Diagnostic models built for our population
Most published clinical AI was trained on European, American, and East Asian cohorts. We partner with hospitals to build diagnostic models on our own data, with our own clinicians.
The data gap
Thyroid classifiers trained on East Asian cohorts. Capsule models trained on Europeans. Deployed in Riyadh, their calibration drifts in ways nobody publishes. We'd rather close that gap.
- Models trained on GCC patient data.
- A physician owner on every project.
- Open to IRB and validation partnerships.
Capsule Endoscopy Model
Small-bowel imaging, recalibrated.
A research-stage capsule endoscopy reading model trained on GCC patient data, because small-bowel disease here doesn't look like European cohorts.
Learn moreProductRectal Manometry Model
Anorectal physiology, properly interpreted.
A research-stage model for high-resolution anorectal manometry, calibrated to the patients these tests are actually run on here.
Learn moreProductThyroid Ultrasound Model
Nodule risk stratification, recalibrated.
A research-stage thyroid nodule model trained on Saudi and GCC ultrasound data, where iodine status and malignancy patterns diverge from existing cohorts.
Learn moreProductScopeReady
Predicting retained food before the scope.
A research-stage model that predicts retained gastric food before upper endoscopy, so prep can be personalized and day-of cancellations caught early.
Learn moreProductIBD Microbiome Model
Predicting IBD from the gut microbiome.
A research-stage model that predicts inflammatory bowel disease from fecal microbiome signatures — aiming at earlier, non-invasive triage before endoscopy.
Learn morePeer-reviewed work
The same team, publishing real models
Our diagnostic work is grounded in peer-reviewed machine-learning research, not just demos.
Partner with us
Research partnerships: IRBs, faculty co-investigators, shared publications.