Artificial Intelligence to Identify Lung Contusions Using Chest X-rays in Blunt Thoracic Trauma An Observational Study

What this Trial Is About

Researchers are evaluating the accuracy of a computer-based system called LUNIT in interpreting chest X-rays to detect lung contusions in patients with blunt chest trauma. This observational study compares the results of LUNIT-assisted readings by expert radiologists with the standard radiologist interpretations, using chest HRCT scans as the gold standard. Lung contusions are important to identify accurately because they can lead to serious complications like respiratory failure and infections. The study involves reviewing past cases of patients who came to the Emergency Department of IRCCS AOU of Bologna between June 1, 2014, and June 1, 2024. These patients had blunt chest trauma and received a chest HRCT scan within 48 hours after a standard chest X-ray. Two expert radiologists will analyze the chest X-rays using the LUNIT software, and their findings will be compared to the original radiologist reports and the HRCT results. This is a retrospective, cross-sectional study without any changes to the usual care provided. Participants' chest X-rays and HRCT images will be reviewed to measure how well the LUNIT system detects lung contusions compared to standard methods. The main outcome is the sensitivity of lung contusion detection on chest X-rays using LUNIT, tracked up to about one year after the study. The study aims to see if the software can help reduce the need for HRCT scans, which would lower radiation exposure and speed up emergency department care.

AI in the Identification of Lung Contusions Through Chest Radiological Examination in Blunt Thoracic Trauma