Artificial intelligence diagnoses breast cancer, a world first

“I was impressed with the results of the study, the very high levels of precision and the breadth of detection capabilities offered by Ibex’s AI technology, similar to those of expert pathologists”. It is in these terms that Stuart Schnitt – physician, chief of the department of oncological breast pathology at the DanaFarber/Brigham and Women’s Cancer Center and professor of pathology at Harvard Medical School – speaks of the breast cancer diagnosis made by artificial intelligence.

This world first is the result of a collaboration between the Institut Curie and Ibex Medical Analyltics. AI, renamed Galen Breast, was developed to reduce errors and improve the quality of breast cancer diagnosis. For the first time, the work of this technology has been validated in clinical use.

First validated clinical results

artificial intelligence at the origin of this first diagnosis of breast cancer acquired his knowledge thanks to the deep learning (deep learning methods). To be able to identify over 50 specific breast features, she had to sift through over two million image samples (including random breast cases and samples with uncommon features), all controlled. by 18 specialists.

Currently, the algorithm can only diagnose slide biopsy specimens. But that doesn’t stop Galen Breast from being very relevant. The results published in the study demonstrate that AI was able to accurately distinguish between different types of breast cancer but also to establish the grade of early cancers.

In the future, the work of Institut Curie and Ibex Medical Analytics will aim to train AI to detect breast pathologies directly on surgical samples.

Help, verify, automate

What can replace doctors specializing in this pathology? This is not at all the vocation of Galen Breast. “The system is mainly developed to be a support for pathologists, to check that they have not missed anything when reading it (diagnosis)” explain to Science and Future Suzette Delaloge, medical oncologist specializing in breast cancer and program director Interception by Gustave Roussy.

Today, breast cancer affects 2.2 million people each year. The issue of diagnosis as fast as precise is at the heart of the concerns, the reduction in the number of pathologists and the increase in the incidence of this cancer making the task more difficult. An AI like Galen Breast could therefore contribute to improving diagnosis but also better care.

In the future “AI could give an indication of treatment that can be coupled with surgery, thanks to the determination of hormone receptors or proliferation indexes” explain to Science and Future Marc Bollet, oncologist and radiotherapist at the French Breast Institute.

We would love to say thanks to the author of this write-up for this remarkable content

Artificial intelligence diagnoses breast cancer, a world first

We have our social media pages here and other pages on related topics here.