Related Experiment Video
Updated: Jun 6, 2025
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
Published on: December 15, 2023
Predicting abnormal C-reactive protein level for improving utilization by deep neural network model
Donghua Mo1, Shilong Xiong1, Tianxing Ji1
1Clinical Laboratory Medicine Department, The Second Affiliated Hospital of Guangzhou Medical University, Guangzhou, China.
Insights
Deep neural network (DNN) models effectively predict C-reactive protein (CRP) levels using complete blood count (CBC) data. This AI approach can improve the clinical utility of CRP testing, aiding in inflammatory diagnostics.
Area of Science:
- Biomedical Informatics
- Artificial Intelligence in Healthcare
- Clinical Diagnostics
Background:
- C-reactive protein (CRP) is a key inflammatory biomarker.
- Current clinical use of CRP testing suffers from overuse and underuse due to insufficient evidence-based guidelines.
- Predictive models are needed to optimize CRP test ordering.
Purpose of the Study:
- To develop and validate deep neural network (DNN) models for predicting normal and abnormal C-reactive protein (CRP) levels.
- To enhance the appropriate and intelligent ordering of CRP tests in clinical practice.
- To leverage complete blood count (CBC) parameters for CRP level prediction.
Main Methods:
- Utilized a large dataset of 53,834 medical records for model development.
- Employed complete blood count (CBC) parameters as feature vectors.
- Compared DNN models against other machine learning algorithms including support vector classification, logistic regression, decision trees, and random forests.
- Externally validated the best performing DNN models on an independent dataset of 20,723 samples using discrimination, calibration curves, and decision curve analysis.
Main Results:
- Deep neural network (DNN) models demonstrated superior performance with the highest Area Under the Receiver Operating Characteristic Curve (AUC).
- Internal validation showed an AUC of 0.818, balanced accuracy of 0.741, and F1 score of 0.649.
- External validation yielded comparable results with an AUC of 0.817, balanced accuracy of 0.741, and F1 score of 0.641.
- The CRP-C2 model was identified as the target model due to its lowest Brier score (0.154) and excellent calibration (y=1.001x-0.010).
Conclusions:
- DNN models provide moderate performance in distinguishing binary C-reactive protein (CRP) levels, outperforming baseline methods.
- The developed models exhibit good generalization and calibration, indicating reliability.
- The CRP-C2 model can optimize CRP test utilization and support inflammatory diagnostics, particularly in primary care settings where CBC data is available but CRP testing may be limited.
Background:
C-reactive protein (CRP) is an inflammatory biomarker frequently used in clinical practice. However, insufficient evidence-based ordering inevitably results in its overuse or underuse. This study aims to predict its normal and abnormal levels using the deep neural network (DNN) models, helping clinicians order this item more appropriately and intelligently.
Methods:
We considered complete blood count (CBC) parameters as feature vectors and 10 mg/L as a cutoff value for CRP. Several models, including linear support vector classification, logistic regression, decision trees, random forests, and DNN, were developed based on a dataset of 53834 medical records to predict binary output. We externally validated DNN models on independent 20723 samples through discrimination, calibration curve, and decision curve analysis.
Results:
DNN models has the best area under the receiver operating characteristic curves (AUC). Learning curves revealed that models' AUC, balanced accuracy, and F1 score do not significantly and continuously improve following increasing data volume. In internal validation, the AUC, balanced accuracy, and the F1 score of 10 models were 0.818 (0.95 CI: 0.812-0.824), 0.741 (0.95 CI: 0.736-0.747), and 0.649 (0.95 CI: 0.643-0.656), respectively. These metrics were 0.817 (0.95 CI: 0.816-0.817), 0.741 (0.95 CI: 0.740-0.742), and 0.641 (0.95 CI: 0.640-0.642), respectively, in external validation. AUC and balanced accuracy shown no significant difference (P-values were 0.106 and 0.339). CRP10-C2 model has the lowest Brier score of 0.154, AUC of 0.818, and calibration curve formula of y=1.001x-0.010, which was identified as a target model to deploy in the app.
Conclusions:
DNN models obtained moderate performance, surpassing baseline indices in distinguishing binary CRP levels. They are good generalizations and well-calibrated. The CRP-C2 model can enhance CRP utilization by informing the orders appropriately and can contribute to inflammatory diagnostics in primary health care where CBC is available, but the CRP test is inaccessible.
