Cerebral malaria, caused predominantly by Plasmodium falciparum, remains a leading cause of long-term neurological disability in children within endemic regions. Even with treatment, it may lead to cognitive delays, motor/speech difficulties, or memory or learning disabilities. In this study, we present a novel nonlinear compartmental model for malaria transmission based on the SITRM framework, which incorporates critical clinical features such as treatment failure, reinfection, and awareness-driven behavioral changes. We derive the basic reproduction number R0, analyze the local and global stability of the disease-free equilibrium, and examine the endemic behavior of the system under varying transmission conditions. To address the numerical challenges posed by the model’s nonlinearities, we employ an Artificial Neural Network (ANN) trained via the Levenberg-Marquardt Backpropagation (LMB) algorithm. This ANN-LMB framework is trained on high-fidelity solutions generated using the classical fourth-order Runge-Kutta method and is designed to replicate both disease-free and endemic equilibrium states with high accuracy. The proposed model achieves remarkably low mean squared errors, with <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12190_2025_2615_Article_IEq1.gif" Format="GIF" Height="17" Rendition="HTML" Resolution="72" Type="Linedraw" Width="104" /> </InlineMediaObject> <EquationSource Format="TEX">\(2.3255 \times 10^{-10}\)</EquationSource> </InlineEquation> in the disease-free scenario (<InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12190_2025_2615_Article_IEq2.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="51" /> </InlineMediaObject> <EquationSource Format="TEX">\(R_0 < 1\)</EquationSource> </InlineEquation>) and <InlineEquation ID="IEq3"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12190_2025_2615_Article_IEq3.gif" Format="GIF" Height="17" Rendition="HTML" Resolution="72" Type="Linedraw" Width="104" /> </InlineMediaObject> <EquationSource Format="TEX">\(1.8918 \times 10^{-11}\)</EquationSource> </InlineEquation> in the endemic case (<InlineEquation ID="IEq4"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12190_2025_2615_Article_IEq4.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="51" /> </InlineMediaObject> <EquationSource Format="TEX">\(R_0 > 1\)</EquationSource> </InlineEquation>), with regression coefficients (R ≈ 1) in both, demonstrating near-perfect predictive performance. Compared to conventional solvers, the ANN-based approach offers enhanced computational efficiency and stability, even in the presence of parameter-induced stiffness. Importantly, the inclusion of media-driven awareness and neuro-disability dynamics adds new layers of realism to malaria modeling. The findings validate the ANN-LMB model as a powerful and novel tool for simulating malaria epidemiology with neurological outcomes, providing actionable insights for targeted interventions and public health planning in high-burden regions.

Nonlinear modeling of cerebral malaria transmission with neuro-disability via ANN-LMB enhanced SITRM model

期刊论文
发表日期：

2025年8月21日

查看全文

Rahat Zarin,
Kamel Guedri,
Basim M. Makhdoum,
Hatoon A. Niyazi,
Hamiden Abd El-Wahed Khalifa

摘要

Cerebral malaria, caused predominantly by Plasmodium falciparum, remains a leading cause of long-term neurological disability in children within endemic regions. Even with treatment, it may lead to cognitive delays, motor/speech difficulties, or memory or learning disabilities. In this study, we present a novel nonlinear compartmental model for malaria transmission based on the SITRM framework, which incorporates critical clinical features such as treatment failure, reinfection, and awareness-driven behavioral changes. We derive the basic reproduction number R₀, analyze the local and global stability of the disease-free equilibrium, and examine the endemic behavior of the system under varying transmission conditions. To address the numerical challenges posed by the model’s nonlinearities, we employ an Artificial Neural Network (ANN) trained via the Levenberg-Marquardt Backpropagation (LMB) algorithm. This ANN-LMB framework is trained on high-fidelity solutions generated using the classical fourth-order Runge-Kutta method and is designed to replicate both disease-free and endemic equilibrium states with high accuracy. The proposed model achieves remarkably low mean squared errors, with \(2.3255 \times 10^{-10}\) in the disease-free scenario ( \(R_0 < 1\) ) and \(1.8918 \times 10^{-11}\) in the endemic case ( \(R_0 > 1\) ), with regression coefficients (R ≈ 1) in both, demonstrating near-perfect predictive performance. Compared to conventional solvers, the ANN-based approach offers enhanced computational efficiency and stability, even in the presence of parameter-induced stiffness. Importantly, the inclusion of media-driven awareness and neuro-disability dynamics adds new layers of realism to malaria modeling. The findings validate the ANN-LMB model as a powerful and novel tool for simulating malaria epidemiology with neurological outcomes, providing actionable insights for targeted interventions and public health planning in high-burden regions.