Artificial intelligence (AI) imaging and volumetric breast density algorithms can help predict long-term risk of breast cancer, according to researchers from Mayo Clinic and the University of California in San Francisco.

The study, published in the Journal of Clinical Oncology, was designed to determine if an AI algorithm could contribute to long-term risk prediction for advanced and interval cancers. Researchers used images from 2412 women with invasive breast cancer and 4995 matched controls who had digital mammograms performed two to five years before cancer diagnosis.

The performance of conventional assessment measures – traditional radiologist interpretation and BI-Rads density categories – were compared with the ability of newer AI-powered techniques such as precise volumetric density measures and image-based malignancy risk scores for long-term risk prediction of advanced and interval cancers.

They found the ScreenPoint Medical Transpara AI score improves long-term risk prediction when combined with clinical risk factors including breast density for overall invasive cancers, screen-detected, advanced and nonadvanced cancers. For interval cancers, the Volpara TruDensity AI algorithm measures remained of greatest importance for discrimination, even years before the cancer.

The Transpara exam score is an image-based risk tool that categorises exams using a ten-point scale and is used for concurrent reading of mammograms. The higher the score, the higher the risk of cancer in the mammogram. Clinical research conducted with Transpara shows that scores between one and seven (low risk) have a 99.97% negative predictive value. The TruDensity algorithm uses a combination of x-ray physics and machine learning to generate an accurate volumetric measure of breast composition to eliminate variability that can arise from human interpretation.

“While we have known for decades that density and breast cancer risk are correlated, recent research has really pushed forward our ability to better understand the effects of density combined with image-based risk to drive personalised medicine for women,” said Nico Karssemeijer, chief scientific officer at Screenpoint Medical, and a faculty member at Radboud University.

Used globally, Transpara has been used to analyse more than four million mammograms. Research shows that up to 45% of interval cancers can be found earlier using Transpara, while helping reduce workload and optimise workflow.

“Breast density is a critical factor in assessing breast cancer risk, and an objective, volumetric measurement of density is pivotal,” said Ralph Highnam, chief science officer at Volpara Health. “Through the power of AI, we can uncover valuable insights that help clinicians identify individuals at risk for cancer and tailor personalised screening and prevention strategies.”

Volpara’s software is used to assess breast density for more than six million women annually. TruDensity is proven to reduce reader variability.

“Early identification of image-based risk can give a more effective screening and care pathway for high-risk women,” said Karssemeijer. “This approach should not only save lives but enable women to experience the least disruption in their lives when care is required. Screening today requires this type of focused, personalised approach. We hope that the work we are doing empowers women and their providers to deliver the right care in the right way at the right times.”

ScreenPoint Medical translates edge machine-learning research into technology accessible by radiologists to improve screening workflow, decision confidence and breast cancer risk assessment. Transpara is used by radiologists globally because it has been developed by experts in machine learning and image analysis and updated with user feedback from breast imagers.

New Zealand-based Volpara Health makes software to save families from cancer. Healthcare providers use Volpara to understand cancer risk, empower patients in personal care decisions, and guide recommendations about additional imaging, genetic testing and other interventions. The AI-powered image analysis enables radiologists to quantify breast tissue with precision and helps technologists produce mammograms with optimal image quality, positioning, compression and dose.