AI more accurate than docs in challenging breast cancer diagnoses
An artificial intelligence system has outperformed pathologists in differentiating atypia from ductal carcinoma in situ—considered to be the greatest challenge in breast cancer diagnosis.
An artificial intelligence system has outperformed pathologists in differentiating atypia from ductal carcinoma in situ—considered to be the greatest challenge in breast cancer diagnosis.
In a diagnostic study involving 240 breast biopsy images, the performance of the AI system was compared with independent interpretations from 87 practicing U.S. pathologists.
“In the classification tasks of atypia and DCIS versus benign and DCIS versus atypia, the associated sensitivities are higher than the sensitivity of the practicing pathologists who independently interpreted the same specimens,” according to the study’s authors.
Results of the study, supported by the National Cancer Institute of the National Institutes of Health, were published last week in JAMA Network Open.
Also See: Algorithm as accurate as radiologists in assessing breast density, cancer risk
“Medical images of breast biopsies contain a great deal of complex data, and interpreting them can be very subjective,” says senior author Joann Elmore, professor of medicine at UCLA’s David Geffen School of Medicine and a researcher at the UCLA Jonsson Comprehensive Cancer Center.
“Distinguishing breast atypia from ductal carcinoma in situ is important clinically but very challenging for pathologists,” Elmore adds. “Sometimes, doctors do not even agree with their previous diagnosis when they are shown the same case a year later.”
Nonetheless, the AI system correctly determined whether scans showed DCIS or atypia more often than pathologists. In the study, the pathologists’ average sensitivity was 0.70 while the computer-based automated approach to interpreting breast pathology produced a sensitivity between 0.88 and 0.89 in differentiating DCIS from atypia.
“These results are very encouraging,” concludes Elmore. “There is low accuracy among practicing pathologists in the U.S. when it comes to the diagnosis of atypia and ductal carcinoma in situ, and the computer-based automated approach shows great promise.”
Going forward, researchers are looking to train the AI system to diagnose melanoma.
In a diagnostic study involving 240 breast biopsy images, the performance of the AI system was compared with independent interpretations from 87 practicing U.S. pathologists.
“In the classification tasks of atypia and DCIS versus benign and DCIS versus atypia, the associated sensitivities are higher than the sensitivity of the practicing pathologists who independently interpreted the same specimens,” according to the study’s authors.
Results of the study, supported by the National Cancer Institute of the National Institutes of Health, were published last week in JAMA Network Open.
Also See: Algorithm as accurate as radiologists in assessing breast density, cancer risk
“Medical images of breast biopsies contain a great deal of complex data, and interpreting them can be very subjective,” says senior author Joann Elmore, professor of medicine at UCLA’s David Geffen School of Medicine and a researcher at the UCLA Jonsson Comprehensive Cancer Center.
“Distinguishing breast atypia from ductal carcinoma in situ is important clinically but very challenging for pathologists,” Elmore adds. “Sometimes, doctors do not even agree with their previous diagnosis when they are shown the same case a year later.”
Nonetheless, the AI system correctly determined whether scans showed DCIS or atypia more often than pathologists. In the study, the pathologists’ average sensitivity was 0.70 while the computer-based automated approach to interpreting breast pathology produced a sensitivity between 0.88 and 0.89 in differentiating DCIS from atypia.
“These results are very encouraging,” concludes Elmore. “There is low accuracy among practicing pathologists in the U.S. when it comes to the diagnosis of atypia and ductal carcinoma in situ, and the computer-based automated approach shows great promise.”
Going forward, researchers are looking to train the AI system to diagnose melanoma.
More for you
Loading data for hdm_tax_topic #better-outcomes...