Page – Welspun Health

WellSpan uses artificial intelligence (AI) to enhance the evaluation of screening mammograms, helping radiologists quickly and accurately analyze images to detect breast cancer.

“This improves accuracy,” said Dr. Heidi Baylis, vice president of diagnostic services and chief medical officer. “It's really a second set of eyes. We have doctors and AI solutions. The AI ​​might highlight the chest area and say, 'Don't miss this.' Within minutes, you can tap the radiologist on the shoulder and point out areas of concern. ”

One in eight women in the United States will develop breast cancer. The earlier cancer is detected, the faster treatment can be provided and the better the chance of survival.

AI training

AI engineers have developed software that looks for patterns in the chest that may indicate areas of possible cancer.

The AI ​​solutions used at WellSpan are trained on more than 1 million mammograms from clinical sources around the world. This tool constantly receives new data and improves its functionality more and more.

This is like having an additional doctor examine your mammogram, and is clinically proven to improve the accuracy of interpretation by radiologists.

AI action

Training-based AI mammography solution:

• Detect areas that the radiologist thinks should be examined more closely and mark them with a diamond or circle.
• Assigns an overall score from 1 to 10 to the mammography region of interest. A higher score indicates an important area of ​​concern that requires further evaluation and verification by a physician.

Although technology can provide assistance, the physician remains the basis of the assessment.

“Any areas with signs of concern should always be reviewed by a physician,” says Dr. Bayliss. “AI doesn't read mammograms. Artificial intelligence solutions are just solutions in a toolkit that helps doctors escalate areas of concern for quick evaluation. If they agree that it is a genuine concern, we will proceed to the next step of assessment with the patient.”

AI immediacy

The system is so fast that it can provide results in just a few minutes and quickly alert radiologists to areas of potential abnormality, Dr. Bayliss said. In some cases, the radiologist may inform you that further testing is required before the patient leaves the imaging area. Patients can receive this treatment as soon as possible, often within the same day.

“We don't want you to have days of anxiety until you come back,” she says. “We want to address problems quickly and provide a one-stop solution for diagnosis, whether benign or cancerous. We can detect cancer early and make it completely curable. We want to treat it when it happens. It's simple, but we want to save lives.”

AI accuracy

WellSpan's testing of the AI ​​tool found more patients with areas of concern that required further testing. Previously, mammograms showed areas of concern in about 5.5 out of 1,000 patients. But with AI, the hope is to increase that number to 8 out of 1,000.

Research has confirmed that it facilitates the discovery of potential concerns. A study in Sweden found that reading mammograms with the help of AI detects 20% more cancers than routine double readings by two different radiologists.

The Swedish study also found another important finding. The use of AI did not increase false positives, meaning that mammograms that appeared to yield concerning results did not detect cancer in the breasts. Similarly, WellSpan found that using this tool reduced patient callback rates for areas of concern.

“We want to avoid calling back people who don't need to be called back,” says Dr. Bayliss. “We don't want to bring people back or cause unnecessary anxiety or upset.”

This AI tool can also help read mammograms with dense breasts, which is determined by the tissue composition of the breast. Women with more glandular and fibrous connective tissue have denser breasts. Due to their structure, dense breasts are difficult to see through on mammograms. Additionally, women with very dense breasts are four to five times more likely to develop breast cancer than women with fatty breasts.

WellSpan already uses 3D mammography to help evaluate women with dense breasts. AI tools further enhance the ability to accurately read these mammograms.

Next steps for AI

One of the next AI enhancements WellSpan will offer for women's breast health later this year is the High Risk Breast Cancer Assessment Questionnaire. Although mammography is recommended for women over 40, women at high risk for breast cancer should start getting tested as young as 25.

The key is to know when your background and personal health history puts you in that high-risk group. This is important for patients and their families.

“If your mother, aunt, or other family member has a history of cancer, or if you have other risk factors, your doctor can guide you through this questionnaire,” Dr. Bayliss says. “We want to identify these very young people who are at risk at a stage where they can be cured.”

An extensive questionnaire administered by doctors uses information from other women collected in a large database to predict a patient's breast cancer risk with the help of AI.

To learn more about mammography or to schedule a screening visit: here.

Dr. Heidi Bayliss, Vice President and Chief Medical Officer of Diagnostic Services