Computer-assisted diagnosis can reduce the risk of missing a breast cancer on a mammography film by more than 50%, according to the latest report by leading researchers in the field.
"We believe the increasingly positive results with computer-assisted diagnosis demonstrate it can serve as a 'second opinion' for traditional screening mammograms," said Kunio Doi, professor of radiology at the University of Chicago, speaking last week in Atlanta at a breast cancer research meeting sponsored by the U.S. Department of Defense.
CAD systems scan digitized mammograms and use sophisticated software algorithms to highlight areas of microcalcification or increased mass that are sometimes missed by even experienced mammogram readers, Doi said. The computer completes its tasks about 10 seconds, he said.
The computer can't replace the radiologist or find tumors that are invisible on the mammogram, Doi noted. But it can help the radiologist by pointing out suspicious areas, especially in the mammograms of young women whose dense breast tissue may obscure signs of cancer.
In the study, Doi and colleagues used a computer-assisted diagnostic prototype to review the mammograms of 12,670 women who received routine breast cancer screening. Of those, 79 developed breast cancer, and 23 of those cancers were missed by the radiologist. When the computer went through the charts, it pointed out 12 of the tumors the doctors originally missed.
By Edward Susman
AuntMinnie.com contributing writer
June 13, 2000
Let AuntMinnie.com know what you think about this story.
Copyright © 2000 AuntMinnie.com


![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=100&q=70&w=100)






![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=112&q=70&w=112)







