Skip to main content

Ask an Expert: Artificial Intelligence and Breast Cancer

Read Time: 3 minutes

Matthew Covington, MD, featured expert in cancer imaging.
Matthew Covington, MD, featured expert in cancer imaging.

Artificial intelligence, or AI, uses computer science and data to help solve complex problems. In cancer care, using this advanced approach can help doctors diagnose, treat, and understand specific types of cancers.

Matthew Covington, MD, an expert radiologist at Huntsman Cancer Institute and the University of Utah, specializes in breast cancer screening. He explains the possible benefits of AI in health care and how it can be used to better detect and treat breast cancer.

Can AI detect cancer?

Yes. Providers are using AI to detect cancer through computer-aided software. For breast cancer, this detection system points out signs of cancer on mammograms. Radiologists like Dr. Covington then review the highlighted regions. By using two approaches, having both human and automated systems look at mammograms can help improve cancer detection.

As of September 2023, AI is not reading mammograms or other breast images on its own. This ensures that a trained medical provider always reads your scans before delivering results.

How accurate is AI at reading a scan compared to a provider?

Doctors and researchers are developing new systems they hope will be as accurate or reliable as a radiologist in specific situations.

Because AI can review and compare large sets of data, it has the potential to be equally as accurate as, if not more than, a human. It may also speed up the time it takes to receive a diagnosis, leading to faster access to cancer treatment.

Mammography room at Huntsman Cancer Institute
Mammography room at Huntsman Cancer Institute

What are the challenges of using AI in cancer care?

Ongoing research is important as new technology advances. Our Breast and Gynecologic Cancers Center is committed to finding ways to enhance cancer prevention, diagnosis, treatment, and survival rates. While we are actively engaged, we do not yet have a system that is ready to take on a large role in patient care. Thorough research is needed to ensure we are introducing something that is accurate and safe.

Diagnostic errors with AI are rare, but they can happen. AI systems learn from the data they are trained on. If the data contains unequal representation, AI can reflect those biases. Like a human provider, biases can happen when diagnosing someone. It is important to remember that AI is simply a tool, just like a mammogram or an ultrasound, and to be aware of errors so we can partner with the system to correct them.

What advice do you have for someone who may be anxious about AI?

It is normal to feel concerned about AI and how it might impact medical decisions. If you are feeling anxious, know that you are not alone in these thoughts. It’s a good idea to openly talk about your worries with your doctors. Open communication between patients and providers about how AI is going to be used during diagnosis and treatment is important. Knowing that AI is being designed to assist health care providers rather than replace them might also provide assurance.

How will AI improve patient care in the future?

Dr. Covington sees a promising future for AI in patient care, especially for cancer treatment. He hopes to use it as a tool to be a better doctor, find cancer earlier, and stage cancers more accurately. He is enthusiastic about working with a system that may help prevent cancer from occurring, leading to fewer cases of metastatic breast cancer. AI will be a powerful tool to delivering a cancer-free frontier.

Cancer touches all of us.