By Stephen Beech
Breast cancer can be detected more accurately using artificial intelligence, according to a new study.
Researchers found that AI improves breast cancer detection accuracy for radiologists when reading screening mammograms, helping them devote more of their attention to suspicious areas.
Previous studies have shown that AI for decision support improves radiologist performance by increasing sensitivity for cancer detection without extending reading time.
However, the positive impact of AI on radiologists’ visual search patterns remained underexplored until now.
Dutch researchers used an eye-tracking system to compare radiologists’ performance and visual search patterns when reading screening mammograms with and without an AI decision support system.
The system included a small camera-based device positioned in front of the screen with two infrared lights and a central camera.
The infrared lights illuminate the radiologist’s eyes, and the reflections were captured by the camera, allowing for the computation of the exact coordinates of the radiologist’s eyes on the screen.
Study joint first author Jessie Gommers, from the Department of Medical Imaging at Radboud University Medical Center, said: “By analyzing this data, we can determine which parts of the mammograms the radiologist focuses on, and for how long, providing valuable insights into their reading patterns.”
Twelve radiologists read mammography examinations from 150 women, 75 with breast cancer and 75 without, for the study, published in the journal Radiology.
Breast cancer detection accuracy among the radiologists was higher with AI support compared with unaided reading.
There was no evidence of a difference in mean sensitivity, specificity or reading time.
Doctoral candidate Gommers said: “The results are encouraging.
“With the availability of the AI information, the radiologists performed significantly better.”
Eye tracking data showed that radiologists spent more time examining regions that contained actual lesions when AI support was available.
Gommers said: “Radiologists seemed to adjust their reading behavior based on the AI’s level of suspicion: when the AI gave a low score, it likely reassured radiologists, helping them move more quickly through clearly normal cases.
“Conversely, high AI scores prompted radiologists to take a second, more careful look, particularly in more challenging or subtle cases.”
She said the AI’s region markings functioned like visual cues, guiding radiologists’ attention to potentially suspicious areas.
She explained that, in essence, the AI acted as an additional set of eyes, providing the radiologists with additional information that enhanced both the accuracy and efficiency of interpretation.
Gommers said: “Overall, AI not only helped radiologists focus on the right cases but also directed their attention to the most relevant regions within those cases, suggesting a meaningful role for AI in improving both performance and efficiency in breast cancer screening.”
She noted that overreliance on erroneous AI suggestions could lead to missed cancers or unnecessary recalls for additional imaging.
However, multiple studies have found that AI can perform as well as radiologists in mammography interpretation, suggesting that the risk of erroneous AI information is relatively low.
To mitigate the risks of errors, Gommers says it is important that the AI is highly accurate and that the radiologists using it feel accountable for their own decisions.
She said: “Educating radiologists on how to critically interpret the AI information is key.”
The researchers are now conducting additional reader studies to explore when AI information should be made available, such as immediately upon opening a case, or on request.
The team is also developing methods to predict if AI is uncertain about its decisions.
Gommers added: “This would enable more selective use of AI support, applying it only when it is likely to provide meaningful benefit.”
