CORA is an efficient annotation tool to optimize manual annotation process of clinical records.
This GUI annotation tool is developed with PyQt5 package in Python.
System required: Python 3.6 or above
Author: Jie Yang.
CORA provides an efficient annotation GUI with customized keywords highlighted during the clinical health record annotation.
- Annotator Interface:
- Automatic Time Cost calculation: CORA will record the time cost for each annotation. Annotator needs to select specific "Patient ID" first and then annotate the specific case, the time cost will be automatical recorded after the annotater switched to another "Patient ID".
- Add text display settings (font size, line number)
- Add
power highlightmodel, to highlight more relevant terms - Support more data format (json, txt, database)
- Save and load project (not annotation result)
- Add button for next/previous patient/record
- Add Index date and shortcut to jump to notes nearby index date
- Include claims labels and EHR records, visualize the time distribution
- Discuss with users within division for needs
- Add result calculation and visualization
- Link to MGB or division datamarts database data selection and filter (maybe a different tool)
- Add AI recommendation (statistical ML, classical deep learning, LLMs), long-term plan
- Experiments and manuscript
- 2024-10-15, speed optimization, and time calculation update
- 2024-06-25, project launched.