Tag: mammography AI

AI for Breast Cancer Risk

Posted on by Jake Fishman

Artificial intelligence may be capable of identifying subtle mammographic signs of breast cancer years before conventional diagnosis, according to a new study published in Radiology. Researchers from Sweden found that three commercially available AI algorithms for mammography screening generated elevated cancer scores as early as 10 years before diagnosis, with detection signals strengthening as diagnosis approached.

Predicting breast cancer risk offers the prospect not only of detecting cancer earlier, but also of tailoring mammography screening to women most likely to benefit from it.

  • Clinical risk calculators like Tyrer-Cuzick and breast density analysis are available, but AI-based algorithms are showing promise by predicting risk from screening mammograms.

In the new study, researchers analyzed 89k mammograms from 31.4k women collected over a 10-year period, drawn from Sweden’s national screening program, where women aged 40-74 undergo biennial mammography interpreted by two radiologists.  

  • During the study period, 12.1k women (39%) were ultimately diagnosed with breast cancer. Three commercially available AI algorithms were used to generate risk scores (Vara AI from Vara, Lunit Insight MMG from Lunit, and MammoScreen from Therapixel). (It’s worth noting all three were originally designed for cancer detection rather than risk prediction.) 

AI scores increased progressively over time in women who later developed cancer, while remaining relatively stable among cancer-free participants…

  • At 90% specificity, AI systems flagged 19%-20% of future breast cancer cases six years before diagnosis.
  • Detection increased to 23%-25% at four years before diagnosis.
  • Performance rose further to 35%-39% at two years before diagnosis.
  • Even 10 years before diagnosis, the systems identified 13%-17% of future cancers.
  • Across all pre-diagnostic examinations, AI achieved AUC values of 0.63-0.67, outperforming mammographic density alone (AUC = 0.57).

The findings suggest that AI tools developed for cancer detection may also have value as early-alert systems for identifying women who could benefit from closer surveillance or supplemental imaging.

  • While prospective validation is still needed, sequential AI scoring may ultimately help identify women who would benefit from supplemental imaging, closer surveillance, or earlier intervention.

The Takeaway

The study adds to growing evidence that mammography AI can extend beyond cancer detection to long-term risk stratification. By identifying subtle imaging patterns years before diagnosis, AI-derived detection scores could provide an additional layer of longitudinal risk monitoring and help guide more personalized screening strategies.

AI Reduces Mammography Workload

Posted on by Brian Casey

Using AI to triage low-risk breast screening exams that don’t need extra review could remove more than three-quarters of mammography cases from radiologists’ workload and allow them to spend more time on high-risk cases. That’s according to a new study in Radiology: Artificial Intelligence that confirms other recent studies. 

Much of recent mammography AI research has focused on its ability to triage low-risk cases to avoid additional radiologist review – saving precious personnel resources.

  • This is particularly valuable in Europe, which uses a double-reading paradigm in which two radiologists review all mammography cases (the U.S. employs single readers but tends to screen women annually rather than every two years). 

The new study comes from France, which employs a slightly different paradigm from the rest of Europe. Double reading is conducted only for lower-risk BI-RADS 1 and 2 cases, while BI-RADS 3-5 go directly to diagnostic workup. 

  • As such, double reading occurs with cases that have low cancer prevalence, which can make it more difficult for radiologists to detect cancers that don’t occur very often.

But what if you offloaded low-risk double reading to AI? 

  • In the new paper, researchers tried that with Therapixel’s MammoScreen AI algorithm, which was employed retrospectively to analyze mammograms from 42.4k women acquired from 2015 to 2019.

AI results were compared to standard radiologist double reading, with the following findings…

  • AI classified 77% of cases as low-risk, meaning these could be safely triaged from the double-reading paradigm.
  • AI missed only one cancer in the low-risk group, a rate the researchers characterized as “small but measurable.” 
  • Eleven cancers were found in the group AI classified as non-low-risk, which would have undergone double reading anyway in the AI triage paradigm.
  • Rates of interval cancer (cancer that occurs between screening rounds) were 5X higher in the cases AI classified as non-low-risk compared to low-risk (2.16 vs. 0.47 cancers per 1k exams). 

Using AI to classify and remove low-risk cases from double reading could therefore save significant resources from the French mammography screening program, with a “small but non-zero risk” of missed cancers.

The Takeaway
The new results track with findings from other recent studies that apply AI to mammography screening, particularly in Europe. While the French reading paradigm is unique, it’s instructive to see that AI maintains its ability to reduce radiologist workload across different types of breast cancer screening programs.

Mammography AI Improves Breast Screening

Posted on by Brian Casey

Radiologists using a commercially available mammography AI algorithm saw improved diagnostic performance in breast cancer screening, mainly due to better specificity. The study adds to a growing body of research supporting mammography AI.

Mammography screening has been one of the most promising use cases for AI, and recent randomized controlled trials have demonstrated that AI can both improve diagnostic accuracy and speed up workflows. 

  • But RCTs are usually performed under highly controlled conditions in high-income Western countries, and the results might not be generalizable to other countries around the world. 

In the new study in Academic Radiology, researchers in Singapore tested Lunit’s Insight MMG algorithm in a retrospective review of a dataset of 302 digital mammograms that was enriched with 89 breast cancers.

  • Researchers noted that many countries have a high breast cancer incidence-to-mortality ratio due to limitations in population-based screening programs, and AI potentially could help. 

The authors focused on AI’s ability to improve the diagnostic performance of nine breast radiologists from four countries in Asia and North Africa who interpreted the mammograms, finding that AI assistance…

  • Improved radiologist accuracy as measured by AUC (from 0.799 to 0.851).
  • Generated a big jump in specificity (from 77% to 88%). 
  • And significantly reduced per-case image interpretation times (from 122 to 83 seconds per case).
  • Without changing sensitivity at a statistically significant level (83% vs. 82%, p = 0.73).

There were some subtle differences in the current study’s findings relative to previous research, some of which were the result of using a cancer-enriched dataset rather than a screening population as would be the case in an RCT.

  • The specificity improvement with AI would reduce unnecessary recalls in a population-based screening program and make mammography more cost-effective – an important consideration in countries with constrained public health budgets.

The Takeaway

The new study doesn’t have the statistical heft of a large, randomized controlled trial, but it still adds to the body of knowledge supporting AI for mammography, especially at facilities that haven’t been party to the large-scale RCTs.

Mammo AI Momentum Builds

Posted on by Brian Casey

Momentum is building toward routine clinical use of AI for breast cancer screening. Several new studies offer even more support for mammography AI, including research published today in Nature Medicine in which AI reduced radiologist workload by over 60% by excluding low-risk studies from human review.

Breast screening has become one of the most promising use cases for AI, with the potential to reduce radiologists’ workload while improving their ability to detect cancer. 

  • For example, the recent MASAI study found that ScreenPoint Medical’s Transpara AI algorithm could replace the second human reader in a double-reading protocol, reducing workload by 44% and improving cancer detection rates by 28%.

The new research in Nature Medicine also used Transpara, as part of the AITIC study in Spain with the goal of seeing if AI could triage low-risk studies so they don’t require review by human radiologists. 

  • AITIC had a prospective design, involving 31k women with screening exams split between 2D mammography (17k) and digital breast tomosynthesis (14k). 

Women in the control arm of the study got conventional double reading by two radiologists – the standard mammography paradigm in Europe.

  • The intervention arm used a partially autonomous AI approach: cases that AI interpreted as low risk were classified as normal and were not reviewed by radiologists, while all other cases were double-read by radiologists using AI support.

In analyzing the results, researchers found…

  • Workload in the AI arm was 64% lower than conventional double reading.
  • AI’s workload reduction was similar between DBT and conventional digital mammography (-66% and -62%, respectively).
  • The AI arm’s cancer detection rate per 1k women was 15% higher (7.3 vs. 6.3 cancers).
  • But the recall rate was also 15% higher.

It’s worth noting that the AITIC study differed from MASAI in its inclusion of DBT screening exams, whereas MASAI only included 2D digital mammography. 

  • While 2D mammography is the norm in Europe, much of the U.S. has switched to DBT for breast screening, so the AITIC results offer good news for U.S. breast imaging practices considering AI adoption.

The Takeaway

The AITIC study’s new results are powerful confirmation of findings from the recent MASAI trial and support broader clinical deployment of mammography AI. Taken together with positive findings from last week’s Nature Cancer articles (see The Wire section in this newsletter), they paint a picture of a technology that’s ready for prime time.

More Positive News on Mammo AI from MASAI

Posted on by Brian Casey

The latest results from the landmark MASAI study of AI for mammography screening show a favorable trend toward reducing the rate of interval cancers, or breast cancers that appear between screening rounds. The new findings – published Friday in The Lancet – also confirm mammography AI’s sharp workload reduction and trend toward higher sensitivity. 

MASAI is a large randomized controlled trial conducted in Sweden that examined the impact of ScreenPoint Medical’s Transpara AI algorithm on breast screening.

  • It’s an important issue, because mammography is one of the radiology segments where AI can provide the most help by reducing radiologist workload while improving cancer detection.

Previous MASAI studies demonstrated that AI can reduce radiologist workload by 44% and improve cancer detection rates by 28%.

  • The findings suggest that AI could eliminate the need for double-reading of most mammograms, a practice that’s common in European screening programs.

The new findings focus specifically on interval cancers, cancers that are missed in one screening round, only to be found later. 

  • Like other MASAI studies, the patient population consisted of 106k women screened with mammography and Transpara AI in Sweden’s national program in 2021 and 2022. 

Results indicated that AI-aided mammography…

  • Cut interval cancer rates by 12% per 1k women (1.55 vs. 1.76).
  • Reduced invasive interval cancers by 16% (75 vs. 89) with 27% fewer cancers of aggressive subtypes (43 vs. 59).
  • Detected 9% more cancers at screening (81% vs. 74%) with comparable specificity (99% for both) and recall rates (1.5% vs. 1.4%).

The researchers acknowledged that the study was not powered to show a statistically significant difference in the interval cancer rate. 

  • But its positive trend indicates that AI can be used to replace double-reading without negative consequences for patients – resulting in a sharp workload reduction for radiologists. 

The Takeaway

Results from the MASAI study on mammography AI just keep on getting better. Last week’s findings indicate that there’s really no reason for European breast screening programs to not dive in and replace their second readers with AI for the majority of exams.

Ensemble Mammo AI Combines Competing Algorithms

Posted on by Brian Casey

If one AI algorithm works great for breast cancer screening, would two be even better? That’s the question addressed by a new study that combined two commercially available AI algorithms and applied them in different configurations to help radiologists interpret mammograms.

Mammography AI is emerging as one of the primary use cases for medical AI, understandable given that breast imaging specialists have to sort through thousands of normal cases to find one cancer. 

Most of these studies applied a single AI algorithm to mammograms, but multiple algorithms are available, so why not see how they work together? 

  • This kind of ensemble approach has already been tried with AI for prostate MRI scans – for example in the PI-CAI challenge – but South Korean researchers writing in European Radiology believed it would be a novel approach for mammography.

So they combined two commercially available algorithms – Lunit’s Insight MMG and ScreenPoint Medical’s Transpara – and used them to analyze 3k screening and diagnostic mammograms.

  • Not only did the authors combine competing algorithms, but they adjusted the ensemble’s output to emphasize five different screening parameters, such as sensitivity and specificity, or by having the algorithms assess cases in different sequences.

The authors assessed ensemble AI’s accuracy and ability to reduce workload by triaging cases that didn’t need radiologist review, finding…

  • Outperformed single-algorithm AI’s sensitivity in Sensitive Mode (84% vs. 81%-82%) with an 18% radiologist workload reduction.
  • Outperformed single-algorithm AI’s specificity in Specific Mode (88% vs. 84%-85%) with a 42% workload reduction.
  • Had 82% sensitivity in Conservative Mode but only reduced workload by 9.8%.
  • Saw little difference in sensitivity based on which algorithm read mammograms first (80.3% and 80.8%), but both approaches reduced workload 50%.

The authors suggested that if applied in routine clinical use, ensemble AI could be tailored based on each breast imaging practice’s preferences and where they felt they needed the most help.

The Takeaway

The new results offer an intriguing application of the ensemble AI strategy to mammography screening. Given the plethora of breast AI algorithms available and the rise of platform AI companies that put dozens of solutions at clinicians’ fingertips, it’s not hard to see this approach being put into clinical practice soon.

Mammo Risk Prediction Improves with AI

Posted on by Brian Casey

Artificial intelligence is beginning to show that it can not only detect breast cancer on mammograms, but it can predict a patient’s future risk of cancer. A new study in JAMA Network Open showed that a U.S. university’s homegrown AI algorithm worked well in predicting breast cancer risk across diverse ethnic groups. 

Breast cancer screening traditionally has used a one-size-fits-all model based on age for determining who gets mammography.

  • But screening might be better tailored to a woman’s risk, which can be calculated from various clinical factors like breast density and family history.

At the same time, research into mammography AI has uncovered an interesting phenomenon – AI algorithms can predict whether a woman will develop breast cancer later in life even if her current mammograms are normal. 

The new study involves a risk prediction algorithm developed at Washington University School of Medicine in St. Louis that uses AI to analyze subtle differences and changes in mammograms over time, including texture, calcification, and breast asymmetry.

  • The algorithm then generates a mammogram risk score that can indicate the risk of developing a new tumor.

In clinical trials in British Columbia, the algorithm was used to analyze full-field digital mammograms of 206.9k women aged 40-74, with up to four years of prior mammograms available. Results were as follows …

  • The algorithm had an AUROC of 0.78 for predicting cancer over the next five years.
  • Performance was higher for women older than 50 compared to 40-50 (AUROC of 0.80 vs. 0.76).
  • Performance was consistent across women of different races.
  • 9% of women had a five-year risk higher than 3%. 

The algorithm’s inclusion of multiple mammography screening rounds is a major advantage over algorithms that use a single mammogram as it can capture changes in the breast over time. 

  • The model also showed consistent performance across ethnic groups, a problem that has befallen other risk prediction algorithms trained mostly on data from White women. 

The Takeaway

The new study advances the field of breast cancer risk prediction with a powerful new approach that supports the concept of more tailored screening. This could make mammography even more effective than the one-size-fits-all approach used for decades.

AI Boosts DBT in Detecting More Breast Cancer

Posted on by Brian Casey

A real-world study of AI for DBT screening found that AI-assisted mammogram interpretation nearly doubled the breast cancer detection rate. Radiologists using iCAD’s ProFound AI software saw sharp improvements across multiple metrics. 

Mammography screening has quickly become one of the most promising use cases for AI. 

  • Multiple large-scale studies published in 2024 and 2025 have documented improved radiologist performance when using AI for mammogram interpretation, with the largest studies performed in Europe.

Another new technology changing mammography screening is digital breast tomosynthesis, which is being rapidly adopted in the U.S. 

  • DBT use in Europe is occurring more slowly, so questions have arisen about whether AI’s benefits for 2D mammography would also be found with 3D systems.

To investigate this question, researchers writing in Clinical Breast Cancer tested radiologist performance for DBT screening before and after implementation of iCAD’s ProFound V2.1 AI algorithm in 2020 at Indiana University. 

  • Interestingly, the pre-AI period included use of iCAD’s older PowerLook CAD software. 

Across the 16.7k DBT cases studied, those with AI saw …

  • A sharp improvement in cancer detection rate per 1k exams (6.1 vs. 3.7).
  • A decline in the abnormal interpretation rate (6.5% vs. 8.2%).
  • Higher PPV1 (rate that abnormal mammograms would be positive) (8.8% vs. 4.2%).
  • Higher PPV3 (rate that biopsies would be positive) (57% vs. 32%). 
  • Higher specificity (94% vs. 92%).
  • No statistically significant change in sensitivity.

The findings on sensitivity are curious given AI’s positive impact on other interpretation metrics.

  • Researchers postulated that there was higher breast cancer incidence in the post-AI implementation period, which could have been caused by AI finding cancers that were missed in the period without AI.

The Takeaway

The radiology world has seen multiple positive studies on AI for mammography, but most of these have come from Europe and involved 2D mammography not DBT. The new results suggest that AI’s benefits will also transfer to DBT, the technology that’s becoming the standard of care for breast screening in the U.S.

How Do Patients Feel about Mammo AI?

Posted on by Brian Casey

As radiology moves (albeit slowly) to adopt clinical AI, how do patients feel about having their images interpreted by a computer? Researchers in a new study in JACR queried patients about their attitudes regarding mammography AI, finding that for the most part the jury is still out. 

Researchers got responses to a 36-question survey from 3.5k patients presenting for breast imaging at eight U.S. practices from 2023-2024, finding …

  • The most common response to four questions on general perceptions of medical AI was “neutral,” with a range of 43-51%. 
  • When asked if using AI for medical tasks was a bad idea, more patients disagreed than agreed (28% vs. 25%). 
  • Regarding confidence that medical AI was safe, patients were more dubious, with higher levels of disagreement (27% vs. 20%).
  • When asked if medical AI was helpful, 43% were neutral but positive attitudes were higher (35% vs. 19%).

The Takeaway

Much like clinicians, patients seem to be taking a wait-and-see attitude toward mammography AI. The new survey does reveal fault lines – like privacy and equitability – that AI developers would do well to address as they work to win broader acceptance for their technology. 

We’re testing a new format today – let us know if you prefer two shorter Top Stories or one longer Top Story with this quick survey!

Patients Want Mammo AI, But Mostly As Backup

Posted on by Brian Casey

Patients support the idea of having AI review their screening mammograms – under certain conditions. That’s according to a new study in Radiology: Imaging Cancer that could have implications for breast imagers seeking to integrate AI into their practices.

Mammography screening has been identified as one of the most promising use cases for AI, but clinical adoption has been sluggish for reasons that range from low reimbursement to concerns about data privacy, security, algorithm bias, and transparency. 

  • Vendors and providers are working on solving many of the problems impeding greater AI use, but patient preference is an often overlooked factor – even as some providers are beginning to offer AI review services for which patients pay out of pocket.

To gain more insight into what patients want, researchers from the University of Texas Southwestern Medical Center surveyed 518 women who got screening mammography over eight months in 2023, finding …

  • 71% preferred that AI be used as a second reader along with a radiologist.
  • Only 4.4% accepted standalone AI interpretation of their images.
  • 74% wanted patient consent before AI review.
  • If AI found an abnormality, 89% wanted a radiologist to review their case, versus 51% who wanted AI to review abnormal findings by radiologists.
  • If AI missed a finding, 58% believed “everyone” should be accountable, while 15% said they would hold the AI manufacturer accountable. 

Patient preference for use of AI in collaboration with radiologists tracks with other recent research. 

  • Patients seem to want humans to retain oversight of AI, and seem to value trust, empathy, and accountability in healthcare – values associated with providers, not machines. 

The findings should also be good news for imaging services companies offering out-of-pocket AI review services. 

The Takeaway

The new findings should be encouraging not only for breast imagers and AI developers, but also for the imaging services companies that are banking on patients to shell out their own money for AI review. As insurance reimbursement for AI languishes, this may be the only way to move mammography AI forward in the short term.

Get every issue of The Imaging Wire, delivered right to your inbox.