AI-Powered Pediatric Measurement for
Safer Emergency Care
Digital Broselow uses an AI model that combines age information with visual cues from images to predict a child's height and weight—accurately and instantly.
Learn MoreHow it works
What makes Digital Broselow tick ?
Secure image collection via a simple web app used in hospitals and schools
Automated face blurring to protect privacy
Background segmentation, normalization, and augmentation to ensure high-quality model training
Grad-CAM visualization for interpretable predictions based on true anatomical feature
Key Innovations
Locally Tailored Design
Most pediatric estimation tools are developed using growth data from Europe or North America, which often do not reflect the physical development patterns of children in Ghana. Digital Broselow is trained on locally collected pediatric data, ensuring predictions are grounded in the real anthropometric characteristics of Ghanaian children. By learning from height, weight, age, and visual features specific to this population, the system delivers more accurate and context-appropriate estimates, reducing reliance on subjective visual judgment and improving clinical confidence during emergency care.
Non-Contact Measurements
In emergency and critical care situations, physically weighing or measuring a child can be unsafe, time-consuming, or simply impractical. Digital Broselow eliminates this challenge by providing rapid, non-contact height and weight estimation directly from images. This approach minimizes unnecessary handling, reduces stress for both patients and caregivers, and allows healthcare providers to act faster—supporting safer, more efficient emergency decision-making when time is critical.
Strong Early Performance
Initial testing on early datasets has shown promising accuracy, with results indicating the system's ability to outperform commonly used estimation tools when applied to Ghanaian children. These early findings demonstrate the model's potential to significantly reduce dosing errors linked to inaccurate weight estimation. While further validation is ongoing, the preliminary results provide strong evidence that Digital Broselow can evolve into a clinically reliable tool, capable of improving pediatric outcomes in real-world hospital and emergency settings.
BUILT FOR
IMPACT
Digital Broselow reflects Dipper Lab's commitment to safe, effective, and context-aware AI healthcare solutions. Once completed, it will serve as a reliable tool for pediatric care in Ghana—with the potential to scale across Africa and other regions facing similar challenges.
QUICK LINKS
© 2025. All rights reserved