Cost-Sensitive Machine Learning Classification for Mass Tuberculosis Verbal Screening

TL;DR

This study demonstrates that cost-sensitive machine learning models, particularly XGBoost with adjusted class weights, outperform traditional clinician-defined score-based methods in TB verbal screening, achieving higher sensitivity and specificity with limited data.

Contribution

The paper introduces a cost-sensitive machine learning approach for TB verbal screening that improves detection accuracy over traditional methods, even with small datasets.

Findings

XGBoost with class weight adjustment achieved 96.64% sensitivity.

Specificity improved by 13.19% over traditional methods.

Only 2000 data points were needed for model convergence.

Abstract

Score-based algorithms for tuberculosis (TB) verbal screening perform poorly, causing misclassification that leads to missed cases and unnecessary costly laboratory tests for false positives. We compared score-based classification defined by clinicians to machine learning classification such as SVM-RBF, logistic regression, and XGBoost. We restricted our analyses to data from adults, the population most affected by TB, and investigated the difference between untuned and unweighted classifiers to the cost-sensitive ones. Predictions were compared with the corresponding GeneXpert MTB/Rif results. After adjusting the weight of the positive class to 40 for XGBoost, we achieved 96.64% sensitivity and 35.06% specificity. As such, the sensitivity of our identifier increased by 1.26% while specificity increased by 13.19% in absolute value compared to the traditional score-based method defined by our clinicians. Our approach further demonstrated that only 2000 data points were sufficient to enable the model to converge. The results indicate that even with limited data we can actually devise a better method to identify TB suspects from verbal screening.