In Journal of the American College of Radiology : JACR
OBJECTIVE : This quality assurance study assessed the implementation of a combined artificial intelligence (AI) and natural language processing (NLP) program for pulmonary nodule detection in the emergency department setting. The program was designed to function outside of normal reading workflows in order to minimize radiologist interruption.
MATERIALS AND METHODS : 19,246 CT exams including at least some portion of the lung anatomy performed in the emergent setting from October 1, 2021 to June 1, 2022 were processed by the combined AI/NLP program. The program utilized an artificial intelligence algorithm (AI algorithm) trained on 6 mm to 30 mm pulmonary nodules to analyze CT images and a natural language processing tool (NLP) to analyze radiological reports. Cases flagged as negative for pulmonary nodules by the NLP but positive by the AI algorithm were classified as suspected discrepancies. Discrepancies result in secondary review of exams for possible addenda.
RESULTS : Out of 19246 CT exams, 50 exams (0.26%) resulted in secondary review. 34/50 (68%) reviews resulted in addenda. Of the 34 addenda, 20 patients received instruction for new follow up imaging. Median time to addendum was 11 hours. The majority of reviews and addenda resulted from missed pulmonary nodules on CT exams of the abdomen and pelvis.
CONCLUSION : A background QA process utilizing AI and NLP helped improve the detection of pulmonary nodules and resulted in increased numbers of patients receiving appropriate follow up imaging recommendations. This was achieved without disrupting in-shift radiologist workflow or causing significant delays in patient follow for the diagnosed pulmonary nodule.
Cavallo Joseph J, de Oliveira Santo Irene, Mezrich Jonathan L, Forman Howard P
2023-Jan-31
Artificial Intelligence, Natural Language Processing, Pulmonary Nodules, Quality
