Mammography Triage Software

Intelligently queue the most suspicious mammography screening cases to the top of your worklist.

Saige-Q identifies suspicious exams

Saige-Q™ mammography triage software helps radiologists manage their mammography screening worklists. DeepHealth’s powerful new AI technology automatically identifies suspicious screening exams that may need prioritized attention, allowing the user to optimize workflows for efficiency and effectiveness.

Saige-Q is compatible with both FFDM and DBT for an all-in-one triage solution.

The Solution

Typical workstations in mammography clinics are limited to simple sorting. Without triage software, there is no way for radiologists to identify and prioritize cases in the worklist that may have suspicious findings. Saige-Q quickly and accurately identifies these cases, enabling radiologists to address the most suspicious exams first.

How does it work?

Saige-Q uses an artificial intelligence algorithm to analyze incoming screening mammograms and assign a code to each exam. The Saige-Q code indicates the software’s suspicion as to whether findings may be present. For exams deemed suspicious, Saige-Q also generates a preview image of the most suspicious image. Radiologists can view the Saige-Q codes and preview images in a worklist on their workstation to sort and prioritize their cases accordingly.

Saige-Q delivers many benefits for radiologists, facilities, and patients:

  • It is compatible with both full-field digital mammography (FFDM) and digital breast tomosynthesis (DBT)—for an all-in-one triage solution for Hologic mammograms.
  • Radiologists can optimize their workflow forgreater efficiency.
  • Suspicious cases can quickly receive the clinical attention they need.
  • Exams can be intelligently assigned, based on Saige-Q codes.

Saige-Q is accurate:

  • In the pivotal study that led to FDA approval, Saige-Q achieved state-of-the-art AUC performance of 0.966 for FFDM and 0.985 for DBT.
  • At Breast Cancer Surveillance Consortium (BCSC) estimates of radiologist specificity, Saige-Q achieved a sensitivity of 91.2% for FFDM and 95.7% for DBT.
  • Saige-Q maintains high performance across dense and non-dense breasts and across different lesion types (calcifications and soft tissue densities).
  • All of the data used in the FDA pivotal study came from clinical sites that were never used for AI model development, demonstrating that Saige-Q exhibits high performance when applied at new clinical sites.

Saige-Q integrates seamlessly:

  • It functions with any standard clinical worklist software. • It fits seamlessly in the clinical workflow, making it easy to adopt in a normal screening regimen.
  • It is HIPAA- and DICOM-compliant.
  • It provides a range of operating points to optimize software performance for each clinic.

Additional Materials

Access the latest Saige-Q information here. Check back for updates.

Find out more.

To learn more about Saige-Q and to set up a demonstration, contact us: