AI-enhanced rapid diagnostic testing platform for mass opisthorchiasis screening

Area of Science:

• Medical Diagnostics
• Artificial Intelligence in Public Health
• Parasitology

Background:

• Cholangiocarcinoma (CCA) is a significant health concern in the Mekong Basin, strongly associated with Opisthorchis viverrini (OV) infection.
• Early detection of OV infection is vital for at-risk populations in endemic areas.
• Current OV rapid diagnostic tests (RDTs) rely on subjective visual interpretation, leading to variability and data digitization challenges.

Purpose of the Study:

• To introduce the OV-RDT platform, a cloud-based AI system for standardized interpretation of OV-RDT results.
• To facilitate efficient mass screening for opisthorchiasis.
• To enhance early detection and public health interventions in OV-endemic regions.

Main Methods:

• Development of a cross-platform (Android/iOS) cloud-based system with a mobile app, intelligent dashboard, and AI/data processing servers.
• Implementation of AI models (EfficientNet-B5) for image quality assessment and T-band grading of OV-RDT images.
• Extensive field validation in northeastern Thailand, processing over 100,000 samples.

Main Results:

• The AI platform demonstrated high accuracy: 98% for image quality assessment and 95% for OV infection grading.
• The system processed over 100,000 samples effectively during field testing in a high-prevalence region.
• Healthcare workers reported high satisfaction (4.41/5.00) with the user-friendly interface and dashboard capabilities.

Conclusions:

• The OV-RDT platform offers a standardized, AI-driven solution for mass screening of opisthorchiasis.
• The technology shows significant promise for improving early detection and public health strategies in Southeast Asia.
• Further external validation is recommended to confirm generalizability across diverse geographical areas.