Prediction of Multidrug-Resistant Tuberculosis Using Machine Learning Algorithms in SWAT, Pakistan

Abstract

In this paper, we have focused on machine learning (ML) feature selection (FS) algorithms for identifying and diagnosing multidrug-resistant (MDR) tuberculosis (TB). MDR-TB is a universal public health problem, and its early detection has been one of the burning issues. The present study has been conducted in the Malakand Division of Khyber Pakhtunkhwa, Pakistan, to further add to the knowledge on the disease and to deal with the issues of identification and early detection of MDR-TB by ML algorithms. These models also identify the most important factors causing MDR-TB infection whose study gives additional insights into the matter. ML algorithms such as random forest, k-nearest neighbors, support vector machine, logistic regression, leaset absolute shrinkage and selection operator (LASSO), artificial neural networks (ANNs), and decision trees are applied to analyse the case-control dataset. This study reveals that close contacts of MDR-TB patients, smoking, depression, previous TB history, improper treatment, and interruption in first-line TB treatment have a great impact on the status of MDR. Accordingly, weight loss, chest pain, hemoptysis, and fatigue are important symptoms. Based on accuracy, sensitivity, and specificity, SVM and RF are the suggested models to be used for patients’ classifications.

Tuberculosis Disease Diagnosis Based on an Optimized Machine Learning Model

“…GAs are stochastic search techniques, based on natural genetics, which provide robust search capabilities in complex spaces. A GA is an iterative process that solves an optimization problem [ 44 ]. Each solution is obtained by means of an encoding/decoding mechanism, which requires us to represent the solution as a chromosome.…”

Machine learning approaches in the diagnosis of infectious diseases: a review

“…The traditional modeling methods are not suitable for data with missing values, irregular spacing, and high-dimensional features. The majority of the current ML models are developed and evaluated either on artificially generated data or are experimental [11], [22], [43], [45], [55], [56]. In many cases, these models perform better than existing methods based on evaluation metrics such as accuracy and precision.…”

A machine learning model to explore individual risk factors for tuberculosis treatment non-adherence in Mukono district

“…Of the chronic diseases, TB, due to its infectious nature, increasing antimicrobial resistance and prevalence in low-and-middle-income (LMIC) countries, for example, India (32,48), Pakistan (49), Morocco (50), Kenya (51), and Uganda (13) has been widely studied. Characteristically, these previous studies gathered demographic and medical histories of a cohort and observed their adherence and outcomes.…”

“…The researchers then retrospectively applied ML algorithms e.g. random forest, support vector machine (48,49), survival analysis (32,51), and logistic regression (13,50) to determine variables predictive of treatment failure or non-adherence.…”

“…Newer work has explored the utility of advanced machine learning techniques such as support vector machines (SVM), Artificial Neural Networks (ANN), and more for improved classification accuracy. A study by Mian et al, (49) out that the two best-performing algorithms were; SVM and RF with an accuracy of 78% and 74% respectively for patients' classification. Methodologically, our work was related to other studies predicting outcomes by inferring a model being trained by a set of historical data (32,56,57).…”