Connect with us

Diagnosis

AI outperforms breast density for breast cancer risk

Published

on

An artificial intelligence model predicts five-year breast cancer risk more accurately than traditional breast density assessment, new research shows.

The image-only AI model, called Clairity Breast, showed that women in its high-risk group had more than a fourfold higher cancer incidence than those in the average-risk group (5.9 per cent vs 1.3 per cent).

Breast density, which refers to the amount of fibrous and glandular tissue compared to fatty tissue in the breast, has traditionally been used as one indicator of cancer risk. Dense breast tissue can both mask tumours and slightly increase cancer risk.

Researchers from Harvard Medical School and University Hospital RWTH Aachen in Germany tested the FDA-authorised AI model on 245,232 screening mammograms from US and European sites taken between 2011 and 2017.

The AI was trained on 421,499 mammograms from 27 facilities across Europe, South America and the US, learning to identify subtle tissue patterns that predict cancer development within five years. It uses a deep convolutional neural network (a pattern-recognition algorithm) to generate risk probabilities.

Traditional density assessment showed minimal difference in outcomes: 3.2 per cent for dense versus 2.7 per cent for non-dense breasts.

“Over two million women are diagnosed with breast cancer annually, and for most, it comes as a complete shock,” said Dr Constance D. Lehman, professor of radiology at Harvard Medical School. “Only 5 to 10 per cent of breast cancer cases are considered hereditary, and breast density alone is a very weak predictor of risk.”

The AI categorised risk using National Comprehensive Cancer Network thresholds: average (less than 1.7 per cent), intermediate (1.7-3.0 per cent) and high (greater than 3.0 per cent) five-year risk.

“The model is able to detect changes in the breast tissue that the human eye can’t see,” Dr Lehman explained. “This is a job that radiologists just can’t perform. It’s a separate task from detection and diagnosis, and it will open a whole new field of medicine, leveraging the power of AI and untapped information in the image.”

The findings have particular significance for younger women. While the American Cancer Society recommends optional annual screening from age 40 for average-risk women, those under 40 represent the fastest-growing group diagnosed with breast cancer and advanced disease.

“An AI image-based risk score can help us identify high-risk women more accurately than traditional methods and determine who may need screening at an earlier age,” Dr Lehman said. “We already screen some women in their 30s when they are clearly at high risk based on family history or genetics. In the future, a baseline mammogram at 30 could allow women with a high image-based risk score to join that earlier, more effective screening pathway.”

Dr Christiane Kuhl, director of the Department of Diagnostic and Interventional Radiology at University Hospital RWTH Aachen, who presented the findings, emphasised the clinical implications.

“The results of this large-scale analysis demonstrate that AI risk models provide far stronger and more precise risk stratification for five-year cancer prediction than breast density alone,” she said. “Our findings support the use of image-only AI as a complement to traditional markers supporting a more personalised approach to screening.”

Currently, 32 US states have breast density legislation requiring healthcare providers to inform women of their density status after screening mammograms. The researchers suggest this information could be enhanced with AI risk scores.

“We’d like to see women given information on their breast density and their AI image-based risk score,” Dr Lehman said. “We can do better than just looking at a mammogram and saying, ‘It is dense or not dense’ to inform women of their risk.”

The technology represents what its developers describe as a shift in breast cancer screening from population-based to personalised risk assessment, potentially enabling earlier intervention for high-risk women while reducing unnecessary procedures for those at lower risk.

Diagnosis

CEM shows promise for screening high-risk breast cancer patients

Published

on

Contrast-enhanced mammography (CEM) has shown promise for women at higher risk of breast cancer, with follow-up screening showing greater specificity and accuracy.

Cancer detection rates remained consistent during repeat screening, while specificity improved compared with baseline examinations.

Specificity shows how accurately a test identifies people who do not have the disease, helping to reduce unnecessary follow-up procedures.

Researchers at Memorial Sloan Kettering Cancer Center analysed 6,911 contrast-enhanced mammography screens carried out among 2,756 women between 2015 and 2021.

Contrast-enhanced mammography, or CEM, combines standard mammography with an injected contrast agent that highlights areas of increased blood flow. Cancerous tumours often develop a greater blood supply.

The team compared 1,575 baseline screens with 3,336 incidence screens.

The researchers classified a screen as baseline if the woman had no previous CEM or had not undergone breast MRI in the previous three years.

Incidence screens were follow-up examinations carried out after previous screening.

After adjusting for the number of screens each woman received, the researchers found no statistically significant difference in cancer detection rates between the groups.

However, specificity reached 91.5 per cent during incidence screening, compared with 83.4 per cent for baseline screening.

Overall accuracy was also higher for incidence screening, at 91.4 per cent compared with 83.5 per cent.

“The ability of contrast-enhanced mammography to help detect cancer during prevalence screening in women at increased risk for breast cancer is maintained in subsequent incidence screens, with better specificity and accuracy,” the research team wrote.

Baseline screening detected 19 cancers per 1,000 examinations, compared with 11.6 per 1,000 incidence screens.

Sensitivity, which measures how well a test identifies people who have a disease, was 87.1 per cent for baseline screening and 86.1 per cent for incidence screening.

Contrast enhancement alone helped detect 18 of the 30 cancers found during baseline screening and 36 of the 62 found during incidence screening.

The researchers also identified 14 interval cancers across the examinations, with no evidence of a difference between the two groups.

An interval cancer is diagnosed after a screening result appears normal but before the next scheduled examination.

The retrospective study used previously collected clinical records rather than following participants in a newly designed trial.

The researchers described CEM as a reasonable screening tool for women with dense breasts.

Dense breasts contain more fibrous and glandular tissue, which can make cancer more difficult to detect with standard mammography.

CEM has previously shown greater sensitivity than ultrasound, digital mammography and digital breast tomosynthesis, according to the article.

Digital breast tomosynthesis takes several low-dose X-ray images from different angles to create a three-dimensional view of breast tissue.

The technique also takes less time and generally costs less than breast MRI, with previous research suggesting its performance is not inferior to MRI.

CEM is currently used for some screening purposes outside its approved indications.

The researchers said it could also help address health inequalities affecting women who face barriers to accessing MRI.

“In addition, women have reported a preference for CEM over MRI,” they wrote.

The researchers said a future article would provide full details about the interval cancers identified in the study.

An accompanying editorial said the technique could become a practical part of breast cancer screening for selected groups, although challenges remain around wider adoption.

“Whether this potential ultimately translates into widespread implementation will depend on future studies evaluating not only diagnostic performance, but also patient outcomes, health care utilisation, and real-world feasibility across diverse practice settings,” wrote Dr Vivianne Aguilera Freitas of the University of Toronto.

Continue Reading

Insight

Experimental treatment significantly slows progression of fatal brain disease in women, study finds

Published

on

Davunetide may significantly slow the progression of a fatal brain disease in women, according to a new analysis of clinical trial data.

The findings indicate that women and men with progressive supranuclear palsy (PSP) may respond differently to the experimental treatment.

Progressive supranuclear palsy, or PSP, is a rare and fatal neurodegenerative disease.

Researchers at Tel Aviv University led the analysis and said the results reinforce the need for sex-specific approaches to neurodegenerative diseases.

Neurodegenerative diseases are conditions in which nerve cells in the brain or nervous system gradually lose function and die.

The team reanalysed data from a 52-week international clinical trial involving more than 300 people with PSP.

The disease is caused by the abnormal accumulation of tau protein in the brain. Tau is a protein found in nerve cells that builds up abnormally in people with PSP.

There is currently no effective drug treatment for the disease.

The work was led by professor Illana Gozes of the Sagol School of Neuroscience and the Gray Faculty of Medical and Health Sciences at Tel Aviv University.

The research team included current and former students Dr Guy Shapira, Jason Blatt and Liri Guz, together with professor Noam Shomron.

The original clinical trial found that Davunetide was safe but ineffective.

However, the researchers separated female and male participants and re-examined the data using updated assessment measures recommended by the FDA.

Women treated with Davunetide experienced a significant slowing of disease progression, while no similar effect was observed in men.

The treatment helped preserve essential movement and functional abilities, including balance, fine motor skills and everyday tasks such as using cutlery, buttoning clothes and washing the face and hands.

Fine motor skills are the small, precise movements needed for tasks involving the fingers and hands.

Treated women also showed significant improvements in language ability, working memory and overall cognitive function.

Cognitive function covers mental abilities such as memory, attention, language and problem-solving.

The analysis also identified profound molecular differences between women and men.

The relationship between levels of pathological tau in cerebrospinal fluid and clinical symptoms was completely reversed between the sexes.

Cerebrospinal fluid is the clear liquid surrounding the brain and spinal cord. A biomarker is a measurable sign that can indicate disease activity.

For example, language abilities declined significantly as tau pathology increased in women, but not in men.

The researchers said this suggests the disease mechanisms may work differently in women and men, potentially explaining their different responses to treatment.

According to professor Gozes, overlooking biological differences between the sexes may hide a genuine treatment effect.

“Our data show that analysing women and men separately is not merely a statistical exercise, but an essential tool for developing more effective treatments for neurodegenerative brain diseases,” she said.

The researchers believe the findings provide a strong scientific basis for future clinical trials and treatment protocols designed from the outset to account for patients’ sex.

These trials could evaluate Davunetide as a targeted treatment for women with PSP.

They said the approach may also pave the way for more precise treatments for tau-related diseases, including Alzheimer’s disease and other neurodegenerative brain disorders.

The study was supported by ExoNavis Therapeutics, which is developing Davunetide for brain diseases under licence from Ramot, Tel Aviv University’s technology transfer company.

Continue Reading

Pregnancy

UK research paves way for new preeclampsia therapies

Published

on

A preeclampsia study has found unusual cell activity in mothers and babies that could reveal new targets for treatment.

The condition affects 2 to 4 per cent of pregnancies worldwide and is a leading cause of maternal and foetal mortality.

There is currently no cure, and severe cases can put both the mother and baby at risk.

Scientists from UCL and University College London Hospitals found that stressed placental cells, poorly functioning blood vessels and an overactive immune response all contribute to the condition.

Preeclampsia causes high blood pressure during pregnancy. It can affect blood flow to the baby and cause symptoms such as swelling, headaches, blurred vision and pain under the ribs.

Without treatment, it can damage the mother’s health, slow the baby’s growth and, in severe cases, become life-threatening.

Previous research has focused only on the placenta, the organ that develops during pregnancy to support the baby’s growth, rather than the tissues around it.

The researchers said the findings could reveal new therapeutic targets, which are biological processes that future treatments could be designed to alter.

Senior author professor Sara Hillman, of the UCL EGA Institute for Women’s Health, said: “We studied individual cells from both the mother and the baby to see how their activity changes in healthy pregnancies compared with preeclampsia.

“This helped us to confirm some changes already suspected in the condition and also discover new ones.”

The team studied 20 pregnant women recruited at UCLH, including 10 with severe preeclampsia and 10 without the condition.

They used genomic testing to examine individual cells in the placenta and other tissues where cells from the developing baby and mother come into contact.

Genomic testing examines genetic information to help researchers understand how cells behave and the roles they may play.

The other tissues studied were the myometrium, the muscular layer of the womb, and the chorioamniotic membranes, which surround the baby during pregnancy.

The team compared cells from healthy pregnancies and those affected by preeclampsia at different gestational ages, meaning different stages of pregnancy.

They used technology that can read the genetic information of thousands of individual cells at the same time, allowing them to see what each cell was doing and where it was located in the tissue.

In preeclamptic pregnancies where babies were born prematurely, before 37 weeks, during the third trimester, placental cells showed signs of stress and low oxygen levels.

The cells also did not use energy in the normal way.

Some cells responsible for reshaping the mother’s blood vessels were not working properly, the researchers found, which may affect blood flow to the baby.

There were also signs of an overactive immune response in the placenta, nearby tissues and the mother’s blood.

The researchers said this response, together with other stress molecules released by the placenta, helps explain why preeclampsia affects the whole body and can become serious.

They hope the findings will help researchers find treatments for the condition and potentially save lives.

Co-lead author Dr Yara Sanchez Corrales, of the UCL Great Ormond Street Institute of Child Health, said: “These findings point to specific biological processes that could be targeted with treatments. Acting early in pregnancy, especially in more severe early-onset cases, could help improve outcomes and reduce the high risks associated with severe preeclampsia.

“We hope that our findings may set us on the path to reducing premature births and fatalities associated with preeclampsia.”

Co-lead author Mr Theodoros Xenakis, of the UCL Great Ormond Street Institute of Child Health, said: “Future studies may provide an even clearer picture of the biological changes linked to the disease by including more participants and using even more precise methods.”

Continue Reading

Trending

Copyright © 2025 Aspect Health Media Ltd. All Rights Reserved.