Henry gas solubility optimization double machine learning classifier for neurosurgical patients.

Area of Science:

• Neurosurgery
• Medical Informatics
• Artificial Intelligence

Background:

• Head trauma significantly impacts patient outcomes across all age groups.
• Accurate prediction of head trauma prognosis is crucial for effective neurosurgical patient management.
• Machine learning (ML) offers potential for improving outcome prediction in neurosurgery.

Purpose of the Study:

• To compare various ML algorithms for predicting head trauma patient outcomes.
• To develop and evaluate a novel hybrid ML model for enhanced outcome prediction.
• To investigate the relationship between patient age, trauma mode, and outcome.

Main Methods:

• Comparative analysis of k-nearest neighbor, Random Forest (RF), C4.5, Artificial Neural Network, and Support Vector Machine (SVM).
• Development of a double classifier using Henry Gas Solubility Optimization (HGSO) and Aquila Optimizer (AQO) for feature selection.
• Implementation of a hybrid RF-SVM model integrated with HGSO for outcome classification into mortality, morbidity, improved, or same status.

Main Results:

• The proposed hybrid RF-SVM and HGSO model achieved a high accuracy of 99.2%.
• The model demonstrated superior performance compared to individual ML classifiers.
• The study identified relationships between age, mode of trauma, and patient outcome.

Conclusions:

• The developed hybrid ML model significantly enhances the accuracy of head trauma outcome prediction in neurosurgical patients.
• This approach offers a valuable tool for clinicians to manage neurosurgical care more effectively.
• The findings support the use of advanced ML techniques in personalized neurosurgical treatment strategies.