CT CAD may help chest trauma patients avoid surgery
A new CT computer-aided detection (CAD) application to detect and measure pneumothoraces in trauma patients helps physicians make quicker and more accurate decisions in emergency room settings, according to a study performed at Massachusetts General Hospital and Harvard Medical School in Boston. The research appears in the March issue of the American Journal of Roentgenology.
The size of a pneumothorax is an important index to guide the emergency treatment of trauma patients-chest tube drainage. The purpose of this study was to develop and validate an automated computer-aided volumetry scheme for detection and measurement of pneumothoraces for trauma patients imaged with multi-detector CT (MDCT), according to the study's lead author Wenli Cai, PhD, and colleagues.
Three pigs and 68 trauma patients with at least one diagnosed occult pneumothorax were selected for the development and validation of our computer-aided volumetry scheme for pneumothorax. Computer-aided volumetry of pneumothorax consisted of five automated steps: extraction of pleural region, detection of pneumothorax candidates, delineation of the detected pneumothorax candidates, reduction of false-positive findings, and report of the volumetric measurement of pneumothoraces.
According to the results, 30 to 39 percent of all patients suffering from chest trauma have pneumothorax.
In the animal study, the computer-aided volumetry scheme yielded a mean value of 24.27 mL compared with 25 mL of air volume manually injected in each scan, the authors wrote. The correlation coefficients were 0.999 and 0.997 for the in vivo and ex vivo comparison, respectively. In the patient study, the sensitivity of our computer-aided volumetry scheme was 100 percent with a false-positive rate of 0.15 per case for 32 occult pneumothoraces 25 mL. The correlation coefficient was 0.999 for manual volumetry comparison. This automated computer-aided volumetry scheme took approximately 3 minutes to finish the detection and measurement per case, the authors noted.
"Aside from being quick and accurate, MDCT and the computer-aided method may help avoid unnecessary surgeries, too. When the pneumothorax is small and the patient is stable, physicians do not need to perform surgery. The computer-aided quantification method can quickly show us how large the pneumothorax is and give accurate monitoring information over the course of several days," said Cai.
The size of a pneumothorax is an important index to guide the emergency treatment of trauma patients-chest tube drainage. The purpose of this study was to develop and validate an automated computer-aided volumetry scheme for detection and measurement of pneumothoraces for trauma patients imaged with multi-detector CT (MDCT), according to the study's lead author Wenli Cai, PhD, and colleagues.
Three pigs and 68 trauma patients with at least one diagnosed occult pneumothorax were selected for the development and validation of our computer-aided volumetry scheme for pneumothorax. Computer-aided volumetry of pneumothorax consisted of five automated steps: extraction of pleural region, detection of pneumothorax candidates, delineation of the detected pneumothorax candidates, reduction of false-positive findings, and report of the volumetric measurement of pneumothoraces.
According to the results, 30 to 39 percent of all patients suffering from chest trauma have pneumothorax.
In the animal study, the computer-aided volumetry scheme yielded a mean value of 24.27 mL compared with 25 mL of air volume manually injected in each scan, the authors wrote. The correlation coefficients were 0.999 and 0.997 for the in vivo and ex vivo comparison, respectively. In the patient study, the sensitivity of our computer-aided volumetry scheme was 100 percent with a false-positive rate of 0.15 per case for 32 occult pneumothoraces 25 mL. The correlation coefficient was 0.999 for manual volumetry comparison. This automated computer-aided volumetry scheme took approximately 3 minutes to finish the detection and measurement per case, the authors noted.
"Aside from being quick and accurate, MDCT and the computer-aided method may help avoid unnecessary surgeries, too. When the pneumothorax is small and the patient is stable, physicians do not need to perform surgery. The computer-aided quantification method can quickly show us how large the pneumothorax is and give accurate monitoring information over the course of several days," said Cai.