What question did this study set out to answer?

To develop and evaluate an accurate, interpretable machine learning model (Graph Neuro Boost) to predict hospital length of stay for patients diagnosed with coronary artery disease.

August 15, 2026Connection ScienceOpen Access

Enhanced CAD patient length of stay prediction using the Graph Neuro Boost model

View Full Paper

Ask AI

Bookmark

Authors

GMGeetha Pratyusha Miriyala ASArun Kumar Sinha

Discussion

Member takes

Overview

Machine learning study demonstrates 98.57% accuracy in predicting hospital length of stay for coronary artery disease patients, highlighting potential for improved clinical resource management.

Key Points

To develop and evaluate an accurate, interpretable machine learning model (Graph Neuro Boost) to predict hospital length of stay for patients diagnosed with coronary artery disease.
Extracted and preprocessed patient records from the MIMIC-III clinical database.
Developed a Graph Neuro Feature Selector that combined autoencoder-based embedded methods with graph theory and Kruskal's algorithm to identify key predictive variables.
Trained an XGBoost model optimized via Bayesian optimization with balanced Expected Improvement and Upper Confidence Bound, utilizing SHAP analysis to assess feature importance.
The Graph Neuro Feature Selector identified 16 critical clinical features in 2.68 seconds.
Graph Neuro Boost achieved an overall prediction accuracy of 98.57%, demonstrating superior performance compared to baseline models across multiple evaluation metrics.

Cite This Study

Miriyala et al. (2026) studied this question.

synapsesocial.com/papers/6a80199875c2e31742c85895 https://doi.org/10.1080/09540091.2026.2692841

View Full Paper

Ask AI

Bookmark

August 15, 2026Connection ScienceOpen Access

Enhanced CAD patient length of stay prediction using the Graph Neuro Boost model

View Full Paper

Ask AI

Bookmark

Authors

GMGeetha Pratyusha Miriyala ASArun Kumar Sinha

Discussion

Member takes

Overview

Machine learning study demonstrates 98.57% accuracy in predicting hospital length of stay for coronary artery disease patients, highlighting potential for improved clinical resource management.

Key Points

To develop and evaluate an accurate, interpretable machine learning model (Graph Neuro Boost) to predict hospital length of stay for patients diagnosed with coronary artery disease.
Extracted and preprocessed patient records from the MIMIC-III clinical database.
Developed a Graph Neuro Feature Selector that combined autoencoder-based embedded methods with graph theory and Kruskal's algorithm to identify key predictive variables.
Trained an XGBoost model optimized via Bayesian optimization with balanced Expected Improvement and Upper Confidence Bound, utilizing SHAP analysis to assess feature importance.
The Graph Neuro Feature Selector identified 16 critical clinical features in 2.68 seconds.
Graph Neuro Boost achieved an overall prediction accuracy of 98.57%, demonstrating superior performance compared to baseline models across multiple evaluation metrics.

Cite This Study

Miriyala et al. (2026) studied this question.

synapsesocial.com/papers/6a80199875c2e31742c85895 https://doi.org/10.1080/09540091.2026.2692841

View Full Paper

Ask AI

Bookmark