An AI platform for predicting and processing medical coding information for optimization and quality monitoring of clinical treatment and invoicing

Operating time: Oct 2022-Sep 2024
Source of funding: Invest BW

Correct coding of medical diagnoses and treatments are prerequisites for software support of the processes of medical service providers. However, due to the complexity of relevant standards (ICD, OPS), these codes are very time-consuming and error-prone. In particular, comorbidities are often not prioritized very often in clinical documentation, which can lead to incomplete coding, incomplete billing, and suboptimal treatment.

With AI4MedCode, an AI platform is being developed to integrate and represent various data sources (doctor's letter, medical history sheet, examination results, etc.) in knowledge graphs, using machine learning to generate ICD and OPS codes, as well as providing services based on these codes, which can

  • improve the efficiency and quality of coding (AI4MedCode Assistant),
  • suggest guideline-compliant processes or check for them (AI4MedCode-ProWarn and QA), and
  • improve billing and control (AI4MedCode-MIS).

These services will improve the efficiency and quality of treatment and billing processes in hospitals.



  1. Wang, Y., Dima, C., & Staab, S. (2023). WikiMed-DE: Constructing a Silver-Standard Dataset for German Biomedical Entity Linking using Wikipedia and Wikidata. The 4th Wikidata Workshop @ ISWC 2023.

Project Members

To the top of the page