Comparing year-long findings with and without adjunctive artificial intelligence (AI) at a breast cancer screening program in Spain, researchers found the combination of digital breast tomosynthesis and AI had a 92.5 percent accuracy for diagnosing cancer in patients with elevated risk.
The use of adjunctive artificial intelligence (AI) doubled the positive predictive value (PPV) of digital mammography (DM) exams overall and led to greater than 90 percent accuracy for DM and digital breast tomosynthesis (DBT) in detecting breast cancer in women with elevated risk, according to research findings presented recently at the European Congress of Radiology (ECR) conference in Vienna, Austria.
For the study, researchers compared the use of adjunctive AI (Transpara® version 1.7.0, ScreenPoint Medical) in 11,988 women (between the ages of 50 and 74) who had DM or DBT screening exams versus 16,555 women screened with DM or DBT the year before without AI support. In the AI group, 5,049 women had DM screening with the Hologic Selenia device and 6,949 women had DBT screening with the Hologic Selenia Dimensions device, according to the study. For the non-AI cohort, 7,229 women had DM screening and 9,326 women had DBT.
The researchers found the combination of AI and DM had an overall cancer detection rate of 8.1/1000 in comparison to 5.1/1000 for DM alone and had an accuracy rate of 95.1 percent for detecting malignancy in patients with an elevated risk for breast cancer. Adjunctive AI also doubled the overall positive predictive value (PPV) for DM (12.2 percent) in comparison to DM alone (6.1 percent), according to the study authors.
For patients with DBT and AI, the researchers reported a cancer detection rate of 9.6/1000 and a PPV of 16.5 percent in comparison to 5.8/1000 and 13.1 percent, respectively, for DBT alone. For the elevated risk group, the study authors said the combination of DBT and AI correctly diagnosed 92.5 percent of breast cancer.
“Artificial intelligence as support for readings in breast cancer screening practice increases cancer detection rate and positive predictive value for both DM and DBT,” noted Esperanza Elias Cabot, MD, who is affiliated with the Breast Cancer Unit in the Department of Radiology at the Hospital Universitario Reina Sofia in Cordoba, Spain.
(Editor’s note: For related content, see “Digital Breast Tomosynthesis and Breast Density: What a New Study Reveals” and “Medicare Mammography Study Shows Black Women Had Less Initial Access to Imaging Advances than White Women.”)
Dr. Cabot also noted a 1.7 percent decrease in the recall rate with adjunctive AI and DM (6.6 percent) in comparison to DM only 8.3 percent). She also pointed out 1.4 percent increase in the recall rate for DBT and AI (5.8 percent) in comparison to DBT alone (4.4 percent).
Of the four lesions missed by AI/DBT but detected by radiologists in the high-risk group, Dr. Cabot said two were invasive carcinomas with architectural distortion and the other two were invasive ductal carcinoma masses. Two lesions missed by AI/DM in the high-risk group included an invasive ductal carcinoma mass and a ductal carcinoma in situ with architectural distortion, according to the study.
Reference
1. Cabot EE. Artificial intelligence (AI) improves breast cancer screening performance for both digital mammography and digital breast tomosynthesis. Presented at the European Congress of Radiology, March 1-5, Vienna, Spain. Available at: https://connect.myesr.org/course/ai-in-breast-imaging-2/#1 . Accessed March 6, 2023.
CT Study: AI Algorithm Comparable to Radiologists in Differentiating Small Renal Masses
May 14th 2024An emerging deep learning algorithm had a lower AUC and sensitivity than urological radiologists for differentiating between small renal masses on computed tomography (CT) scans but had a 21 percent higher sensitivity rate than non-urological radiologists, according to new research.
Current Insights and Emerging Roles for Contrast-Enhanced Mammography
May 10th 2024In a recent lecture at the 2024 ARRS Annual Meeting, Jordana Phillips, MD, discussed the role of contrast-enhanced mammography in staging breast cancer, evaluating response to neoadjuvant chemotherapy and recalls from screening.
Study Finds High Concordance Between AI and Radiologists for Cervical Spine Fractures on CT
May 6th 2024Researchers found a 98.3 percent concordance between attending radiology reports and AI assessments for possible cervical spine fractures on CT, according to new research presented at the 2024 ARRS Annual Meeting.
FDA Clears AI-Powered Qualitative Perfusion Mapping for Cone-Beam CT
May 6th 2024Reportedly validated in more than 10 clinical trials, the AngioFlow perfusion imaging software enables timely identification of brain regions with cerebral blood flow reduction and those with significant hypoperfusion.