Every year, more than 2.3 million women worldwide are diagnosed with breast cancer, and nearly 700,000 still die from it, according to the World Health Organization1. But behind these shocking figures, a quiet revolution is underway.
As we markPink October 2025, advances in artificial intelligence are reshaping the landscape of breast cancer screening and treatment. Gone are the days of stressful mammograms and delayed diagnoses: a new generation of smart tools promises gentler exams, more accurate results, and more effective prevention.
From the United States to India, via Europe, researchers are coming up with a host of innovations: predictive mammograms, contactless imaging, mobile screening in rural areas, and even works of art created from biomedical data.
Artificial intelligence is no longer just a computational tool, but a compassionate ally. It puts science at the service of compassion, offering women more preventive, personalized, and less invasive medical care.
This special article explores the major breakthroughs of 2025: from the first U.S. clinical trials to mobile apps in India, from MIT’s Mirai predictive model to the Data Bloom art installation presented in Paris, AI is emerging as a new weapon against breast cancer—one that is as powerful as it is human.
A new era in screening: The United States launches the first large-scale clinical trial
In 2025, the United States will host the first large-scale global clinical trial to assess the real-world impact of AI on breast cancer screening2.
This program, led by the Mayo Clinic and funded by the National Institutes of Health (NIH), will involve 80,000 patients across 20 medical centers.
Objective: To compare the performance of artificial intelligence with that of experienced radiologists in a representative sample of women aged 40 to 74.
Researchers hope that AI will make it possible to:
- reduce false negatives by 30%, that is, cancers that go undetected during standard screenings,
- improve early detection by 20%,
- and speed up diagnosis by 25% through the automation of image interpretation.
The study also aims to assess the emotional impact on patients: more reliable detection, fewer unnecessary tests, and clearer communication about individual risks.
“Our goal is to make screening smarter, but also more humane,” explains Professor Lauren O’Neill, the program’s principal investigator. “AI helps us see what the human eye cannot detect, but it does not replace a doctor’s judgment.”
Mirai, the MIT AI that tailors screening to each woman
At the heart of this transformation, the Mirai model developed by the Massachusetts Institute of Technology (MIT) represents a new generation of so-called “predictive” artificial intelligence3.
By combining mammogram analysis, clinical data, and family history, Mirai establishes a personalized risk profile for each patient.
This approach breaks with traditional protocols, under which all women of the same age followed an identical screening schedule. Now, AI adjusts the frequency and type of exams based on the actual risk of developing cancer over the next five years.
The results of the clinical trials are spectacular:
- Mirai has helpedidentify an additional 41% of cancers at an early stage,
- while reducing the number of unnecessary mammograms by 23%.
In France, several teaching hospitals are already testing this approach as part of the European AI4Health program.
“Artificial intelligence is paving the way for more equitable healthcare: it tailors monitoring to each woman’s unique biological profile,” explains researcher Maria Konovalova, a member of the MIT Jameel Clinic.
Toward Pain-Free Screening: The Era of “No Touch, No Radiation”
Technology isn't just about precision—it's also transforming the patient experience. In India, researchers at Apollo Cradle Hospital and the West Bengal Department of Health are testing a contactless, radiation-free, and painless screening method based on AI-driven infrared thermography 4.5
How it works: Thermal cameras detect changes in heat and micro-blood flow in breast tissue. AI then identifies abnormalities that are invisible to the naked eye, sometimes up to 18 months before they become palpable.
The initial results are impressive:
- 89% diagnostic accuracy,
- a 70% reduction in invasive procedures,
- and costs one-tenth as much as a standard mammogram.
These systems are already in use in more than 200 Indian villages, thanks to AI-powered mobile apps used by community health workers.
“AI makes preventive care accessible in areas where medical infrastructure is lacking,” says Dr. Neelima Kanth, a gynecologist in Hyderabad. “It helps save lives without pain or fear.”
Multimodal models: when multiple technologies come together
In Europe, a team of researchers supported by the Horizon Europe program (CORDIS) has developed a diagnostic support system (CAD) that combines several types of medical data:
- X-rays,
- thermal imaging,
- and digital pathology analyses6.
This “multimodal” system enables AI to cross-reference visual and biological signals, providing a more detailed analysis of tumors hidden within dense tissue—a challenge that affects nearly 45% of women.
The results published on arXiv in 2025 are remarkable:
- 42% improvement in the detection of “invisible” tumors,
- 35% reduction in reading errors,
- and a 28% overall improvement in accuracy.
This model, called Density-AI, is currently being tested in 12 European hospitals, including those in Madrid, Zurich, and Berlin. It could become the new European standard by 2027.
When data becomes art: “Data Bloom,” the poetry of research
Because the fight against breast cancer goes beyond the laboratory, artists and researchers are joining forces to transform scientific data into an emotional work of art. Presented in Paris for Pink October, the “Data Bloom” installation brings together data scientists and visual artists around a unique concept: translating biomedical signals into AI-animated digital flowers.
Each virtual flower represents a tumor profile, a protein sequence, or an immune response. The artwork evolves in real time, symbolizing resilience, diversity, and healing.
“We wanted to turn numbers into emotions,” explains Élodie Rousseau, a French data artist and co-founder of the project. “These virtual flowers pay tribute to life, to science, and to all those who are fighting.”
A more predictive, inclusive, and humane approach to medicine
Breast cancer remains the leading cause of cancer-related deaths among women worldwide, but the AI revolution is opening up unprecedented possibilities. According to the WHO, early detection technologies could reduce mortality by 25% by 2030, saving more than 175,000 lives each year.
But beyond the numbers, it is a philosophy of care that is changing: prevention is becoming more human, and technology more compassionate. From MIT’s AI to Indian mobile apps, these innovations are driven by a single goal: to make science an act of universal kindness.
Learn more
To better understand the origins of the technologies that are revolutionizing medical screening today, read the article Generative AI: The Tip of the Iceberg. Published well before the rise of medical AI, it already explained how generative models could transform research, biology, and our understanding of living organisms—a technological foundation upon which the predictive screening innovations presented in this article now rest.
References
1. World Health Organization. (2025). Global Breast Cancer Factsheet.
https://www.who.int
2. Oncology Central. (2025). First Major AI Trial in Breast Cancer Screening Launches in the USA.
https://www.oncology-central.com
3. Axios San Francisco. (2025). MIT’s Mirai AI Personalizes Breast Screening Schedules.
https://www.axios.com
4. The New Indian Express. (2025). Breast Checks Go High-Tech.
https://www.newindianexpress.com
5. Times of India. (2025). AI Protocol App for Quick Breast Cancer Identification.
https://timesofindia.indiatimes.com
6. CORDIS. (2025). AI: The Key to Smarter Patient-Centered Breast Cancer Care.
https://cordis.europa.eu
