Chest X-ray Triage Solution Sorts Emergency Cases

The algorithm performs at 94-96% sensitivity, 95-99% specificity, and 97-99% accuracy rates, reliably detecting lung nodules, calcifications, consolidation, fibrosis, pneumothorax, pneumoperitoneum, cardiomegaly, pleural effusion, mediastinal widening, atelectasis, and tuberculosis.

CXR Triage generates the location information of detected lesions in in the form of heatmaps and/or contour maps, with an abnormality score reflecting the AI’s calculation of the actual presence of the detected lesion, and also provides a case report summarizing the overall analysis result, narrowed down to each finding. The prioritized cases, according to the abnormality scores, reduce subsequent reading time by 65% for normal cases and 25% for abnormal cases.

“Lunit INSIGHT CXR was designed to assist doctors with faster and more accurate diagnosis,” said Brandon Suh, CEO of Lunit. "We strongly believe that sooner or later AI will become the new standard of care, and AI will be used everywhere as a must-use product. Not only will it be used as a tool to make the workflow more efficient, but it will ensure better diagnosis and healthier life for patients.”

CXR is the main modality used for screening and diagnosis of thoracic injuries in trauma patients, used to visualize rib fractures, lung contusions, pneumothorax and hemothorax, emphysema, diaphragmatic and aortic injury, and fractures of the axial skeleton. In common practice, a CXR taken in the emergency department is assessed by the trauma team.

Related Links:
Lunit