OpenMed — обработка клинических текстов локально

★ 5K

OpenMed is a tool for extracting structured information from clinical text and de-identifying it on hardware you control. Reach for it in medical and research settings where medical records must be handled privately: pull structured entities (diagnoses, medications, measurements) out of free text and strip patient personal data without sending records to someone else's cloud. Its core runtime performs extraction and de-identification locally once the needed models are available; it ships as a Python package (Python 3.10+) with a model catalog and has adapters to remote providers (review network use and each model/dataset's terms separately). Its focus is private clinical-text processing (NER + de-identification) on controlled hardware, not an electronic health record or a general NLP framework: the value is keeping sensitive medical data in-house. Mind your jurisdiction's rules for handling medical data and the models' licenses.