Muniyandi Maruthupandi, N Y Lee
Neurological disorders, diabetes, cancer, and infectious diseases remain major global health concerns, particularly in low- and middle-income countries with insufficient access to accurate and rapid diagnostics. Conventional biochip sensing platforms, while effective, are often constrained by complex instrumentation and have limited capability for handling complex and large datasets. This review aims to address these limitations by evaluating the integration of artificial intelligence (AI) with biochip technology improve biomedical diagnostics. We systematically analyze recent advances in AI-integrated biochips, such as spectroscopic, paper-based, lab-on-chip, and microfluidic platforms integrated with reinforcement learning, machine learning, and deep learning models. These pre-trained AI models simplify pattern recognition, feature extraction, and automated data processing from a variety of biosensor outputs, such as electrochemical, fluorescence, and colorimetric signals. The reviewed studies indicate improved real-time diagnostic sensitivity and accuracy across biomedical applications. Overall, we discuss ongoing challenges and future perspectives toward explainable, robust, and smartphone-assisted AI-integrated biochips for rapid and accurate diagnostics. The review offers a comprehensive overview of AI-integrated biochips to support effective disease detection and clinical decision-making.