Helicobacter pylori (H. pylori) risk factor analysis and prevalence prediction: a machine learning-based approach.

BACKGROUND: Although previous epidemiological studies have examined the potential risk factors that increase the likelihood of acquiring Helicobacter pylori infections, most of these analyses have utilized conventional statistical models, including logistic regression, and have not benefited from advanced machine learning techniques. OBJECTIVE: We examined H. pylori infection risk factors among school children using machine learning algorithms to identify important risk factors as well as to determine whether machine learning can be used to predict H. pylori infection status. METHODS: We applied feature selection and classification algorithms to data from a school-based cross-sectional survey in Ethiopia. The data set included 954 school children with 27 sociodemographic and lifestyle variables. We conducted five runs of tenfold cross-validation on the data. We combined the results of these runs for each combination of feature selection (e.g., Information Gain) and classification (e.g., Support Vector Machines) algorithms. RESULTS: The XGBoost classifier had the highest accuracy in predicting H. pylori infection status with an accuracy of 77%-a 13% improvement from the baseline accuracy of guessing the most frequent class (64% of the samples were H. Pylori negative.) K-Nearest Neighbors showed the worst performance across all classifiers. A similar performance was observed using the F1-score and area under the receiver operating curve (AUROC) classifier evaluation metrics. Among all features, place of residence (with urban residence increasing risk) was the most common risk factor for H. pylori infection, regardless of the feature selection method choice. Additionally, our machine learning algorithms identified other important risk factors for H. pylori infection, such as; electricity usage in the home, toilet type, and waste disposal location. Using a 75% cutoff for robustness, machine learning identified five of the eight significant features found by traditional multivariate logistic regression. However, when a lower robustness threshold is used, machine learning approaches identified more H. pylori risk factors than multivariate logistic regression and suggested risk factors not detected by logistic regression. CONCLUSION: This study provides evidence that machine learning approaches are positioned to uncover H. pylori infection risk factors and predict H. pylori infection status. These approaches identify similar risk factors and predict infection with comparable accuracy to logistic regression, thus they could be used as an alternative method.

Machine learning-based risk factor analysis and prevalence prediction of intestinal parasitic infections using epidemiological survey data.

BACKGROUND: Previous epidemiological studies have examined the prevalence and risk factors for a variety of parasitic illnesses, including protozoan and soil-transmitted helminth (STH, e.g., hookworms and roundworms) infections. Despite advancements in machine learning for data analysis, the majority of these studies use traditional logistic regression to identify significant risk factors. METHODS: In this study, we used data from a survey of 54 risk factors for intestinal parasitosis in 954 Ethiopian school children. We investigated whether machine learning approaches can supplement traditional logistic regression in identifying intestinal parasite infection risk factors. We used feature selection methods such as InfoGain (IG), ReliefF (ReF), Joint Mutual Information (JMI), and Minimum Redundancy Maximum Relevance (MRMR). Additionally, we predicted children's parasitic infection status using classifiers such as Logistic Regression (LR), Support Vector Machines (SVM), Random Forests (RF) and XGBoost (XGB), and compared their accuracy and area under the receiver operating characteristic curve (AUROC) scores. For optimal model training, we performed tenfold cross-validation and tuned the classifier hyperparameters. We balanced our dataset using the Synthetic Minority Oversampling (SMOTE) method. Additionally, we used association rule learning to establish a link between risk factors and parasitic infections. KEY FINDINGS: Our study demonstrated that machine learning could be used in conjunction with logistic regression. Using machine learning, we developed models that accurately predicted four parasitic infections: any parasitic infection at 79.9% accuracy, helminth infection at 84.9%, any STH infection at 95.9%, and protozoan infection at 94.2%. The Random Forests (RF) and Support Vector Machines (SVM) classifiers achieved the highest accuracy when top 20 risk factors were considered using Joint Mutual Information (JMI) or all features were used. The best predictors of infection were socioeconomic, demographic, and hematological characteristics. CONCLUSIONS: We demonstrated that feature selection and association rule learning are useful strategies for detecting risk factors for parasite infection. Additionally, we showed that advanced classifiers might be utilized to predict children's parasitic infection status. When combined with standard logistic regression models, machine learning techniques can identify novel risk factors and predict infection risk.

Concurrent infection of intestinal parasites and Helicobacter pylori among school-age children in Central Ethiopia.

BACKGROUND: Coinfection of multiple intestinal microbial pathogens plays an important role in individuals harboring these organisms. However, data on magnitude and risk factors are scarce from resource limited settings. OBJECTIVE: We examined the prevalence and associated risk factors of intestinal parasites and Helicobacter pylori co-infection among young Ethiopian school children. METHOD: Data from a total of 434 Ethiopian school children from the Ziway region were analyzed in the study. Stool antigen and blood serum antibody tests were used to detect H. pylori, while the presence of any intestinal parasites was detected using direct wet mount microscopy and formol-ether concentration techniques. A structured questionnaire was delivered to mothers and legal guardians of the children by an interviewer to collect data relevant demographic and lifestyle factors. Multivariate logistic regression analysis was performed to assess the association of these sociodemographic characteristics with the coinfection of H. pylori and intestinal parasites. RESULTS: The prevalence of coinfection with any intestinal parasites and Helicobacter pylori was 23.0% (n = 92/400). Univariate analysis showed an increased risk for co-infection among children whose mothers had non-formal education (COR: 1.917, p < 0.01) and those who had no history of child vaccination (COR: 3.455, p = 0.084). Children aged 10-14 and those who lived in a house that had a flush or ventilated latrine were found at lower odds of coinfection between intestinal parasites and Helicobacter pylori (COR: 0.670, p = 0.382; COR: 0.189, p = 0.108). Multivariate regression analysis showed increased odds of co-infection among children whose mothers had non-formal education (AOR: 1.978, p < 0.01). Maternal education was also associated with a two-fold increase in odds for H. pylori and any protozoa co-infection (AOR: 2.047, p < 0.01). CONCLUSION: Our study shows a moderate prevalence of H. pylori and intestinal parasite co-infection and identified maternal education as a significant risk factor among school children.

Individual and household correlates of Helicobacter pylori infection among Young Ethiopian children in Ziway, Central Ethiopia.

BACKGROUND: Investigating distinct individual- and household-level risk factors for acquiring Helicobacter pylori (H. pylori) infection can inform disease prevention efforts and implicate possible routes of transmission. This study determined the magnitude of H. pylori infection among schoolchildren in Ziway, central Ethiopia and identified personal and household correlates of H. pylori infection in young Ethiopian children. METHODS: A total of 434 schoolchildren participated in this cross-sectional study. Infection status was assessed using antigen and antibody rapid tests. Demographic and lifestyle information was obtained from parents via an interviewer-led questionnaire. Univariate and multivariate logistic regressions were performed to assess the relationships between potential individual- and household-level risk factors and H. pylori infection. RESULTS: The prevalence of H. pylori infection was 65.7% (285/434). Of the personal variables assessed, the age group 10-14 years was found to be significantly associated with higher odds of H. pylori infection in univariate analysis (COR = 2.22, 95% CI: 1.06-4.66, p = 0.03) and remained positively correlated after adjusting for confounding factors. Of the household-level factors explored, having a traditional pit or no toilet was found to be significantly associated with 3.93-fold higher odds of H. pylori infection (AOR = 3.93, 95% CI: 1.51-10.3, p = 0.01), while the presence of smokers in the household was associated with 68% lower odds of infection (AOR = 0.32, 95% CI: 0.11-0.89, p = 0.03). CONCLUSION: This study from a developing country provides additional evidence for older age as a personal risk factor for H. pylori infection and identifies correlations between socioeconomic and sanitation household factors and positive childhood infection status. The associations reported here support the hypothesized fecal-oralroute of transmission for H. pylori.

Association between infection with Helicobacter pylori and platelet indices among school-aged children in central Ethiopia: a cross-sectional study.

OBJECTIVE: Previous clinical studies in adults from developed countries have implicated Helicobacter pylori infections in the development of thrombocytopenia. However, studies in children, particularly those from low-income countries, are unusually scarce. We examined the association between H. pylori infection and platelet indices in young Ethiopian school children. DESIGN: Cross-sectional study SETTING: This study was conducted in five elementary schools located in central Ethiopia. PARTICIPANTS: Blood and stool samples were collected from 971 children across five elementary schools in Ethiopia. H. pylori infection was diagnosed using stool antigen and serum antibody tests, and haematological parameters were measured using an automated haematological analyser. An interviewer-led questionnaire administered to mothers provided information on demographic and lifestyle variables. The independent effects of H. pylori infection on platelet indices were determined using multivariate linear and logistic regressions. STUDY OUTCOMES: H. pylori-infected children had a lower average platelet count and mean platelet volume than uninfected after adjusting the potential confounders (adjusted mean difference: -20.80×109/L; 95% CI -33.51 to -8.09×109, p=0.001 and adjusted mean difference: -0.236 fL; 95% CI -0.408 to -0.065, p=0.007, respectively). Additionally, H. pylori-infected children had lower red blood cell counts (adjusted mean difference: -0.118×1012/L; 95% CI -0.200 to -0.036, p=0.005) compared with non-infected. CONCLUSION: Our study from a developing country provides further support for an association between H. pylori infections and reduced platelet indices in young Ethiopian school children, after controlling for potential confounders. Further research is needed, particularly longitudinal studies, to establish causality.