Agentic AI-gated multiplex electrochemical sensing for food-safety decision-making

Area of Science:

• Analytical Chemistry
• Biosensing Technology
• Artificial Intelligence in Food Safety

Background:

• Field-deployable food safety monitoring necessitates rapid detection of chemical residues.
• Existing methods often lack reliability assessment for critical decision-making.
• Simultaneous detection of multiple contaminants is crucial for comprehensive food safety.

Purpose of the Study:

• To develop an AI-gated multiplex electrochemical sensing platform for reliable food safety monitoring.
• To integrate electrochemical impedance spectroscopy (EIS) with AI for enhanced data interpretation and decision logic.
• To simultaneously detect eight common food residues, including antibiotic metabolites and pesticides.

Main Methods:

• A 16-electrode antibody-functionalized array was utilized for multiplex detection.
• Electrochemical impedance spectroscopy (EIS) responses were converted into interpretable electrochemical features.
• AI gates assessed signal quality, baseline drift, dose-response consistency, and prediction uncertainty.

Main Results:

• The platform achieved high analytical performance with R² ≈ 0.975 and low detection limits (0.001-0.06 ng/mL).
• Recoveries were within the acceptable range of 80-120%, with rapid readout in under 5 minutes.
• AI-gated reporting improved accepted-sample accuracy from 89.1% to 98.7% by managing uncertain measurements.

Conclusions:

• The AI-gated electrochemical sensing platform offers a reliable and rapid solution for field-deployable food safety monitoring.
• Integrating EIS with AI-driven reliability assessment significantly enhances the accuracy and trustworthiness of residue detection.
• This technology provides a robust tool for ensuring food safety by minimizing false-negative or false-positive decisions.