Challenge:
Interpretation of chest X-rays is not only a routine but also a significant task for radiologist to inspect incorrect placement of life-supporting implantable medical devices such as endotracheal tube, nasogastric/enteric tubes and central venous lines. Improper placement of those tubes and lines can be life threatening if undetected. Our group has been pioneering in developing artificial intelligence (AI) system for detection of tubes and lines, determination of the region for safe tube and line placements, visualization of safe zone for tube and line placements, and alerting radiologists immediately for cases with incorrect tube/line placement, using a variety of image processing and machine learning techniques. Our system is expected to be extremely useful in a busy clinical environment, where radiologists may only have a very limited amount of time in reading a chest X-ray.
Solution:
Integrate the AI System developed by Matt Brown, PhD which detects placement of lines and tubes and adds overlays to the DICOM images onto a UCLA Health Information Technology OHIA (Office of Health Informatics & Analytics) managed server with an appropriate disaster recovery plan. Integrate the system with PACS to pass identified chest x-rays to the system to be marked-up and returned to PACS. With this solution, a marked-up image is returned to PACS for review within 4 minutes saving time and lives!