Knowledge Extraction and Outcome Prediction using Medical Notes

摘要: The increasing use of electronic health records (EHR) has allowed for an unprecedented ability to perform analysis on patient data. By training a number statistical machine learning classifiers over the unstructured text found in admission notes and operating procedures, prediction surgical procedure’s outcome can be performed. We extend initial bag-of-words model bag-of-concepts model, which uses cTakes UMLS extract medical terms concepts from notes. also improve knowledge extraction. Lastly, we propose exchange component, allows physicians provide feedback results further tune underlying classifier.