AI-Powered Portable Optical Biosensors for Environmental Toxicants and Biomarkers

Abstract

Detecting environmental toxicants and biomarkers is vital for protecting ecosystems and public health, yet current methods often lack portability, speed, and precision for field and point- of-care applications. AI-powered portable optical biosensors address these challenges by combining molecular detection units with compact optical devices and advanced data analytics. These systems offer high sensitivity, specificity, portability, and rapid decision-making.This dissertation focuses on developing AI-powered biosensor platforms for environmental field testing and medical diagnostics. Four platforms were designed to detect microRNAs, protein biomarkers, micro/nanoplastics (MNPs), and per- and polyfluoroalkyl substances (PFAS). The detection relies on light scattering or intensity measurements captured by compact photodetectors or smartphone cameras, analyzed using AI frameworks like computer vision and machine learning for pattern recognition, contaminant level prediction, and scalability via cloud processing. Chapter 2 outlines the targets, detection methods, and data analysis techniques across the projects. Overall, this work highlights the transformative potential of AI-powered biosensors in advancing environmental monitoring, healthcare, and industry, with future research aiming to enhance detection capabilities and expand applications.