Modeling Hepatitis C Transmission Dynamics Using Fractional Order and PINNs
Overview
Hepatitis C virus (HCV) remains a significant global health challenge, particularly in low-income countries like Zimbabwe, where diagnosis and treatment rates are low. This study employs fractional-order mathematical models with Caputo–Fabrizio derivatives and physics-informed neural networks (PINNs) to better capture the complex memory-dependent dynamics of HCV transmission and treatment response.
Background
HCV infects approximately 58 million people worldwide, causing serious liver diseases such as cirrhosis and hepatocellular carcinoma. Despite the availability of highly effective directly acting antivirals (DAAs), access to diagnosis and treatment remains limited in many low- and middle-income countries, including Zimbabwe. The World Health Organization aims to eliminate HCV as a public health threat by 2030, but current treatment rates are far below targets. Modeling approaches that incorporate biological memory effects in disease progression and treatment response can improve understanding and forecasting of HCV dynamics.
Data Highlights
Key epidemiological data include: 58 million global HCV cases; 8 million in Africa; 67,000 estimated chronic HCV cases in Zimbabwe in 2022; only 16% diagnosed and 0.69% treated in Zimbabwe; 12-week generic DAA treatment costs approximately USD 1400 versus average monthly household income of USD 370; global HCV elimination requires increasing treatment from 750,000 to 7.2 million annually.
Key Findings
Fractional-order derivatives, specifically the Caputo–Fabrizio operator, effectively model the non-Markovian memory and hereditary properties of HCV infection and treatment dynamics.
The exponential kernel of the Caputo–Fabrizio derivative aligns with the rapid decay of biological memory in acute HCV viral kinetics and pharmacodynamics of DAAs.
Physics-informed neural networks (PINNs) can integrate fractional-order models with clinical data to improve prediction accuracy of HCV transmission and treatment outcomes.
Zimbabwe faces significant barriers to HCV elimination, including low diagnosis and treatment rates, high treatment costs relative to income, and social factors such as stigma and healthcare access.
Micro-elimination strategies targeting specific populations or regions may be necessary to achieve WHO 2030 elimination goals in resource-limited settings.
Clinical Implications
Incorporating fractional-order models with PINNs provides a more realistic representation of HCV disease progression and treatment response, which can inform public health strategies. Understanding the memory-dependent dynamics of HCV can help optimize timing and targeting of interventions. Addressing barriers to diagnosis and affordable treatment access remains critical for achieving elimination targets, especially in low-income countries like Zimbabwe.
Conclusion
Fractional-order modeling combined with advanced computational techniques offers valuable insights into the complex dynamics of HCV transmission and treatment. These approaches can support more effective planning and implementation of elimination strategies in settings with significant epidemiological and socioeconomic challenges.
References
WHO Global Health Sector Strategy on Viral Hepatitis 2016
Razavi-Shearer D et al. 2018 -- Global prevalence, treatment, and prevention of hepatitis B virus infection
