Data Extraction from Free-Text Reports on Mechanical Thrombectomy in Acute Ischemic Stroke Using ChatGPT: A Retrospective Analysis
Nils Lehnen, Franziska Dorn, Isabella C. Wiest, Hanna Zimmermann, Alexander Radbruch, Jakob Nikolas Kather, Daniel Paech
Abstract
By using data manually extracted by a radiologist as the ground truth, GPT-4 more frequently extracted correct data from free-text reports on mechanical thrombectomy in patients with ischemic stroke compared with GPT-3.5.
Topics & Concepts
MedicineMcNemar's testStroke (engine)Retrospective cohort studyIschemic strokeNeuroradiologyInternal medicineSurgeryEmergency medicineIschemiaNeurologyMathematicsStatisticsMechanical engineeringEngineeringPsychiatryAcute Ischemic Stroke ManagementArtificial Intelligence in Healthcare and EducationVenous Thromboembolism Diagnosis and Management