Clinical and Imaging Model for Predicting Post-Traumatic Nonunion

Table 1_An integrated clinical and imaging model for predicting post-traumatic nonunion.docx is a dataset from a retrospective cohort study of 343 patients with unilateral closed long bone fractures treated by internal fixation. The dataset was created by Bin Wang and last updated on 2026-04-13. It contains clinico-radiological variables used to develop and validate a machine learning model for predicting nonunion.

Use Cases

• Training a binary classifier for nonunion risk based on variables like Injury Severity Index and fracture gap width.
• Benchmarking machine learning algorithms (e.g., random forest, gradient boosting) for clinical prediction tasks.
• Developing interpretable clinical tools (e.g., nomograms) using SHAP analysis on the identified predictors.
• Studying the relationship between imaging features (cystic change volume, callus growth rate) and healing outcomes.

Strengths

• Model was validated on an independent set, achieving an AUC of 0.858 in validation.
• Dataset is based on a cohort of 343 patients, providing a foundation for model development.
• Methodology explicitly avoided information leakage via 10-fold cross-validation and a strict train/validation split.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect temporal or selection bias inherent to the single retrospective cohort study.

Provenance

Source

figshare, author Bin Wang.

Collection Method

Retrospective cohort analysis of patient records and imaging data.

Time Range

null

Freshness

Last updated 2026-04-13 04:35:17; freshness should be verified.

Geography

null