AI Results | Visage Leading the Way in Transparency

As AI-generated measurements and segmentations (e.g., AI Results) have become more commonplace, the need arises to better inform the reporting radiologist that an AI algorithm was involved in the imaging under interpretation. This helps raise awareness that it was an AI algorithm rather than a peer radiologist who created the measurement (and/or AI Result(s); bolsters liability and regulatory protection (Software as a Medical Device [SaMD]); and facilitates data curation for research purposes, among other benefits.

Optimized during our work with one of our academic research partners, Visage 7 (e.g., Visage 7 Enterprise Imaging Platform) can display in Viewer Text that the AI algorithm result is a diagnostic aid, as well as pertinent information about the algorithm, by leveraging the DICOM Content Qualification attribute adhering to the IHE AI Results (AIR) profile. The information displayed is highly customizable, including the text color, contrast, and font size. In this example, the Viewer Text displays:

1. That the AI algorithm is a diagnostic aid
2. The AI algorithm name and version number in a contrasting color (green)
3. In study context - image acquisition information

Visage 7 | Viewer Text - DICOM Content Qualification attribute adhering to the IHE AI Results (AIR) Profile

Similar to the defined wording and format required for nutrition facts labels on prepared food items, progress has been made in labeling AI-generated results to help bring greater transparency. Visage is leading the way in providing a flexible, Optimal AI platform that supports the evolving requirements for successful AI deployment and real-world use.

MingDe Lin, PhD
Visage Imaging

Wolfgang Holler, MS
Visage Imaging