โก Quick Summary
This study introduces a novel algorithm for automatic ICD code prediction using medical named entity recognition (NER) based on the BERT model. The algorithm achieved an impressive precision rate of approximately 90%, demonstrating its effectiveness in processing patient claims.
๐ Key Details
- ๐ Dataset: Medical claims data
- ๐งฉ Features used: Medical entities extracted from patient claims
- โ๏ธ Technology: Medical NER model based on BERT and ClinicalBERT
- ๐ Performance: Precision rate of approximately 90%
๐ Key Takeaways
- ๐ค Algorithm development focuses on automatic ICD encoding using advanced AI techniques.
- ๐ก Medical NER is employed to identify critical medical entities from patient claims.
- ๐ Embeddings are generated for both extracted entities and ICD codes to enhance prediction accuracy.
- ๐ Cosine similarity is used to match patient complaints with relevant ICD codes.
- ๐ Minimal data requirements for training make this algorithm accessible for various healthcare settings.
- ๐ High efficiency of the algorithm indicates its potential for real-world applications in medical coding.
- ๐๏ธ Local database stores embedding vectors and mapped ICD codes for quick retrieval.
๐ Background
The International Classification of Diseases (ICD) is essential for medical coding, providing a standardized way to classify diseases and health conditions. Traditional methods of ICD coding can be labor-intensive and prone to errors. With advancements in artificial intelligence and natural language processing, there is a growing interest in automating this process to improve efficiency and accuracy.
๐๏ธ Study
The researchers aimed to develop an algorithm that leverages medical named entity recognition to enhance the automatic prediction of ICD codes from patient claims. By utilizing the BERT model for NER and ClinicalBERT for generating embeddings, the study sought to create a more effective system for ICD encoding.
๐ Results
The proposed algorithm demonstrated a precision rate of approximately 90% when tested on a medical dataset. This high level of accuracy indicates that the algorithm effectively identifies relevant ICD codes based on the extracted medical entities from patient claims, showcasing its potential for practical application in healthcare settings.
๐ Impact and Implications
The implications of this study are significant for the healthcare industry. By automating the ICD coding process, healthcare providers can reduce administrative burdens, minimize errors, and enhance the overall efficiency of medical coding. This technology could lead to improved patient care and streamlined operations within healthcare systems, ultimately benefiting both providers and patients alike.
๐ฎ Conclusion
This study highlights the transformative potential of artificial intelligence in medical coding through the use of medical named entity recognition. The algorithm’s high precision rate suggests a promising future for automated ICD prediction, paving the way for further research and development in this area. As healthcare continues to evolve, integrating such technologies will be crucial for improving efficiency and accuracy in medical coding practices.
๐ฌ Your comments
What are your thoughts on the use of AI for automatic ICD coding? We would love to hear your insights! ๐ฌ Leave your comments below or connect with us on social media:
Using Medical Named Entity Recognition in Automatic ICD Prediction.
Abstract
The International Classification of Diseases (ICD) serves as a standard in medical coding. Researchers in artificial intelligence, including those focused on natural language processing and machine learning, have made a significant effort to build and develop automatic ICD encoding systems and algorithms. Many algorithms have been developed to implement automatic ICD encoding, but almost all of these algorithms depended on the raw text input without taking into consideration the important medical entities in this input. In this paper, we propose an algorithm for automatically predicting ICD codes based on patient claims. Our algorithm contains several steps for finding the most relevant ICD codes. Primarily, our proposed algorithm employs medical named entity recognition (NER) to find the most important medical entities in a patient claim. For this purpose, the Medical NER model was used based on the BERT model. Next, the algorithm generates embeddings for the extracted entities using the ClinicalBERT model. To identify the most relevant ICD code, the algorithm creates embeddings for an ICD catalog, which contains various information such as chapter descriptions, long descriptions, short descriptions, and ICD codes. The embedding process is primarily based on the long descriptions, and the results are stored in a local database that contains embedding vectors and corresponding mapped ICD codes. The final step of the algorithm calculates the cosine similarity between the embedding vector generated from the patient complaint and the ICD long description vectors. The strength of this new algorithm is that it first detects the medical entities in the textual input and then predicts the most similar ICD codes. Also, our developed algorithm does not need such huge data for training. We tested the developed algorithm on a medical dataset, and the results indicate that the proposed method is highly efficient, achieving a precision rate of approximately 90%.
Author: [‘Kawas M’, ‘Alkhatib B’, ‘Omar K’, ‘Tofelia K’, ‘Dashash M’, ‘Formanowicz D’]
Journal: Biomed Res Int
Citation: Kawas M, et al. Using Medical Named Entity Recognition in Automatic ICD Prediction. Using Medical Named Entity Recognition in Automatic ICD Prediction. 2025; 2025:6117755. doi: 10.1155/bmri/6117755