The results could lead to expanded use of chest radiography in interstitial lung disease clinics, experts reported in the American Journal of Roentgenology.

Experts say this is the first study to assess how preprocessing techniques impact providers' ability to routinely extract similar quantitative imaging features from US scans.

Promising new research methods have been able to train AI to more accurately spot image anomalies.

Even though synthetic samples may not be perfect copies of original images, experts say they can act as a useful substitute for gadolinium-enhanced exams.

Imaging specialists remain irreplaceable, the experts maintained, particularly when looking for clinically significant findings.

This proved true for detecting pneumothorax, as networks developed on hand-labeled images outperformed tools refined via natural language processing.

Massachusetts General Hospital gathered reports from those on the front lines of the coming AI revolution, sharing their tips for success in JACR.

The FDA-approved algorithm can ultimately minimize patients' risk of developing secondary cancers later in life, experts said in Radiology: Artificial Intelligence.

The system also sent nearly 90% of dense breast scans for radiologist review due to abnormal lesions.

By comparison, more than 20% ranked radiology as their first choice when not considering AI's potential impact on the specialty.

Clinicians opted to use the AI tool nearly 60% of the time compared to directly searching for structured exam indications in 41% of cases.

The tool helps pathologists spot areas of interest on biopsy images and is designed to be used as an assistant.