A breakthrough in AI could save thousands of breast cancer patients from surgery

Artificial intelligence can assess breast cancer better than doctors and could save thousands of women from unnecessary surgery, according to a new study.

Breast cancer is the most common cancer in Britain. About 55,000 women and 400 men are diagnosed annually.

In the United States, where the study was conducted, about one in eight women will receive a diagnosis during their lifetime.

During diagnosis, cancers are evaluated by pathologists to determine how abnormal the tissue is.

However, it can be difficult for the human eye to detect non-cancerous cells, which can indicate whether the cancer is growing or disappearing.

But a team of scientists from Northwestern University, Illinois, has developed an AI that can prevent patients from undergoing unnecessary chemotherapy, which can be uncomfortable and harmful.

Authors of the study, published in the journal Nature Medicine, also say AI technology could be invaluable in helping doctors assess cancers and predict individual outcomes.

The assessment that helps determine a patient’s treatment currently focuses solely on the appearance of cancer cells and has remained largely unchanged for decades.

However, many studies of the biology of breast cancer have shown that non-cancerous cells – including those of the immune system and non-cancerous cells that provide shape and structure to tissue – may play an important role in supporting or inhibiting the growth and progression of breast cancer. breast cancer. a cancer.

The Northwestern researchers, led by study author Dr. Lee Cooper, associate professor of pathology at Northwestern University, developed an AI model to assess breast cancer tissue from digital images that assess the appearance of cancerous and non-cancerous cells and the interactions between them. .

“Our study shows how important non-cancer components are to a patient’s outcome,” said Dr. Cooper.

“It is difficult for a pathologist to assess these patterns because they are difficult for the human eye to reliably categorize.”

“The AI ​​model measures these patterns and presents information to the pathologist in a way that makes the AI ​​decision-making process clear.”

The AI ​​model analyzes 26 different characteristics of a patient’s breast tissue to create an overall score and also generates individual scores for cancer, immune and other cells to explain the overall result to the pathologist.

This information can then be used by a patient’s care team in developing an individualized treatment plan.

The new assessment method could also give breast cancer patients a more accurate assessment of the risk associated with their disease and help them make informed decisions about their treatment.

The researchers say it can also help assess response to treatment, which can increase or decrease depending on how the appearance of the tissue changes over time.

For example, the tool could be used to detect the effectiveness of a patient’s immune system in fighting cancer during chemotherapy, which could be used to reduce the duration or intensity of chemotherapy.

To train their AI model, the researchers used hundreds of thousands of digital images of patient tissues by building an international network of medical students and pathologists on multiple continents.

The study was conducted in collaboration with the American Cancer Society (ACS).

The team of Dr. Cooper will then evaluate the model to validate it for future clinical use.