Project Description

CLIKODE

CLIKODE – Automatic Clinical Coding Processing, (3I) Innovation, Investigation of AI models in hospitalization for the codification of Procedures and Diagnostics is a project worth almost 300 thousand euros, benefiting the University Hospital Center of São João, EPE (UHCSJ), leader of the project, in partnership with the Faculty of Medicine of the University of Porto (FMUP) and CINTESIS.

Abstract

The execution of the operation will make it possible to automate the clinical coding of procedures and diagnostics using Artificial Intelligence techniques, with the development of a Learning Engine based on natural language processing of unstructured clinical texts to support the ICD-10-PCS classification. The project will have as context the production of the UHCSJ and the scientific validation of FMUP.

Objectives:

The project’s objectives are clearly defined, with a focus on:

  • Automatic generation of diagnostics and procedures of the ICD10 classification system, based on clinical texts;
  • Ensure that the coding process remains up to date over time;
  • Increase the quality of coding, i.e., increase the average number of coded procedures and diagnoses;
  • Reduce the monitoring structure (time and number of activities carried out) dedicated to the codification process, with a high impact on the number of human resources dedicated to the process in question.

Funding Institution

POCI – FSE in the framework of SATDAP-Training for Public Administration

Global Budget

297,897.74 €

CINTESIS Budget

52,897.70 €

Reference

POCI-05-5762-FSE-000230

Duration

From 07/01/2020 to 31/12/2022

CINTESIS Researchers Involved

Alberto Freitas, Pedro Rodrigues, Ricardo Correia, Fernando Lopes

Beneficiaries

University Hospital Center of São João, EPE (UHCSJ), leader of the project, in partnership with the Faculty of Medicine of the University of Porto (FMUP).