Journal DOI 10.68956/journal.human-performance-futures

Creative Commons source reading · Machine learning for performance

Development and validation of an automatic machine learning-based myocardial contusion prediction model in rib fracture patients: a retrospective multicenter cohort study

Yang Y; Liu P; Jia N; Xu Y; Wu P; Zhang Z; Han T; Jiang Y; Huang X; Li L

Originally published in Front Med (Lausanne) · 2026-09-03

From the article

Selected source text. JournalOS removes markup and shortens excerpts where indicated. The full article provides the complete context.

Abstract

AbstractObjectiveThis study aims to construct an automatic machine learning-based prediction model to accurately identify the risk of myocardial contusion in traumatic rib fracture patients, addressing the core limitations of insufficient sensitivity in traditional diagnostic methods and difficulties in clinical translation of existing prediction models.MethodsA multicenter retrospective study design was adopted, including 1,038 rib fracture patients from 8 hospitals (training set: 808 cases; testing set: 230 cases) from June 2024 to June 2025. An automatic machine learning framework (AutoML) was developed based on the Improved Crested Porcupine Optimization algorithm (ICPO). A dual-stage optimization process achieved key feature screening and hyperparameter tuning, with Synthetic Minority Over-sampling Technique (SMOTE) addressing…

Discussion

This study systematically constructed and validated an automatic machine learning prediction model for myocardial contusion (MC) in rib fracture patients, with its core findings highlighting the clinical value of intelligent technology for optimizing trauma patient risk stratification and guiding early intervention. Given that current clinical practice relies mainly on electrocardiographic changes and traditional serological markers like troponin for early MC diagnosis. However, these anatomical trauma scores do not contain diagnostic elements of myocardial contusion (such as electrical abnormalities or troponin dynamics); in this study, the cardiac AIS code corresponding to myocardial contusion was likewise excluded from the score calculation. Therefore, these variables are mathematically independent as predictive features, yet they…

Conclusion

In summary, this study has preliminarily established an intelligent early-warning system for myocardial contusion risk in rib fracture patients through comprehensive clinical data mining, advanced algorithm optimization, multi-dimensional feature selection validation, and human-centered multi-unit collaborative design. The SHAP explainable platform has enabled refined multimodal feature contribution analysis, establishing a visualized framework for exploring potential high-risk factor combinations that warrant further investigation. It is important to emphasize that SHAP analysis quantifies feature contributions within the predictive model but cannot prove biological mechanisms, causal relationships, or mechanical pathways. These exploratory findings therefore generate hypotheses regarding possible associations between anatomical…

Editorial reading guide

  • What data created and tested the model, and was evaluation independent?
  • What is the cost of a false positive, false negative or out-of-distribution decision?