Development of an AI Assessment System for Pediatric Respiratory Distress : A Prospective Study

Samsung Medical Center2,200 enrolled

Overview

This is a multicenter, prospective observational study designed to collect clinical data for the development of a vision-language model-based artificial intelligence system for automated assessment of pediatric respiratory patterns. The study enrolls pediatric patients aged 0 to 12 years who present to the pediatric emergency departments of participating institutions. Clinical and visual respiratory data are collected along with baseline clinical characteristics, including sex, age, body weight, height, presenting symptoms recorded at emergency department arrival, initial vital signs (body temperature, pulse rate, respiratory rate, blood pressure, and oxygen saturation), severity at presentation assessed by the Korean Triage and Acuity Scale (KTAS), emergency department management and outcomes such as hospital admission or discharge, and other relevant clinical information. These data are used for cohort characterization and for the development and evaluation of an AI-based system that aims to automatically analyze pediatric respiratory patterns and support objective respiratory assessment in pediatric emergency care.

Study Type

OBSERVATIONAL

Enrollment

2,200

Conditions

Pediatric Respiratory Distress

Eligibility

Sex: ALLMin age: 0 YearsMax age: 12 Years

Medical Language ↔ Plain English

Inclusion Criteria: * Pediatric patients aged 0 to 12 years who present to participating pediatric emergency departments. * Patients for whom clinical data, including basic demographic characteristics (e.g., age, body weight), severity at presentation (e.g., KTAS level), vital signs, emergency department management and outcomes, as well as visual respiratory data, are available during emergency care. Exclusion Criteria: * Patients outside the specified age range. * Patients with insufficient or poor-quality clinical or visual respiratory data. * Patients whose data cannot be used due to withdrawal of consent or regulatory restrictions.

Locations (3)

CHA Bundang Medical Center, CHA University, 9, Yatap-ro, Bundang-gu

Seongnam-si, Gyeonggi-do, South Korea

RECRUITING

Asan Medical Center, 88, Olympic-ro 43-gil, Songpa-gu

Seoul, Seoul, South Korea

RECRUITING

Samsung Medical Center, 81, Irwon-ro, Gangnam-gu

Seoul, Seoul, South Korea

RECRUITING

Outcomes

Primary Outcomes

Accuracy of automated respiratory pattern assessment

Performance of a vision-language model-based system in assessing pediatric respiratory patterns using clinical and visual respiratory data collected in pediatric emergency departments.

Time frame: From study start through study completion (up to December 2027)