Diagnosis of vertebral column pathologies using concatenated resampling with machine learning algorithms.

Area of Science:

• Bioinformatics
• Medical Informatics
• Computational Biology

Background:

• Machine learning (ML) is crucial for automated medical diagnosis.
• Existing methods for classifying biomedical attributes face challenges with data imbalance.

Purpose of the Study:

• To propose and evaluate a novel data resampling technique, concatenated resampling (CR).
• To enhance the efficacy of traditional ML algorithms for medical diagnosis, specifically inter-vertebral pathologies.

Main Methods:

• Developed and applied the concatenated resampling (CR) technique to an unbalanced training dataset.
• Evaluated four ML approaches: tree-based ensembles and linear models.
• Compared CR with undersampling and over-sampling techniques.
• Utilized metrics including accuracy, precision, recall, and F1 score for performance analysis.

Main Results:

• The extra tree classifier achieved a 0.99 accuracy when combined with the CR technique.
• Comparative analysis identified the best performing classifier and resampling method.
• CR demonstrated superior performance in improving classification accuracy on imbalanced datasets.

Conclusions:

• The proposed concatenated resampling (CR) technique significantly enhances ML model performance for medical diagnosis.
• CR is effective in addressing data imbalance issues in biomedical datasets.
• The extra tree classifier with CR shows exceptional promise for accurate inter-vertebral pathology diagnosis.