mucus sitting on the polyp’s surface. One study found that the combination of a mucous cap with prominent varicose-appearing vessels was the strongest predictor of sessile serrated lesions during endoscopy.5PubMed Central. Endoscopic Features of Mucous Cap Polyps: A Way to Predict Serrated Polyps These polyps also tend to be larger than typical hyperplastic polyps, often exceeding 10 millimeters, and they have a cloud-like surface with an interrupted mucosal vascular pattern.6Revista de GastroenterologÃa de México (English Edition). Serrated polyps of the colon and rectum: a concise review
Rare and Non-Neoplastic Polyps
Not every polyp found during colonoscopy is on the path to cancer or even part of the serrated family. Juvenile polyps, for example, are hamartomatous growths that are the most common intestinal polyp in children but show up only rarely in adults.7PubMed Central. A juvenile polyp on screening colonoscopy On colonoscopy, juvenile polyps usually appear as smooth, round, cherry-red growths, often with a stalk. They can bleed easily, which is typically how they announce themselves in children. A single juvenile polyp is generally benign, but multiple juvenile polyps or a family history of them can point toward juvenile polyposis syndrome, which does carry an elevated cancer risk and requires closer surveillance.
Inflammatory polyps, sometimes called pseudopolyps, are another non-neoplastic type. They tend to show up in people with inflammatory bowel disease and are essentially tags of regenerating tissue left behind after episodes of severe inflammation. They can look quite dramatic on camera, sometimes forming long, finger-like projections or bridging across the lumen, but they themselves are not precancerous. Their main significance is that they can make it harder for the endoscopist to spot true neoplastic polyps hiding among them.
How Doctors Read a Polyp’s Surface During the Procedure
Modern endoscopy goes well beyond simply spotting a bump on the colon wall. Gastroenterologists now use classification systems that let them predict, with reasonable accuracy, whether a polyp is likely neoplastic or benign before it’s even removed. This “optical diagnosis” approach relies on close inspection of the polyp’s surface pattern, color, and vessel architecture under enhanced imaging modes.
One of the earliest and most studied systems is Kudo’s pit pattern classification, which uses magnification endoscopy to examine the tiny openings (pits) on a polyp’s surface. Round or star-shaped pits suggest normal or hyperplastic tissue. Elongated, branching, or gyrus-like pit patterns point toward neoplastic growth. Featureless or structureless areas raise concern for invasive cancer. A meta-analysis of studies using this system found it achieved about 89% sensitivity and 86% specificity for identifying neoplastic lesions.8PubMed Central. Kudo’s pit pattern classification for colorectal neoplasms: a meta-analysis
Narrow-band imaging, or NBI, takes a different approach. By filtering colonoscope light to blue and green wavelengths, NBI enhances the contrast of blood vessels and surface patterns without any dye. The NICE classification, designed specifically for NBI, sorts polyps into three types based on color, vessel pattern, and surface texture. In a large retrospective study, the NICE system reached sensitivities above 86% and specificities above 85% across its three types.9PubMed Central. Utility of the narrow-band imaging international colorectal endoscopic classification for optical diagnosis of colorectal polyp histology in clinical practice: a retrospective study A separate study using the related JNET classification found that while it correctly classified about 77% of lesions overall, the NICE system performed even better at roughly 94% correct classification.10Scientific Reports. Impact of narrow band imaging in prediction of histology of advanced colorectal neoplasia The practical upside is that an experienced endoscopist can make high-accuracy predictions about a polyp’s nature in real time, potentially allowing tiny, clearly benign-looking polyps to be removed and discarded without the cost of sending each one to pathology.
These classification systems are not foolproof. Their accuracy depends heavily on the endoscopist’s training and experience, and on the quality of the view. A polyp obscured by residual stool or mucus, or one sitting on a difficult fold, may not show its surface features clearly enough for a confident call.11PubMed Central. Colorectal polyp classification and management of complex polyps for surgeon endoscopists
The Bowel Prep Problem
Anyone who has gone through a colonoscopy remembers the bowel preparation more vividly than the procedure itself. Beyond patient comfort, the quality of that prep directly affects which polyps get found and which get missed, particularly the flat, subtle ones. A study comparing detection rates of sessile serrated polyps across different prep quality levels found a stark difference: high-quality preparation yielded a detection rate of 12%, while intermediate and low-quality preps dropped to roughly 4.6% and 4.9%, respectively. In other words, poor prep cut the detection of these already-hard-to-see polyps by about two-thirds.12Clinical Gastroenterology and Hepatology. High-quality Bowel Preparation Is Required for Detection of Sessile Serrated Polyps
This is especially relevant because sessile serrated lesions are the very polyps whose cancer risk was only recently appreciated. They’re pale, flat, often covered in mucus, and preferentially located in the right colon where residual fluid tends to pool. If prep is inadequate, these polyps essentially vanish into the background. When your doctor’s office stresses the importance of completing every ounce of that prep solution, this is the main reason: a clean colon wall is the single most important factor in making subtle polyps visible.
Where Polyps Hide in the Colon
Polyps do not distribute themselves evenly. Their location within the colon matters for both detection and risk. Adenomas and serrated lesions show different geographic preferences, and the anatomy of the colon itself creates blind spots that endoscopists must actively work to overcome.
The colon alternates between segments that are fixed to the back wall of the abdomen (the ascending colon, descending colon, and rectum) and segments that hang more freely on a mesentery (the transverse and sigmoid colons). This anatomy affects how easily the endoscope can be positioned. In fixed segments, maintaining alignment between the scope and the colon’s long axis is harder, which can reduce visibility along certain angles.13Clinical Endoscopy. Assessment of colorectal polyp detection along the short-axis direction of the intestinal lumen during colonoscopic observation: a retrospective study in Japan Polyps nestled behind folds, near the hepatic or splenic flexures, or along the inner curves of haustral folds are the ones most commonly missed during withdrawal.
The right colon, which includes the cecum and ascending colon, is where sessile serrated lesions concentrate. This region also tends to have more residual liquid despite good prep. Traditional hyperplastic polyps, meanwhile, cluster in the rectum and sigmoid. Advanced adenomas are found throughout but have a somewhat higher prevalence in the proximal colon among older adults. Knowing these patterns helps explain why thorough withdrawal technique and adequate inspection time are considered quality metrics for colonoscopy: rushing through the right colon is exactly how the most dangerous flat polyps get missed.
How Artificial Intelligence Is Changing Detection
AI-assisted colonoscopy has moved from experimental concept to clinical reality in the past several years. Computer-aided detection (CADe) systems highlight potential polyps on the live video feed, drawing the endoscopist’s eye to areas they might otherwise overlook. Computer-aided diagnosis (CADx) goes a step further, attempting to classify a polyp’s likely histology in real time based on its visual features.
A prospective study comparing CADx to standard visual inspection for identifying neoplastic polyps found similar accuracy: the AI system achieved about 90% sensitivity and 86% specificity, compared to roughly 88% sensitivity and 83% specificity for standard inspection. Where the AI notably outperformed was in confidence level. Endoscopists felt confident in their assessment about 74% of the time, while the CADx system delivered high-confidence predictions in about 93% of cases.14PubMed. Real-Time Artificial Intelligence-Based Optical Diagnosis of Neoplastic Polyps during Colonoscopy That higher confidence matters because it could allow more “diagnose and discard” decisions for tiny polyps, reducing the volume of tissue sent to pathology labs without sacrificing accuracy.
An important finding for training programs is that AI seems to level the playing field between experienced and novice endoscopists. A study comparing trainees using AI assistance to experts without it found that trainees with CADx achieved essentially the same negative predictive value as unassisted experts for small rectosigmoid polyps, around 90%.15PubMed. The Augmented Colonoscopy With Computer-Aided Polyp Characterization Study The AI alone, without any human input, performed slightly better still at about 93%. Researchers are also working to make these systems more transparent, developing explainable AI approaches that align with how endoscopists already think about features like color, vessel pattern, and surface texture, rather than simply outputting a black-box prediction.16Scientific Reports. A novel approach to overcome black box of AI for optical diagnosis in colonoscopy
What Happens After a Polyp Is Found
Most polyps are removed during the same colonoscopy in which they’re discovered. The choice of removal technique depends on the polyp’s size, shape, and location. Small, stalked polyps are typically snared with a wire loop and cut using electrical current. For flat polyps, endoscopic mucosal resection lifts the lesion off the underlying muscle layer by injecting fluid beneath it, creating a cushion that allows safe removal. Cold snare techniques, which cut without electrical energy, have gained popularity for their safety profile, particularly for larger flat polyps where the risk of bleeding or perforation needs to be minimized.17PubMed Central. Cold snare endoscopic mucosal resection for the removal of large nonpedunculated colon polyps
If a polyp shows signs that cancer has already invaded the deeper submucosal layer, endoscopic removal alone may not be sufficient, and surgical options come into play. Visual clues that raise suspicion for deep invasion include a depressed surface, irregular or absent pit patterns, and thickened, distorted blood vessels visible under enhanced imaging. This is why real-time classification during the procedure matters so much: it helps the endoscopist decide on the spot whether to proceed with resection or refer for a different approach.
Follow-Up Schedules Based on What Was Found
The type, number, and size of polyps removed at colonoscopy directly determine how soon you need to come back. Current guidelines lay out a tiered system. If you had just one or two small tubular adenomas (under 10 millimeters) completely removed during a high-quality exam, your next colonoscopy is recommended in seven to ten years. Three or four small tubular adenomas push that interval to three to five years. Five to ten adenomas warrant a repeat in three years. Any adenoma 10 millimeters or larger, any with villous tissue, or any with high-grade dysplasia calls for a three-year return. More than ten adenomas found at once trigger a one-year follow-up.18Gastroenterology. Follow-up after colonoscopy and polypectomy
Sessile serrated lesions have their own surveillance track. Because they were historically underrecognized and their cancer pathway is different from adenomas, guidelines now specify intervals based on their size, number, and whether they show dysplasia. A large sessile serrated lesion or one with dysplasia generally calls for a three-year follow-up, similar to an advanced adenoma. Small hyperplastic polyps in the rectum, by contrast, are treated the same as having no polyps at all for surveillance purposes.
These intervals assume the colonoscopy itself was thorough and the prep was good. If the endoscopist could not see the entire colon clearly, or if a large polyp was removed piecemeal and there’s concern about completeness, the recommended return may be much sooner, sometimes within three to six months. The visual appearance of the polyp during the procedure, combined with the pathologist’s reading of the tissue afterward, together determine what your follow-up looks like for years to come.
Virtual Colonoscopy and Its Limits for Polyp Identification
CT colonography, commonly called virtual colonoscopy, uses a CT scanner to produce detailed images of the colon without inserting an endoscope. For larger polyps, its accuracy approaches that of conventional colonoscopy. Research from major centers has shown detection rates comparable to optical colonoscopy for polyps over 10 millimeters in diameter.19American Journal of Roentgenology (AJR). Colon cancer screening with virtual colonoscopy: promise, polyps, politics However, this method struggles with the very polyps that are most easily missed during standard colonoscopy too: flat and sessile lesions. Since virtual colonoscopy relies on the polyp protruding enough to create a visible contour change on the scan, lesions that barely rise above the surface often go undetected.
Virtual colonoscopy also cannot characterize a polyp’s surface in the way that NBI or magnification endoscopy can. You get shape and size information but no color, no vessel detail, and no pit pattern analysis. If a suspicious polyp is found, you’ll still need a conventional colonoscopy to remove it and get a tissue sample. For people at average risk who are unable or unwilling to undergo standard colonoscopy, virtual colonoscopy offers a reasonable screening alternative for catching larger, clinically significant polyps, but it has real blind spots for flat and serrated lesions that modern endoscopy is increasingly designed to find.