A keyphrase-based approach for interpretable ICD-10 code classification of Spanish medical reports

Area of Science:

• Medical Informatics
• Natural Language Processing
• Health Information Management

Background:

• Manual assignment of International Classification of Diseases, 10th Revision (ICD-10) codes to Electronic Health Records (EHRs) is resource-intensive.
• Developing automated systems is crucial for efficient health information management.

Purpose of the Study:

• To present an alternative system for automatically recommending ICD-10 codes for EHRs.
• To improve the efficiency and accuracy of medical coding processes.

Main Methods:

• Characterizing ICD-10 codes using keyphrases derived from EHRs and statistical analysis.
• Developing an information model to recommend codes based on EHR textual content.

Main Results:

• The proposed system demonstrates competitive performance compared to state-of-the-art approaches.
• The system can be combined with existing methods to enhance code recommendation accuracy.

Conclusions:

• The automated ICD-10 code recommendation system is effective and interpretable.
• Keyphrase associations provide valuable insights for both the system and other machine learning techniques.