Stanford AI tool boosts ultrasound efficacy: 3 things to know

An AI tool developed and evaluated by researchers at Stanford (Calif.) Medicine was able to identify 44% more lesions than human interpreters when reviewing prostate biopsy ultrasound images, according to a study published Jan. 28 in European Urology Oncology. 

The tool, called ProCUSNet, was designed by Mirabela Rusu, PhD, an assistant professor in the department of radiology, an Aug. 13 news release from Stanford Medicine said. 

Here are three things to know about the tool:

1. Traditionally, ultrasound imaging is used to guide prostate biopsy procedures where 12 to 16 “scattered samples” are acquired and reviewed for signs of cancer. These samples provide “limited insight” and lead up to 52% of clinically significant cancers to go undetected, the release said.
2. ProCUSNet was able to detect 82% of clinically significant cancers and 44% more lesions than human interpreters reviewing the same ultrasound images.
3. As the tool was developed with data from a specific ultrasound system, “broader validation across different devices and practice settings will be necessary,” the release said. “Future studies will also explore faster integration, prospective clinical trials and adaptation to other ultrasound modalities.”

Read the full study here.

The post Stanford AI tool boosts ultrasound efficacy: 3 things to know appeared first on Becker’s Hospital Review | Healthcare News & Analysis.