AI model detects malpositioned tubes in intubated patients

A group led by researchers at Seoul National University Hospital and including colleagues at AI software firm Lunit tested a deep-learning model trained on chest x-rays from intubated patients in intensive care units. The model performed well and could potentially be used to alert clinicians of cases where tubes need to be repositioned, according to the group.

“A deep-learning system exhibited excellent performance in identifying the presence of ET [endotracheal tube] and malposition of ET on chest radiographs,” noted corresponding author Dr. Eui Jin Hwang, in a thoracic imaging poster…