Femtech World explores the latest research and science developments in the world of women’s health.

AI models independently interpret mammograms

AI models have shown excellent performance for detecting breast cancers on mammography images.

The algorithms, submitted for a 2023 AI Challenge hosted by the Radiological Society of North America (RSNA), demonstrated increased screening sensitivity while maintaining low recall rates, according to a new study.

The goal of the Challenge was to source AI models that improve the automation of cancer detection in screening mammograms, helping radiologists work more efficiently, improving the quality and safety of patient care, and potentially reducing costs and unnecessary medical procedures.

A research team evaluated 1,537 working algorithms submitted to the Challenge, testing them on a set of 10,830 single-breast exams – completely separate from the training dataset – that were confirmed by pathology results as positive or negative for cancer.

The algorithms yielded median rates of 98.7 per cent specificity for confirming no cancer was present on mammography images, 27.6 per cent sensitivity for positively identifying cancer, and a recall rate, the percentage of the cases that AI judged positive, of 1.7 per cenr.

When the researchers combined the top three and top 10 performing algorithms, it boosted sensitivity to 60.7 per cent and 67.8 per cent, respectively.

According to the researchers, creating an ensemble of the 10 best-performing algorithms produced performance that is close to that of an average screening radiologist in Europe or Australia.

Individual algorithms showed significant differences in performance depending on factors such as the type of cancer, the manufacturer of the imaging equipment and the clinical site where the images were acquired.

Overall, the algorithms had greater sensitivity for detecting invasive cancers than for noninvasive cancers.

Since many of the participants’ AI models are open source, the results of the Challenge may contribute to the further improvement of both experimental and commercial AI tools for mammography, with the goal of improving breast cancer outcomes worldwide.

The research team plans to conduct follow-up studies to benchmark the performance of the top Challenge algorithms against commercially available products using a larger and more diverse dataset.

Cracking the cold case of endometriosis with big data

Records from millions of patients at UC health centers found correlations between endometriosis, one of the most common diseases in women, and a bounty of other diseases.

Scientists at UCSF have found that endometriosis often occurs alongside conditions like cancer, Crohn’s disease, and migraine.

The research could improve how endometriosis is diagnosed and, ultimately, how it is treated; and it paints the sharpest portrait yet of a condition that is as mysterious as it is prevalent.

The study used computational methods developed at UCSF to analyse anonymised patient records collected at the University of California’s six health centers.

Using algorithms developed for the task, researchers hunted for connections linking endometriosis with the rest of each patient’s health history.

Endometriosis patients were compared with patients who did not have it, and categorised the patients with endo into groups based on shared health histories.

The findings from the UCSF data were mapped against the rest of the UC’s health data to see if they held up across California.

The team say they found over 600 correlations between endometriosis and other conditions, ranging from infertility, autoimmune disease, and acid-reflux, to cancers, asthma, and eye-related diseases.

Some patients had migraines, bolstering previous studies suggesting that migraine drugs might help treat endometriosis.

The study supports the growing understanding of endometriosis as a “multi-system” disorder – a disease arising from dysfunction throughout the body.

Respiratory viruses can wake up breast cancer cells in lungs

Researchers have found the first direct evidence that common respiratory infections, including Covid-19 and influenza, can awaken dormant breast cancer cells that have spread to the lungs, setting the stage for new metastatic tumours.

The findings, obtained in mice, were supported by research showing increases in death and in metastatic lung disease among cancer survivors infected with SARS-CoV-2, the virus that causes Covid-19.

“Our findings indicate that individuals with a history of cancer may benefit from taking precautions against respiratory viruses, such as vaccination when available, and discussing any concerns with their healthcare providers,” said Julio Aguirre-Ghiso, a co-leader of the study and director of MECCC’s Cancer Dormancy Institute.

Prior to the study, some evidence suggested that inflammatory processes can awaken disseminated cancer cells (DCCs) – cells that have broken away from a primary tumor and spread to distant organs, often lying dormant for extended periods.

“During the COVID-19 pandemic, anecdotal reports suggested a possible increase in cancer death rates, bolstering the idea that severe inflammation might contribute to arousing dormant DCCs,” said Dr. Aguirre-Ghiso, who also serves as leader of MECCC’s Tumor Microenvironment and Metastasis Research Programme.

Researchers tested this hypothesis using Dr. Aguirre-Ghiso’s laboratory’s unique mouse models of metastatic breast cancer, which include dormant DCCs in the lungs and therefore closely resemble a key feature of the disease in humans.

The researchers exposed mice to SARS-CoV-2 or influenza virus. In both cases, the respiratory infections triggered the awakening of dormant DCCs in the lungs, leading to a massive expansion of metastatic cells within days of infection and the appearance of metastatic lesions within two weeks.

Molecular analyses revealed that the awakening of dormant DCCs is driven by interleukin-6 (IL-6), a protein that immune cells release in response to infections or injuries.

The Covid-19 pandemic offered a unique opportunity to investigate the effect of respiratory virus infections, in this case from the SARS-CoV-2 virus, on cancer progression.

The research team analysed two large databases and found support for their hypothesis that respiratory infections in cancer patients in remission are linked to cancer metastasis.

The UK Biobank is a general population cohort in which some of the more than 500,000 participants were diagnosed with cancer and other diseases prior to the Covid-19 pandemic.

Researchers from Utrecht University and Imperial College London investigated whether a Covid-19 infection increased the risk of cancer-related mortality among participants with cancer.

They focused on cancer survivors who had been diagnosed at least five years before the pandemic, ensuring they were likely in remission.

Among them, 487 individuals tested positive for COVID-19 and these were compared to 4,350 matched controls who tested negative.

After excluding those cancer patients who died from Covid-19, the researchers found that cancer patients who tested positive for Covid-19 faced an almost doubling of risk of dying from cancer compared to those patients with cancer who had tested negative.

From the second population study, the U.S. Flatiron Health database, researchers drew data pertaining to female breast cancer patients seen at 280 U.S. cancer clinics.

They compared the incidence of metastases to the lung among Covid-19-negative patients and Covid-19-positive patients (36,216 and 532 patients respectively).

During the follow-up period of approximately 52 months, those patients who came down with Covid-19 were almost 50 per cent more likely to experience metastatic progression to the lungs compared with patients with breast cancer without a diagnosis of Covid-19.

“Our findings suggest that cancer survivors may be at increased risk of metastatic relapse after common respiratory viral infections,” said Dr. Vermeulen.

Losing weight before IVF may increase chance of pregnancy

A systematic review and meta-analysis of randomised controlled trials (RCTs) has assessed whether weight loss interventions before in vitro fertilization (IVF) improved reproductive outcomes.

The review found that weight loss interventions before IVF could increase the chances of pregnancy, especially in unassisted conception, although the effect on live births was unclear.

The findings are published in Annals of Internal Medicine.

Researchers from the University of Oxford reviewed 12 RCTs comprising 1,921 patients conducted between 1980 through 27 of May 2025.

Inclusion criteria included studies conducted on women at least 18 years old with a BMI of 27 kg/m2 or greater who were seeking IVF with or without intracytoplasmic sperm injection treatment for infertility.

Outcomes of interest were the number of participants achieving pregnancy without IVF (unassisted pregnancy), with IVF (treatment-induced pregnancy), overall (unassisted plus treatment-induced) and those delivering a live infant.

The researchers found that participants were typically women in their early 30s with a median baseline BMI of 33.6 kg/m2.

Weight loss interventions studied included low-energy diets, an exercise program accompanied by healthy eating advice, and pharmacotherapy accompanied by diet and physical activity advice.

Overall, weight loss interventions before IVF were associated with greater unassisted pregnancy rates. Evidence was inconclusive on the effect of weight loss interventions on treatment-induced pregnancies.

Evidence on the association between weight loss interventions before IVF and live births was uncertain, although there was moderate certainty of no association with pregnancy loss.

The findings suggest that weight loss interventions before IVF increase total pregnancies, mainly through an increase in unassisted pregnancy rates.

However, further high-quality clinical trials testing different weight loss interventions, particularly those known to achieve greatest weight losses, such as low-energy total diet replacement programmes, are needed.