At the 2022 Radiological Society of North America (RSNA) Annual Meeting, Peter R. Eby, MD, of Virginia Mason Medical Center presented the results of the largest mammographic image quality evaluation to date, titled “Reduction in technical repeat and recall rate after implementation of artificial intelligence driven quality improvement software.”
Following the adoption of Volpara Analytics software in 2019, significant improvements in image quality metrics related to breast positioning and compression pressure were observed, alongside a significant 78% reduction in the technical repeats and recalls.
Volpara Analytics uses artificial intelligence (AI) to automatically and objectively measure the image quality of every mammogram image taken. The quantifiable metrics give technologists feedback at the image and study level to identify areas for improvement.
In a quality improvement review, the researchers evaluated “baseline” data from the first 12 months of AI usage (April 2019 – March 2020). These data were compared to a “current” dataset, based on the most recent 12 months of usage (April 2021 to March 2022). In total, “baseline” versus “current” data were evaluated from 42 technologists and over 210,000 images.
The comparison found 6% increases in both the overall mean Quality Score and the proportion of images scored by the software as “Perfect” or “Good” positioning, and an 8% increase in the proportion of images meeting target compression pressure. These improvements in objectively measured image quality indicators corresponded with technical repeats and recall rates decreasing from 0.77% (“baseline”) to 0.17% (“current”).
Lead study author Dr. Eby noted “this study highlights the potential for AI in revolutionizing what has traditionally been a very manual process, by providing continuous mammography image quality feedback to technologists on an unprecedented scale.”
Reducing unnecessary repeat imaging due to image quality issues improves patient care and the overall patient experience. “There are also benefits for mammography facilities in terms of lowering costs and improving workflow efficiencies for imaging center staff,” said Co-author and Global Clinical Research Lead at Volpara Health, Dr. Ariane Chan.
“AI in breast care is typically associated with CAD. We are hoping that improved image quality leads to improvements in cancer detection―it can improve mammography quality and density assessment, and transform patient care as well,” said Volpara CEO Teri Thomas. “Through the objective analysis of images, Volpara Analytics can help facilities improve image quality and reduce technical repeats and recalls.”