Validation of Pediatric Weight Estimation Tools and Development of an AI-Driven Mobile Application

Introduction: Rapid and accurate estimation of weight is important inpediatric emergencies for safe drug administration and equipment sizing. We validated the Broselow tape, Dose by Growth tape, and Advanced Pediatric Life Support formula and developed new locally validated weight estimation formulas, which form the basis for an AI-driven mobile application (MedicAI).


Methods: We conducted a multicenter, cross-sectional study from December 2023 through June 2024 at two tertiary hospitals, enrolling 1,312 children. Weight estimates using the Broselow and Dose by Growth tapes and APLS formula were compared with actual weight. Concordance, correlation (Pearson/Spearman), and agreement (Cohen's Kappa, Bland–Altman) were analyzed. Regression modeling generated new formulas suitable for AI application.


Results: The Broselow tape was most concordant (62.7%) with realweights, followed by Dose by Growth (58.8%) and APLS (45.4%). Agreement analysis showed moderate concordance for Broselow and Dose by Growth and fair for APLS, all of which exceeded the clinically acceptable ranges (−5 to +5 kg). New formulas were derived: 0.4 × age (months) + 4.3 (infants) and 2 × age (years) + 8 (children). These were integrated in MedicAI, an AI-powered mobile app for real-time weight estimation from child height.


Conclusions: Broselow tape is the optimal traditional technique, but integration of locally validated formulas into an AI-powered mobile app offers real-time accurate weight estimation that has the potential to revolutionize pediatric emergency care globally, especially in resource-scarceareas.