Extraction of Radiological Characteristics From Free-Text Imaging Reports Using Natural Language Processing Among Patients With Ischemic and Hemorrhagic Stroke: Algorithm Development and Validation.
David E HsuAbdulaziz Tijjani BakoThomas B H PotterAlan P PanGavin W BritzJonika TannousFarhaan S VahidyPublished in: JMIR AI (2023)
Our proposed NLP pipeline achieved high performance and has the potential to improve medical research and patient safety.
Keyphrases
- patient safety
- quality improvement
- high resolution
- atrial fibrillation
- machine learning
- healthcare
- cerebral ischemia
- autism spectrum disorder
- deep learning
- smoking cessation
- ischemia reperfusion injury
- adverse drug
- risk assessment
- oxidative stress
- subarachnoid hemorrhage
- blood brain barrier
- fluorescence imaging
- climate change