AI-Enhanced Cervical Cancer Screening: Integrating automated diagnostics and self-sampling platforms

Abstract

This is a systematic review that summarizes the existing body of literature about the incorporation of artificial intelligence (AI) and automated diagnostic frameworks and scalable self-sampling systems to detect cervical cancer screening. The analysis particularly reviews the AI use in risk stratification and image diagnostics applications and compares the performance metrics, scalability prospects, economic analysis, and practical implementation of the applications. The extensive review of peer-reviewed publications by analyzing articles published during the 10-year interval between January 2014 and June 2024 shows that deep learning architecture consistently demonstrates sensitivity and specificity rates over 90%, which are much higher in controlled research studies than traditional cytology and colposcopy processes 1,37. The integration of patient self-sampling approaches contributes greatly to the screening participation rates, especially when it comes to underserved and resource-restricted settings. Extensive implementation literature shows significant potential in terms of reducing costs and improving access to healthcare services, but major issues still exist when it comes to data heterogeneity, clinical validation processes, and integration frameworks in health systems. The aggregate results highlight the potential sponsorship these AI-based solutions have to achieve in enhancing the global cervical cancer prevention strategies and bring them closer to the potential goals of elimination set by the World Health Organization.