A handheld multifunctional smartphone platform integrated with 3D printing portable device: On-site evaluation for glutathione and azodicarbonamide with machine learning.

Researchers

Journal

Modalities

Models

Abstract

Azodicarbonamide (ADA) in flour can be easily decomposed to semi-carbazide and biuret, exhibiting strong genotoxicity in vitro and carcinogenicity. Glutathione (GSH) can be conjugated with some ketone-containing compounds and unsaturated aldehydes to form toxic metabolites. Here, a novel ratio fluorescence probe based on blue emitting biomass-derived carbon dots (BCDs) and yellow emitting 2,3-diaminophenazine (OxOPD) was prepared for the bifunctional determination of glutathione (GSH) and ADA. This strategy includes three processes: (1) Ag+ oxidizes o-phenylenediamine (OPD) to produce OxOPD. The peak at 562 nm was enhanced, and the peak at 442 nm was reduced due to fluorescence resonance energy transfer (FRET), (2) glutathione binds Ag+ and inhibits the production of OxOPD, (3) ADA oxidizes GSH to form GSSG, resulting in the release of Ag+ by GSH. Therefore, the newly designed ratio fluorescence probe can be based on the intensity ratio (I442/I562) changes and significant fluorescent color changes to detect GSH and ADA. Moreover, a smartphone WeChat applet and a yolov3-assisted deep learning classification model have been developed to quickly detect GSH and ADA on-site based on an image processing algorithm. These results indicate that smartphone ratiometric fluorescence sensing combined with machine learning has broad prospects for biomedical analysis.Copyright © 2021 Elsevier B.V. All rights reserved.