Publications | Md. Abu Bokkor Shiddik

BMC MedicineSpringer2026Open Access

Spatiotemporal patterns of climate-sensitive vector-borne diseases in Bangladesh: leveraging machine learning and spatial regression for intervention strategies

Md. Siddikur Rahman, Md. Abu Bokkor Shiddik, Arman Hossain Chowdhury

64

Districts

2017–20

Data Span

β 16.64

Temp. Effect

5 Diseases

VBD Coverage

XGBoost

Best Model

↗ DOI

Rahman, M., Shiddik, M.A.B. & Chowdhury, A.H. Spatiotemporal patterns of climate-sensitive vector-borne diseases in Bangladesh: leveraging machine learning and spatial regression for intervention strategies. BMC Med (2026). https://doi.org/10.1186/s12916-026-04857-1

Abstract

Background: Bangladesh has seen a noticeable rise in vector-borne diseases (VBD) attributed to climate change. Accurately mapping, predicting, and identifying risk factors of VBD is essential for prevention and control strategies.

Methods: District-level reported cases of VBD (dengue, chikungunya, malaria, filariasis, and yellow fever) were obtained from the Bangladesh Bureau of Statistics alongside meteorological data from NASA (2017–2020). Exploratory spatial analysis, spatial regression, and tree-based ML models were applied for prediction, risk factor identification, and correlation analysis.

Results: Dengue was the most prevalent VBD, peaking in 2019. Dhaka, Pirojpur, Jessore, Bandarban, Rangamati, and Narail exhibited higher incidence rates. Spatial regression identified mean temperature as a major risk factor (β = 16.64, s.e. = 6.39). XGBoost also highlighted mean temperature as the primary determinant, alongside GDP, population size, and healthcare infrastructure.

Conclusions: Findings incorporating the One Health perspective provide insights for early-warning, prevention, and control strategies using ML to combat infectious diseases in Bangladesh and similar endemic countries.

Figures

Health Science ReportsWiley2026Open Access

Global Prediction of Dengue Incidence Using an Explainable Artificial Intelligence-Driven ConvLSTM Integrating Environmental, Health, and Socio-Economic Determinants

Md. Abu Bokkor Shiddik

118

Countries

2000–21

Data Span

R² 0.77

ConvLSTM Fit

SHAP+IG+LRP

XAI Methods

ConvLSTM

Best Model

↗ DOI View in Journal

Shiddik, M.A.B. (2026). Global Prediction of Dengue Incidence Using an Explainable Artificial Intelligence-Driven ConvLSTM Integrating Environmental, Health, and Socio-Economic Determinants. Health Science Reports. https://doi.org/10.1002/hsr2.72280

Abstract

Background: Dengue fever is a rapidly expanding vector-borne disease posing significant global epidemiological challenges. Accurate and interpretable forecasting is essential for timely interventions, yet most models overlook spatiotemporal, sex-specific, and country-level heterogeneity in disease dynamics.

Methods: A ConvLSTM network was applied to predict dengue incidence across 118 countries from 2000 to 2021 using 20 climatic, environmental, health system, and socio-economic predictors. Feature contributions were assessed through SHAP values, permutation importance, perturbation sensitivity, integrated gradients (IG), and layer-wise relevance propagation (LRP).

Results: ConvLSTM achieved the best predictive performance (R² = 0.7731). Annual freshwater withdrawals dominated male dengue incidence (SHAP: 44.37%), while hospital bed density had greater influence for females (SHAP: 31.86%). Temperature anomalies contributed consistently to both sexes (SHAP: 11.51%). Sensitivity analysis showed dengue incidence varied from −65% to +91% under ±50% predictor perturbations.

Conclusion: The ConvLSTM–XAI framework provides transparent, sex-aware, and country-specific dengue forecasts supporting targeted, climate-resilient dengue control strategies.

Figures

Health Science ReportsWiley2026Open Access

Explainable Artificial Intelligence in Healthcare: Current Landscape, Challenges, and Future Directions

Md. Abu Bokkor Shiddik

70

Studies Reviewed

2017–25

Data Span

SHAP 54%

Top XAI Method

6 Databases

Sources

Systematic Review

Study Design

↗ DOI View in Journal

Shiddik, M.A.B. (2026). Explainable Artificial Intelligence in Healthcare: Current Landscape, Challenges, and Future Directions. Health Science Reports. https://doi.org/10.1002/hsr2.72172

Abstract

Background: AI, particularly ML and Deep Learning, is transforming healthcare by enabling improved diagnosis, prognosis, and personalized treatments. However, the opacity of many AI "black box" models limits interpretability, clinician trust, and real-world adoption. XAI has emerged to address these limitations by providing transparent and actionable insights. This systematic review synthesizes current evidence on XAI in healthcare, mapping AI models to XAI techniques, domains, and clinical applications.

Methods: A systematic search was conducted across six databases (Elsevier, Springer, Taylor & Francis, Semantic Scholar, ACM, and IEEE Xplore) for peer-reviewed articles published between 2017 and 2025, following PRISMA guidelines.

Results: Seventy studies were included, spanning oncology (40%), cardiology (21%), infectious diseases (14%), neurology (11%), and clinical decision support (13%). Deep learning models were most frequently applied (76%). SHAP (54%) and LIME (30%) were the most commonly used XAI techniques. Only 12 studies explicitly addressed ethical or regulatory considerations.

Conclusion: XAI enhances transparency, clinician trust, and decision-making in healthcare AI. Future work should focus on hybrid, clinically validated XAI models, comprehensive ethical compliance, and user-centered implementations.

Figures

Scientific ReportsNature Portfolio2025Open Access

Utilizing artificial intelligence to predict and analyze socioeconomic, environmental, and healthcare factors driving tuberculosis globally

Md. Siddikur Rahman, Abu Bokkor Shiddik

194

Countries

2000–22

Data Span

SHAP 0.874

Peak Impact

R² 0.67

Model Fit

XGBoost

Best Model

↗ DOI View in Journal

Rahman, M.S. & Shiddik, A.B. (2025). Utilizing artificial intelligence to predict and analyze socioeconomic, environmental, and healthcare factors driving tuberculosis globally. Scientific Reports. https://doi.org/10.1038/s41598-025-96973-w

Abstract

Tuberculosis (TB) is a major global health issue, contributing significantly to mortality and morbidity rates worldwide. This study employed the advanced machine learning model XGBoost, combined with XAI and spatial analysis, to describe global TB incidence and mortality rates across 194 countries from 2000 to 2022. Spatial autocorrelation analysis utilizing Moran's I revealed geographical clusters and significant determinants affecting TB incidence. Treatment success rates and MDR-TB treatment initiation were identified as pivotal determinants. Confirmed cases of MDR-TB had the most significant impact on TB incidence (SHAP = 0.874). Additionally, air pollution had a notable impact (SHAP = 1.36). The XGBoost model demonstrated the best predictive performance (RMSE = 0.88, R² = 0.67, Adjusted R² = 0.65). This study highlights the potential of integrating XGBoost and XAI methodologies as a holistic framework for effectively tackling global tuberculosis incidence and mortality rates.

Figures

Health Science ReportsWiley2025Open Access

Dengue Early Warning System and Outbreak Prediction Tool in Bangladesh Using Interpretable Tree-Based Machine Learning Model

Md. Siddikur Rahman, Miftahuzzannat Amrin, Md. Abu Bokkor Shiddik

LightGBM

Best Model

2000–21

Data Span

27°C

Optimal Mean Temp

82%

Risk Humidity

SHAP

XAI Method

↗ DOI View in Journal

Rahman, M.S., Amrin, M. & Shiddik, M.A.B. (2025). Dengue Early Warning System and Outbreak Prediction Tool in Bangladesh Using Interpretable Tree-Based Machine Learning Model. Health Science Reports, 8(5), e70726. https://doi.org/10.1002/hsr2.70726

Abstract

Background: A life-threatening vector-borne disease, dengue fever poses significant global public health and economic threats, including Bangladesh. This study proposes an interpretable tree-based ML model for dengue early warning systems using climatic, sociodemographic, and landscape factors.

Methods: High-performance ML algorithms — Random Forests, XGBoost, and LightGBM — were applied to surveillance data from January 2000 to December 2021. SHAP values identified the most significant dengue drivers.

Results: Optimal thresholds: mean temperature 27°C, minimum temperature 22°C, humidity 82%. LightGBM accurately forecasts dengue, with agricultural land, population density, and minimum temperature as key outbreak drivers.

Conclusion: The proposed ML model functions as an early warning system, providing a framework for sophisticated analytical techniques in public health.

Figures

Global EpidemiologyElsevier2025Open Access

Explainable artificial intelligence for predicting dengue outbreaks in Bangladesh using eco-climatic triggers

Md. Siddikur Rahman, Md. Abu Bokkor Shiddik

AUC 0.89

Performance

2000–23

Data Span

SHAP+LIME

XAI Methods

2024–30

Forecast

XGBoost

Best Model

↗ DOI View in Journal

Rahman, M.S. & Shiddik, M.A.B. (2025). Explainable artificial intelligence for predicting dengue outbreaks in Bangladesh using eco-climatic triggers. Global Epidemiology, 10, 100210. https://doi.org/10.1016/j.gloepi.2025.100210

Abstract

Background: Dengue represents a significant public health threat in Bangladesh, characterized by complex ecological transmission dynamics. This study employs state-of-the-art AI methods to identify eco-climatic factors in predicting dengue outbreaks.

Methods: XGBoost and LightGBM combined with XAI (SHAP and LIME) evaluated predictive performance of dengue determinants from 2000 to 2023. Models also predicted dengue cases and early warning trends from 2024 to 2030.

Findings: Bangladesh experienced the highest dengue cases in August, most fatalities in November. XGBoost achieved AUC = 0.89 and Log Loss = 0.64. Key predictors: population density, precipitation, temperature, and land-use patterns.

Interpretation: This study showcases XAI's potential in uncovering dengue outbreak complexities, providing a robust tool for public health interventions and generating timely alerts to prevent outbreaks.

Figures

Global EpidemiologyElsevier2025Open Access

Reflections on explainable artificial intelligence for predicting dengue outbreaks in Bangladesh

Md. Siddikur Rahman, Md. Abu Bokkor Shiddik

↗ DOI View in Journal

Rahman, M.S. & Shiddik, M.A.B. (2025). Reflections on explainable artificial intelligence for predicting dengue outbreaks in Bangladesh. Global Epidemiology, p.100230. https://doi.org/10.1016/j.gloepi.2025.100230

Abstract

This paper reflects on XAI's vast implications in public health research and its potential to inform climate-sensitive disease prediction in low-income settings. A novel framework combining predictive accuracy with interpretability through XAI enables actionable insights for early warning systems (EWS) and dengue prevention in densely populated and climate-sensitive regions like Bangladesh.

Figures

Scientific ReportsNature Portfolio2025Open Access

Unraveling global malaria incidence and mortality using machine learning and artificial intelligence–driven spatial analysis

Md. Siddikur Rahman, Md. Abu Bokkor Shiddik

106

Countries

2000–22

Data Span

R² 0.93

Model Fit

RMSE 0.63

Error

XGBoost

Best Model

↗ DOI View in Journal

Rahman, M.S. & Shiddik, M.A.B. (2025). Unraveling global malaria incidence and mortality using machine learning and AI-driven spatial analysis. Scientific Reports, 15(1), 28334. https://doi.org/10.1038/s41598-025-12872-0

Abstract

This study utilized spatial analysis, advanced ML, and XAI to identify high-risk areas, uncover key determinants, and predict malaria outcomes across 106 countries (2000–2022). XGBoost combined with XAI and causal AI techniques was employed. Spatial analyses (Getis-Ord Gi* and Moran's I) detected significant geographical clusters. In 2022, malaria cases reached 251.75 million; Nigeria recorded the highest incidence (1,332.99 million cases). The XGBoost model demonstrated best predictive performance (RMSE = 0.63, R² = 0.93). Key determinants included access to basic sanitation, electricity availability, population growth, and under-5 mortality rate.

Figures

Int. J. Statistical SciencesBanglaJOL2025

Data-Driven Dengue Prevention Strategies in Bangladesh using Explainable Artificial Intelligence and Causal Inference

Md. Siddikur Rahman, Md. Abu Bokkor Shiddik

2010–24

Data Span

41.87%

Climate Factor

DoWhy

Causal Framework

XGBoost

Best Model

↗ DOI View in Journal

Rahman, M.S. & Shiddik, M.A.B. (2025). Data-Driven Dengue Prevention Strategies in Bangladesh using Explainable Artificial Intelligence and Causal Inference. Int. J. Statistical Sciences, 25(2), 101-113. https://doi.org/10.3329/ijss.v25i2.85772

Abstract

This study employed high-performance ML (XGBoost) combined with XAI and causal inference (DoWhy framework) to evaluate dengue incidence in Bangladesh from January 2010 to December 2024. Dengue cases exceeded 100,000 total in 2019, 2023, and 2024, with monthly peaks of 50,000. The XGBoost model was the top performer (RMSE: 11,365.2; MAE: 7,014.68). Climate indicators were the strongest contributor to dengue prediction (41.87%), followed by sociodemographic (35.03%), healthcare (19.16%), and landscape factors (3.95%).

Figures

Int. J. Health GeographicsSpringer2025Open Access

Leveraging explainable artificial intelligence and spatial analysis for communicable diseases in Asia (2000–2022) based on health, climate, and socioeconomic factors

Md. Siddikur Rahman, Md. Abu Bokkor Shiddik

41

Countries

2000–22

Data Span

R² 0.91

TB Model Fit

Gi*+Moran

Spatial Methods

XGBoost

Best Model

↗ DOI View in Journal

Rahman, M.S. & Shiddik, M.A.B. (2025). Leveraging explainable artificial intelligence and spatial analysis for communicable diseases in Asia (2000-2022). Int. J. Health Geographics. https://doi.org/10.1186/s12942-025-00433-7

Abstract

Background: Communicable diseases remain a significant challenge in Asia. This study applies ML and spatial analysis to examine patterns and determinants across 41 countries (2000–2022).

Results: TB cases declined from 8.01M (2000) to 7.54M (2022), with hotspots in India (Gi* = 3.07) and Nepal (Gi* = 4.67). Malaria dropped from 27.00M to 7.96M, yet Bangladesh (Gi* = 4.13) and Pakistan (Gi* = 4.17) exhibited sustained risk. Dengue peaked at 2.71M cases in 2019. XGBoost achieved RMSE = 0.94, R² = 0.91 for tuberculosis.

Conclusion: Integrating ML, spatial analysis, and XAI uncovers disease determinants and guides targeted interventions across Asia.

Figures

BMJ OpenBMJ2026Open Access

Bayesian spatiotemporal modelling of neonatal, infant and under-5 mortality (2000–2022) in 41 Asian countries: a population-level observational study

Md. Siddikur Rahman, Md. Abu Bokkor Shiddik

41

Countries

2000–22

Data Span

R² 0.99

Neonatal Model

r = −0.79

Rubella Corr.

Bayesian

Approach

↗ DOI View in Journal

Rahman, M.S. & Shiddik, M.A.B. (2026). Bayesian spatiotemporal modelling of neonatal, infant and under-5 mortality (2000-2022) in 41 Asian countries. BMJ Open, 16(2), e108632. https://doi.org/10.1136/bmjopen-2025-108632

Abstract

Background: Child mortality continues to pose a major public health challenge across Asia. This study examines trends in under-5, infant, and neonatal mortality and identifies key determinants using spatiotemporal modelling across 41 Asian countries (2000–2022).

Results: Under-5 mortality decreased significantly from 46.73 to 18.53 per 1000 live births. Strong negative associations were found with vaccination coverage — rubella (r = −0.79), DTP (r = −0.74). Afghanistan (77.12), Bangladesh (57.12), and Myanmar (63.16) were identified as high-risk hotspots. Predictive accuracy was highest for neonatal mortality (R² = 0.99; RMSE = 1.93).

Conclusion: Immunisation and maternal education are critical to reducing mortality. The white-box modelling framework supports data-driven policy planning toward SDG 3.2.

Figures

Trop. Med. & Infect. Dis.MDPI2026Open Access

District-Level Dengue Early Warning Prediction System in Bangladesh Using Hybrid Explainable AI and Bayesian Deep Learning

Md. Abu Bokkor Shiddik, Farzana Zannat Toshi, Sadia Yesmin, Md. Siddikur Rahman

64

Districts

2017–24

Data Span

Acc 0.93

MLP Accuracy

AUC 0.99

ROC-AUC

SHAP 0.314

Humidity Peak

↗ DOI View in Journal

Shiddik, M.A.B., Toshi, F.Z., Yesmin, S. & Rahman, M.S. (2026). District-Level Dengue Early Warning Prediction System in Bangladesh Using Hybrid Explainable AI and Bayesian Deep Learning. Tropical Medicine and Infectious Disease, 11(3), 73. https://doi.org/10.3390/tropicalmed11030073

Abstract

In Bangladesh, 321,179 dengue cases were reported in 2023, followed by 101,214 cases in 2024. This study developed a district-level dengue early warning system integrating climatic, socio-demographic, economic, healthcare, and environmental determinants across all 64 districts (2017–2024). ML and deep learning approaches — MLP and ConvLSTM — were combined with SHAP-based XAI and Bayesian spatio-temporal models. Climate was the strongest predictor (humidity SHAP = 0.314; minimum temperature SHAP = 0.146). Poverty (SHAP = 0.193) and healthcare capacity also contributed significantly. The MLP model achieved accuracy = 0.93, ROC-AUC = 0.99. The integrated framework delivers transparent, interpretable district-level early warnings for dengue outbreak preparedness.

Figures