AI-based ICD coding and classification approaches using discharge summaries: A systematic literature review

Kaur, Rajvir; Ginige, Jeewani Anupama; Obst, Oliver

doi:10.1016/j.eswa.2022.118997

Cited by 21 publications

(16 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Like other literature reviews [13][14][15][16][17][18][19], this research has been carried out following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analysis) guidelines, whose objective is to help authors of systematic reviews, scoping reviews, among others, to generate a document free of bias.…”

Section: Methodsmentioning

confidence: 99%

A Scoping Review of Pipeline Maintenance Methodologies Based on Industry 4.0

et al. 2022

View full text Add to dashboard Cite

The fourth industrial revolution was a milestone at the industrial level. It forced most industries to evolve technically and for their collaborators to prepare and advance together with technology; the oil industry was no exception. It develops its activities in dangerous and dynamic environments and needs to protect its human resources, equipment, and infrastructure. This article presents a scoping review, based on the PRISMA guidelines, of pipeline maintenance methodologies based on Industry 4.0. From the first collection of 123 articles from prestigious databases such as SpringerLink, MDPI, Scopus, IEEEXplore and ACM, a final sample of 31 articles was obtained. Here, technologies that enhance preventive and predictive maintenance systems are discussed. The results show that predictive maintenance compared to preventive maintenance has a percentage difference in upkeep time optimization of 38% in the last five years. This difference was corroborated with a Student’s t-test for independent samples, with a significance of 0.023. Likewise, the most used technologies were analyzed, with artificial intelligence standing out with 45.16%.

show abstract

Section: Methodsmentioning

confidence: 99%

A Scoping Review of Pipeline Maintenance Methodologies Based on Industry 4.0

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Automated Medical Coding (AMC) is the idea that artificial intelligence can automate clinical coding. In recent years, there has been a significant increase in AMCrelated work [14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32] through deep learning. Although research in this field has grown, this problem is far from being solved [18,33].For instance, automated coding remains a complex problem because extracting knowledge from patients' clinical records is challenging.…”

Section: Background and Significancementioning

confidence: 99%

Medical Diagnosis Coding Automation: Similarity Search vs. Generative AI

Klotzman

2024

Preprint

View full text Add to dashboard Cite

Objective: This study aims to predict ICD-10-CM codes for medical diagnoses from short diagnosis descriptions and compare two distinct approaches: similarity search and using a generative model with few-shot learning. Materials and Methods: The text-embedding-ada-002 model was used to embed textual descriptions of 2023 ICD-10-CM diagnosis codes, provided by the Cen-ters provided for Medicare & Medicaid Services. GPT-4 used few-shot learning. Both models underwent performance testing on 666 data points from the eICU Collaborative Research Database. Results: The text-embedding-ada-002 model successfully identified the relevant code from a set of similar codes 80 percent of the time, while GPT-4 achieved a 50 percent accuracy in predicting the correct code. Discussion: The work implies that text-embedding-ada-002 could automate medical coding better than GPT-4, highlighting potential limitations of generative language models for complicated tasks like this. Conclusion: The research shows that text-embedding-ada-002 outperforms GPT- 4 in medical coding, highlighting embedding models usefulness in the domain of medical coding.

show abstract

“…(23) A systemic review of studies from 2010 to 2021 provided an overview of automatic ICD coding assignment systems that utilized NLP, machine learning, and deep learning techniques, and concluded that deep learning models were found to be better than other traditional machine learning models when automating clinical coding systems. (24) Utilizing NLP techniques such as Word Embedding (a representation of words and phrases by vectors in a low-dimensional space such that it retains semantic and syntactic information) and a Convolutional Neural Network model (a deep learning algorithm that captures hierarchical patterns in textual data utilizing convolutional layers), another study processed 21,953 clinical records from ve departments, signi cantly enhancing the accuracy of automated ICD-10 code predictions and potentially easing the manual coding process for physicians. ( 25) A similar study analyzed the use of a natural language processing-bidirectional recurrent neural network (NLP-BIRNN) algorithm to optimize the medical records and identi ed areas of error by medical coders.…”

Section: Ai For Assigning Icd Codesmentioning

confidence: 99%

Fine-Tuning for Accuracy: Evaluation of GPT for Automatic Assignment of ICD Codes to Clinical Documentation

Nawab,

Fernbach,

Atreya

et al. 2024

Preprint

View full text Add to dashboard Cite

Background: Assignment of International Classification of Disease (ICD) codes to clinical documentation is a tedious but important task that is mostly done manually. This study evaluated the widely popular OpenAI’s Generative Pretrained Model (GPT) 3.5 Turbo in facilitating the automation of assigning ICD codes to clinical notes. Methods: We identified the 10 most prevalent ICD-10 codes in the Medical Information Mart for Intensive Care (MIMIC-IV) dataset. We selected 200 notes for each code, and then split them equally into two groups of 100 each (randomly selected) for training and testing. We then passed each note to GPT 3.5 Turbo via OpenAI’s API, prompting the model to assign ICD-10 codes to each note. We evaluated the model’s response for the presence of the target ICD-10 code. After fine-tuning the GPT model on the training data, we repeated the process with the test data, comparing the fine-tuned model’s performance against the default model. Results: Initially the target ICD-10 code was present in the assigned codes by the default GPT 3.5 Turbo model in 29.7% of the cases. After fine-tuning with 100 notes for each top code, the accuracy improved to 62.6%. Conclusions: Historically, GPT’s performance for healthcare related tasks is sub-optimal. Fine-tuning as in this study provides great potential for improved performance, highlighting a path forward for integration of Artificial Intelligence (AI) in healthcare for improved efficiency and accuracy of this administrative task. Future research should focus on expanding the training datasets with specialized data and exploring the potential integration of these models into existing healthcare systems to maximize their utility and reliability.

show abstract

AI-based ICD coding and classification approaches using discharge summaries: A systematic literature review

Cited by 21 publications

References 43 publications

A Scoping Review of Pipeline Maintenance Methodologies Based on Industry 4.0

A Scoping Review of Pipeline Maintenance Methodologies Based on Industry 4.0

Medical Diagnosis Coding Automation: Similarity Search vs. Generative AI

Fine-Tuning for Accuracy: Evaluation of GPT for Automatic Assignment of ICD Codes to Clinical Documentation

Contact Info

Product

Resources

About