Creating an AI-Based Symptom Checker for Low-Resource Healthcare Settings with Explainability Features

Abstract

The rise of AI-driven health technologies presents opportunities for transformative change in underserved healthcare systems. This paper proposes the design and implementation of an AI-based symptom checker tailored to low-resource settings. Emphasizing algorithmic transparency and interpretability, the proposed system integrates lightweight machine learning models with explainability features to aid both clinicians and patients. We assess usability, diagnostic accuracy, and alignment with the constraints of limited digital infrastructure. Through literature synthesis and prototype testing, this paper outlines design parameters and implementation pathways for practical deployment in global health contexts.