Developers introduced the MedJev tool for extracting clinical data

Developers presented MedJev — a specialized tool for automated extraction of structured information from unstructured medical records. The software is oriented towards local deployment, allowing for the processing of confidential patient data without transmitting it to external servers.

The system operates on a single graphics card, providing high-speed processing of text notes. A key feature of the solution is the provision of confidence scores for each extracted field, enabling medical specialists to assess the algorithm's confidence in the accuracy of the obtained data. The tool is distributed without subscription fees, lowering the entry barrier for medical institutions.

The architecture of MedJev is optimized for on-premise operation, eliminating dependence on cloud APIs. This is critically important for complying with security protocols when working with personal medical information. The project's source code is available in an open repository, allowing for the integration of the solution into existing hospital information systems.

The implementation of such automation tools reduces the time spent on routine document processing. The transition from manual data entry to automated extraction with verification through confidence scores reduces the likelihood of errors when filling out electronic patient records and accelerates clinical documentation workflow.