Researchers in the United States developed the model to grade cancers by looking at the cells of patients and evaluating what kind of treatment they need.
Though this grading is currently undertaken by pathologists, they are less able to pick up on non-cancerous cells which can indicate whether the cancer will grow or recede.
The scientists, from Northwestern University, in Illinois, say the robotic tool could prevent patients from undergoing unnecessary chemotherapy treatment, which can be unpleasant and harmful.
The authors behind the study, published in the journal Nature Medicine, also say the AI technology could be invaluable in helping doctors to assess cancers and predicting individual outcomes.
Breast cancer is so wide-reaching that around one in eight women across the US will receive a diagnosis in their lifetimes.
It’s also the most common cancer in the UK, with around 55,000 women and 400 men diagnosed with breast cancer every year.
During diagnosis, a pathologist reviews a patient’s cancerous tissue to determine how abnormal the tissue is in a process called grading.
Grading, which helps to determine what treatment a patient will receive, currently focuses solely on the appearance of cancer cells and has remained largely unchanged for decades.
However, many studies of breast cancer biology have shown that non-cancerous cells – including those from the immune system and non-cancerous cells that provide form and structure to the tissue – can play an important role in sustaining or inhibiting the growth and progress of a cancer.
The Northwestern researchers, headed by study author Dr. Lee Cooper, an associate professor of pathology at Northwestern University Feinberg School of Medicine, built an AI model to evaluate breast cancer tissue from digital images that evaluate the appearance both of cancerous and non-cancerous cells and the interactions between them.
“Our study demonstrates the importance of non-cancer components in determining a patient’s outcome,” Dr. Cooper explained.
“The importance of these elements was known from biological studies, but this knowledge has not been effectively translated to clinical use.
“These patterns are challenging for a pathologist to evaluate as they can be difficult for the human eye to categorize reliably.
“The AI model measures these patterns and presents information to the pathologist in a way that makes the AI decision-making process clear to the pathologist.”
The AI model analyses 26 different properties of a patient’s breast tissue to generate an overall prognostic score, as well as generating individual scores for cancer, immune and stromal cells to explain this overall score to the pathologist.
In some patients, a favorable score may be due to the properties of their immune cells. Whereas, for others, it could be down to the properties of their cancer cells.
This information could subsequently be used by a patient’s care team in creating an individualized treatment plan.
The new evaluation method could also provide breast cancer patients with a more accurate estimate of the risk associated with their disease; empowering them to make informed decisions about their clinical care.
The researchers say it could also help to assess therapeutic response treatments, allowing it to be escalated or de-escalated depending on how the appearance of the tissue changes over time.
For instance, the tool could be used to recognize the effectiveness of a patient’s immune system in targeting their cancer during chemotherapy, which could be used to reduce the duration or intensity of chemotherapy.
“We also hope that this model could reduce disparities for patients who are diagnosed in community settings,” Dr. Cooper added.
“These patients may not have access to a pathologist who specializes in breast cancer, and our AI model could help a generalist pathologist when evaluating breast cancers.”
The study was conducted in collaboration with the American Cancer Society (ACS), which created a unique dataset of breast cancer patients from more than 423 US counties through their Cancer Prevention Studies.
In this collaboration, Northwestern developed the AI software while scientists at the ACS and National Cancer Institute provided expertise on breast cancer epidemiology and clinical outcomes.
To train their AI model, the researchers used hundreds of thousands of human-generated annotations of cells and tissue structures within digital images of patient tissues by creating an international network of medical students and pathologists across several continents.
These volunteers provided this data through a website over the course of several years to make it possible for the AI model to reliably interpret images of breast cancer tissue.
Next, Dr. Cooper’s team will evaluate the model to validate it for future clinical use.
The researchers are additionally working to develop models for more specific types of breast cancers, such as triple-negative or HER2-positive.
Invasive breast cancer encompasses several different categories, and the important tissue patterns may vary across these categories.
“This will improve our ability to predict outcomes and will provide further insights into the biology of breast cancers,” Dr. Cooper said.
The post “Artificial intelligence outperforms doctors in assessing breast cancer” appeared first on Zenger.
Produced in association with SWNS Talker
“What’s the latest with Florida Man?”
Get news, handpicked just for you, in your box.