ORIGINAL RESEARCH article
Front. Genet.
Sec. Statistical Genetics and Methodology
Volume 16 - 2025 | doi: 10.3389/fgene.2025.1636065
This article is part of the Research TopicMulti-Omics Insights into Autoimmune Diseases and Major Chronic Non-Communicable DiseasesView all articles
An Interpretable Machine Learning M odel for Predicting Myocardial Injury in Patients with High Cervical Spinal Cord Injury
Provisionally accepted
- 1First Affiliated Hospital of Fujian Medical University, Fuzhou, China
- 2Fudan University Shanghai Cancer Center Department of Anesthesiology, Shanghai, China
- 3Fujian Medical University Union Hospital, Fuzhou, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
High cervical spinal cord injury (HCSCI) is associated with severe au tonomic dysfunction and an increased risk of cardiovascular complications, including myocardial injury. However, early identification of myocardial inj ury remains challenging due to the lack of predictive tools.A total of 454 patients with HCSCI were retrospectively enrolled and divided into myocardial injury (n = 101) and non-injury (n = 353) groups.Univariate and multivariate logistic regression analyses were used to ide ntify independent risk factors. Four machine learning (ML) models-Logist ic Regression, Gradient Boosting Machine (GBM), Neural Network, and A daBoost-were constructed to predict myocardial injury, and model perfor mance was evaluated using area under the curve (AUC), F1 score, and average precision (AP). SHAP (SHapley Additive exPlanations) was appli ed for model interpretability.Multivariate analysis identified dyspnea (OR = 3.32, 95% CI: 1.49-7. 39) and low hematocrit (OR = 2.18, 95% CI: 1.04-4.57) as independent predictors of myocardial injury. Among the ML models, the Neural Netwo rk model achieved the highest AUC and F1 score in the testing set and demonstrated superior calibration and net clinical benefit. SHAP analysis revealed that dyspnea, LDL, spinal cord segment level, paralysis status, hematocrit, and myocardial injury stage were the top predictors. Individu alized SHAP force plots illustrated the contribution of each feature to pre diction outcomes.We developed an interpretable ML model capable of accurately predi cting myocardial injury in patients with HCSCI. The Neural Network mod el showed the best overall performance and, with SHAP interpretation, pr ovided transparent and individualized risk insights, supporting early diagn osis and targeted management in clinical practice.
Keywords: High cervical spinal cord injury, Myocardial injury, machine le arning, Neural Network, Shap, risk prediction, Interpretable artificial intelligence
Received: 27 May 2025; Accepted: 14 Jul 2025.
Copyright: © 2025 Li, Zhang, Liao, Wei, Huang, Lin, Chen and Chen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Hui Chen, First Affiliated Hospital of Fujian Medical University, Fuzhou, China
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.