AnonCAT: De-identification NLP with CogStack / MedCAT

Cogstack previews the performance of NLP text redaction on real-world data in NHS hospitals, showing the power of a localised NLP model with performances in the order of >99% accuracy in redacting personal health identifiers with minimal impact on clinical text.

Read the research paper

© 2020 CogStack - PhiDataLab | Made by Suara