Consecutive cohort analysis with NLP and multi-finding AI algorithm for chest radiographs: Multidimensional opportunity for QA/QC.
Author
Garza Frias, Emiliano | Mass General Brigham, US
Scientific poster presentation (W5B-SPCH-4) at RSNA 2023, 26 – 30. November 2023 in Chicago, US
Purpose
Are consecutive cohort analyses with NLP and comprehensive decision-support AI for CXR a valid tool for quality control?
Method
NLP-based radiology reports search was performed on 3,760 adult CXR studies and compared with the result of a comprehensive AI algorithm for chest X-ray (annalise.ai).
Results
NLP extracted reports and AI output showed high levels of concordance. The AI model showed good performance at detecting over- and undercalls.
Key Takeaway
NLP and comprehensive AI discrepancy analysis can be a valuable approach to quality assessment and control.
Disclaimer
Harrison.ai Radiology Solutions were previously marketed as Annalise.ai solutions.
