Case Study: InformAI achieves 97% organ volume estimation accuracy with V7 Darwin

How InformAI used V7 to build an organ volume estimation model achieving 97% accuracy

InformAI, a healthcare informatics company, wanted to improve organ transplant decision-making by building a fully automated organ volume estimation model for CT scan analysis. Using V7 Darwin, they needed a secure way to annotate DICOM scans and create ground truth for training a model that could accurately segment organs and compare donor and recipient sizes.

With V7, InformAI built browser-based medical labeling workflows, kept sensitive data in their own S3 bucket, and enabled remote radiologist collaboration with review by a tenured radiologist. The result was a model that reached 97.0% accuracy, helping InformAI’s TransplantAI product estimate organ size more precisely and reduce reliance on manual, approximate sizing methods in the time-critical transplant process.

InformAI

Britton Marlatt

Senior Data Scientist