AI4EU - COLONLP

AI4EU - COLONLP

Colon cancer screening

Transformers & Graph Algorithms against colon cancer (AI4EU – ColoNLP)

Development of a tool that, given a free medical text, using natural language processing (NLP) techniques, is able to automatically extract and identify ICD-10 diagnosis and procedure codes.

Intelligent health

Problem

The ultimate goal is to help, with the extraction of ICD-10 codes and in combination with molecular characteristics, to improve the algorithm available to Amadix for the prediction of colorectal cancer and to identify the population at high risk for colorectal cancer.

Results

The main idea is, using natural language processing (NLP) techniques focused on the Spanish language, to identify on these documents what other pathologies the patients suffer from and map them into the ICD-10 codes, which are standard codes for diseases and procedures, and to identify different comorbidities that could affect the development of colorectal cancer. Once the clinical records have been analyzed to obtain the data, these data will be combined with the results of the blood tests developed by AMADIX, to create Artificial Intelligence based models that allow:

  • Generate action plans aimed at the target population to prevent colorectal cancer or detect it at an early stage.
  • To identify potential profiles of people at high risk of developing cancer in its early stages and thereby reduce treatment costs and reduce the risk of death.
  • Centralize information to generate statistics and detect problems more quickly.

Skip to content