Mammography vs. Tomosynthesis: What’s the Difference?

Tomosynthesis, often called 3D mammography, takes multiple low-dose X-ray images of the breast from different angles and reconstructs them into thin slices, while standard digital mammography captures a single flat image from each direction. That distinction in how the pictures are acquired changes what radiologists can see, how often they call patients back for extra testing, and which cancers get caught early. The real-world differences are meaningful but more nuanced than marketing materials tend to suggest.

How the Two Technologies Actually Work

In a conventional digital mammogram, the X-ray tube fires once from a fixed position, producing a two-dimensional image of the compressed breast. Everything inside the breast is flattened into a single plane, which means overlapping tissue can hide a tumor or, just as often, mimic one. During a tomosynthesis exam, the X-ray tube sweeps in a short arc, typically over about fifteen degrees, firing a series of low-dose exposures. Software then reconstructs those exposures into a stack of thin slices, usually about one millimeter thick. Radiologists scroll through those slices the way you might flip through pages of a book, examining each layer of tissue individually.

The practical effect is that overlapping tissue gets separated. A dense cluster of normal breast tissue that might look suspicious on a flat image becomes clearly benign when you can see it in its own slice. Conversely, a small mass buried behind dense tissue becomes visible because it is no longer hidden by the layers in front of and behind it. The first breast-specific tomosynthesis system received FDA approval in 2011, and since then several manufacturers have released their own versions.¹ Today, a large share of screening mammograms in the United States are performed with tomosynthesis.

Finding More Cancers

The clearest advantage of tomosynthesis over standard mammography is a higher cancer detection rate. A population-based screening program in Norway found that tomosynthesis detected roughly 9.4 cancers per 1,000 women screened, compared with 6.1 per 1,000 for standard digital mammography.² A large U.S. study of women aged 40 to 64 reported a detection rate of about 4.9 cancers per 1,000 screens with tomosynthesis versus 3.8 with standard mammography.³ And a decade-long screening comparison found tomosynthesis detected about 5.3 cancers per 1,000 compared with 4 per 1,000 for digital mammography.⁴

Importantly, the extra cancers that tomosynthesis catches are mostly invasive cancers rather than ductal carcinoma in situ, the very early, non-invasive form that is more controversial to treat.⁵ That matters because finding more invasive cancers while not inflating the count of borderline diagnoses is exactly the profile you want from a screening upgrade. It means the technology is catching biologically significant disease, not just adding to overdiagnosis.

Fewer Callbacks, Most of the Time

Being called back after a screening mammogram for additional images is stressful, and the majority of callbacks turn out to be false alarms. One of the early selling points of tomosynthesis was that it would reduce these false-positive recalls, and the evidence largely supports that. A major JAMA study found that adding tomosynthesis to digital mammography dropped the recall rate by about 16 per 1,000 screens.⁶ Over a full decade of annual screening, the cumulative chance of at least one false-positive recall was roughly 50% with tomosynthesis versus 56% with digital mammography.⁷ That gap is real, but notice how high both numbers still are: even with the newer technology, annual screening leads to about a coin-flip chance of being called back at some point over ten years.

For biennial screening, the difference narrows further. Cumulative false-positive recall rates dropped to about 36% for tomosynthesis versus 38% for digital mammography, and the differences in biopsy recommendations between the two technologies essentially vanished.⁷ So the recall advantage of tomosynthesis is most meaningful for people who screen every year.

The picture is also not uniformly positive. Data from the Malmö Breast Tomosynthesis Screening Trial in Sweden actually showed a higher false-positive recall rate with tomosynthesis than with standard mammography, particularly in the trial’s first year when radiologists were still adapting to the new images.⁸ Part of the explanation is that tomosynthesis reveals architectural distortions, subtle tissue disruptions that can signal cancer but are also frequently benign. Radiologists seeing these for the first time on the new modality called patients back more often until they built experience interpreting them. The false-positive rate in that trial settled down by the second year, illustrating how much radiologist familiarity affects outcomes.

The Interval Cancer Question

An interval cancer is one that shows up between routine screenings, meaning it was either missed at the last screening or grew quickly afterward. If tomosynthesis finds more cancers upfront, you might expect fewer interval cancers. The evidence here is genuinely mixed. The Malmö trial found a lower interval cancer rate with tomosynthesis: about 1.6 per 1,000 women versus 2.8 per 1,000 in a control group screened with digital mammography.⁹

But a large U.S. study published in JAMA found no significant difference in interval invasive cancer rates between the two technologies across risk categories and breast density groups.¹⁰ A randomized trial in Italy similarly found that interval cancer incidence was essentially the same in both arms.¹¹ This is one of the unresolved questions in the field. Detecting more cancers at screening is clearly a benefit, but whether it actually translates into catching cancers earlier enough to prevent them from becoming interval cancers remains uncertain.

Dense Breasts and Who Benefits Most

Breast density is one of the biggest factors in mammographic accuracy. Dense tissue appears white on a mammogram, and so do tumors, making it harder to distinguish one from the other. Tomosynthesis helps here because scrolling through individual slices separates dense tissue from underlying masses. A large Norwegian study found that the true-positive rate for tomosynthesis was higher across most density categories, primarily because it caught more spiculated masses and architectural distortions that standard mammography missed.¹² The improvement was statistically strong for women with heterogeneously dense or scattered fibroglandular breast tissue (the two middle density categories, which happen to include the majority of women). For extremely dense breasts and almost entirely fatty breasts, the differences were smaller and not always statistically reliable.

A study of women with a family history of breast cancer found that tomosynthesis reduced the recall rate and improved specificity compared with digital mammography, particularly among women with one first-degree relative with the disease. However, actual cancer detection rates were not significantly different between the two technologies in that population.¹³ That finding is a useful reminder that the benefits of tomosynthesis are not always about finding more cancers. Sometimes the payoff is fewer false alarms, which for a high-anxiety population like women with family histories can be its own kind of clinical benefit.

What About Radiation Dose

Because tomosynthesis involves multiple exposures, the natural concern is that you are getting more radiation. The answer depends on how the exam is configured. When tomosynthesis is performed as a stand-alone exam, the dose is comparable to or even lower than a standard two-view mammogram, with dose ratios ranging from about 0.34 to 1.17 depending on the system and number of views.¹⁴ The problem arises when clinics perform both a full standard mammogram and a tomosynthesis exam at the same visit, which roughly doubles the dose.

The workaround is synthetic mammography: software generates a standard-looking 2D image from the same tomosynthesis data, so there is no need for a separate conventional exposure. A meta-analysis found that these synthetic images are diagnostically comparable to conventional mammograms, with similar sensitivity and specificity.¹⁵ A reader study comparing the two approaches directly found no significant difference in performance.¹⁶ Using synthetic images instead of a separate conventional mammogram cuts the dose by roughly 45%.¹⁴ Most facilities today use this combined approach, making the radiation concern largely a solved problem.

Synthetic mammograms are not perfect substitutes, though. They can produce a higher false-positive rate for calcifications, may blur some asymmetries, and occasionally introduce imaging artifacts.¹⁷ Radiologists need to be aware of these quirks, but they do not generally change the screening workflow in a meaningful way for patients.

It Takes Longer to Read

There is no free lunch in breast imaging. Scrolling through dozens of thin slices takes more time than reviewing a flat image. One study found that interpretation of combined tomosynthesis and mammography exams averaged about 2.8 minutes versus 1.9 minutes for digital mammography alone, a roughly 47% increase in reading time per case.¹⁸ Another study using a case-mix scenario found even larger differences, with combined reading averaging 77 seconds versus 33 seconds for digital mammography.¹⁹

Experience makes a measurable difference. Radiologists with five or more years of tomosynthesis experience read cases faster than their less-experienced colleagues, and reading time decreased after less than nine months of accumulated screening experience.²⁰ Recall rates relative to standard mammography also dropped promptly once radiologists began reading tomosynthesis, without an extended learning period.²¹ So while there is a genuine workflow cost, it diminishes as radiologists get comfortable with the technology.

What the Exam Feels Like

From your perspective as a patient, a tomosynthesis exam is nearly identical to a standard mammogram. The machine looks the same, the breast is still compressed between two plates, and the exam takes only slightly longer because the X-ray tube needs a few extra seconds for its sweeping arc. One study found that reduced compression was feasible during tomosynthesis and that patients experienced significantly less pain and anxiety with lighter compression compared with standard compression.²² This does not mean every facility uses lighter compression for tomosynthesis. But the technology does allow some room to reduce it because the 3D reconstruction partly compensates for the tissue overlap that compression was needed to control. Over time, you may see more facilities take advantage of this.

Architectural Distortions and How Tomosynthesis Changes What Gets Found

Tomosynthesis is particularly good at finding architectural distortions, areas where the breast tissue looks pulled or disrupted. On a flat mammogram, these can be subtle or invisible; on tomosynthesis, they pop out. A study of symptomatic patients found that architectural distortion seen on tomosynthesis had a positive predictive value for malignancy of about 42%, and nearly half the cases were visible on only one mammographic view, making them easy to miss on standard imaging.²³ The flip side is that tomosynthesis also finds more benign-looking distortions that would never have been flagged on standard mammography, contributing to the higher false-positive rate seen in some trials during early adoption.

Cost and Access

Tomosynthesis machines are more expensive than standard digital mammography units, and reading times are longer, both of which add costs. In many countries, the question of whether the clinical benefit justifies the expense is still being debated. A cost-effectiveness analysis modeling a 30-year screening horizon in Brazil found that tomosynthesis plus synthetic mammography actually saved about €954 per patient over that period compared with digital mammography, because early detection offset downstream treatment costs, while also gaining quality-adjusted life years.²⁴ That said, cost-effectiveness models are sensitive to local healthcare prices, screening intervals, and population demographics, so a finding in one country does not automatically apply elsewhere. In the United States, Medicare and most private insurers now cover tomosynthesis as a routine screening option, which has removed the cost barrier for many patients.

When Mammography and Tomosynthesis Are Not Enough

Neither standard mammography nor tomosynthesis is the end of the conversation for every patient. For women at very high risk, such as BRCA mutation carriers, MRI is substantially more sensitive. One study of BRCA carriers and women at high familial risk found MRI sensitivity of 90% compared with about 38% for mammography and ultrasound combined.²⁵ Supplemental ultrasound has also been shown to find additional cancers in women with dense breasts, though at the cost of lower specificity and more false positives. These modalities are generally not substitutes for mammography but additions to it for specific risk groups. For average-risk screening, the choice between standard mammography and tomosynthesis is the relevant one, and tomosynthesis has become the default at many large facilities.

Artificial Intelligence and Where This Is Heading

The volume of image data produced by tomosynthesis, sometimes hundreds of slices per exam, is one reason reading times are longer. AI-based triage systems are being developed to help with exactly this bottleneck. A recent paired noninferiority trial tested an AI system designed to automatically flag suspicious regions in both standard mammography and tomosynthesis images, outputting a risk score and a visual overlay marking the areas of concern.²⁶ The idea is not to replace radiologists but to route clearly normal exams away from detailed human review and direct attention toward the most suspicious cases. Whether these systems ultimately shorten reading times enough to close the gap between tomosynthesis and standard mammography workflow is still being studied, but it is the most active area of development in breast imaging right now.