Related Experiment Video
Updated: Dec 31, 2025
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
Use of natural language processing to improve predictive models for imaging utilization in children presenting to the
Xingyu Zhang1, M Fernanda Bellolio2, Pau Medrano-Gracia3
1Department of Systems, Populations and Leadership, University of Michigan School of Nursing, Ann Arbor, USA. zhangxyu@umich.edu.
Insights
Predictive models incorporating patient factors and natural language processing (NLP) accurately forecast medical imaging use in pediatric emergency departments (EDs). Combining structured and unstructured data significantly improved prediction accuracy for X-rays and CT scans.
Area of Science:
- Pediatric Emergency Medicine
- Health Informatics
- Medical Imaging Utilization
Background:
- Medical imaging is frequently utilized in pediatric emergency departments (EDs).
- Predicting imaging needs can optimize resource allocation and patient care.
- Understanding factors influencing imaging decisions is crucial for improving ED efficiency.
Purpose of the Study:
- To investigate the link between patient characteristics (socioeconomic, demographic, clinical) and medical imaging use in pediatric ED visits.
- To develop predictive models for medical imaging utilization in pediatric EDs, incorporating both structured and unstructured patient data.
- To assess the performance of predictive models using natural language processing (NLP) techniques on free-text clinical notes.
Main Methods:
- Utilized data from the 2012-2016 National Hospital Ambulatory Medical Care Survey for pediatric ED visits.
- Developed multivariable logistic regression models using structured (e.g., vital signs, age) and unstructured (e.g., reason for visit, nursing notes) variables.
- Applied NLP techniques to extract relevant information from unstructured clinical data for enhanced predictive modeling.
Main Results:
- Approximately 30.3% of pediatric ED visits involved medical imaging (X-ray: 25.0%, CT: 4.9%).
- Models incorporating unstructured data (NLP-extracted) showed higher predictive accuracy (c-statistics up to 0.85 for CT) than structured data alone (c-statistics up to 0.77 for CT).
- Combined models using both structured and unstructured data achieved the highest predictive performance (c-statistics up to 0.87 for CT).
Conclusions:
- Socioeconomic, demographic, and clinical factors at ED triage are associated with medical imaging utilization in pediatric patients.
- Predictive models integrating structured and unstructured data, particularly leveraging NLP, outperform models using only structured information.
- NLP holds significant potential for improving the prediction of resource utilization, such as medical imaging, in pediatric ED settings.
Objective:
To examine the association between the medical imaging utilization and information related to patients' socioeconomic, demographic and clinical factors during the patients' ED visits; and to develop predictive models using these associated factors including natural language elements to predict the medical imaging utilization at pediatric ED.
Methods:
Pediatric patients' data from the 2012-2016 United States National Hospital Ambulatory Medical Care Survey was included to build the models to predict the use of imaging in children presenting to the ED. Multivariable logistic regression models were built with structured variables such as temperature, heart rate, age, and unstructured variables such as reason for visit, free text nursing notes and combined data available at triage. NLP techniques were used to extract information from the unstructured data.
Results:
Of the 27,665 pediatric ED visits included in the study, 8394 (30.3%) received medical imaging in the ED, including 6922 (25.0%) who had an X-ray and 1367 (4.9%) who had a computed tomography (CT) scan. In the predictive model including only structured variables, the c-statistic was 0.71 (95% CI: 0.70-0.71) for any imaging use, 0.69 (95% CI: 0.68-0.70) for X-ray, and 0.77 (95% CI: 0.76-0.78) for CT. Models including only unstructured information had c-statistics of 0.81 (95% CI: 0.81-0.82) for any imaging use, 0.82 (95% CI: 0.82-0.83) for X-ray, and 0.85 (95% CI: 0.83-0.86) for CT scans. When both structured variables and free text variables were included, the c-statistics reached 0.82 (95% CI: 0.82-0.83) for any imaging use, 0.83 (95% CI: 0.83-0.84) for X-ray, and 0.87 (95% CI: 0.86-0.88) for CT.
Conclusions:
Both CT and X-rays are commonly used in the pediatric ED with one third of the visits receiving at least one. Patients' socioeconomic, demographic and clinical factors presented at ED triage period were associated with the medical imaging utilization. Predictive models combining structured and unstructured variables available at triage performed better than models using structured or unstructured variables alone, suggesting the potential for use of NLP in determining resource utilization.
