AI‐Assisted Plasmonic Enhanced Colorimetric Fluidic Device for Hydrogen Peroxide Detection from Cancer Cells

Abstract Hydrogen peroxide (H 2 O 2 ) is an essential molecule to various physiological processes and is commonly used for the detection and monitoring of glucose and cell viability. Furthermore, it is identified as a signal of oncogenic growth due to its widespread presence within the cancer cell environment. However, the low concentrations of H 2 O 2 released by cancer cells' metabolism challenge current detection methods' capabilities and their practicality for translation to clinical applications. Colorimetric assays with simple readouts are a promising solution, provided that their sensitivity and rapidity in detecting H 2 O 2 improve. Here, a plasmonic enhanced nanopatterned platform is proposed coupled with an Amplex Red assay to monitor the color change of H 2 O 2 released from cancer cells. The nanopatterned platform embedded into a multiplexed microfluidic device enhances the kinetics of the reaction ≈7 times. This approach has reached a limit of detection of 1 p m when tested in breast (MCF‐7) and prostate (PC‐3) cancer media. The collected color images are processed and analyzed by a machine learning algorithm that categorizes them into “high” or “low‐to‐no” concentrations of H 2 O 2 with 91% accuracy. This study is a step toward developing a device for highly sensitive H 2 O 2 detection that is easily adaptable, user‐friendly, portable, and automated.