Machine learning-assisted wide-gamut fluorescence visual test paper for propazine determination in fish and seawater samples

Sensors and Actuators B: Chemical(2024)

引用 0|浏览7
暂无评分
摘要
Molecularly imprinted polymer (MIP-QDs) with fluorescence quenching ability toward propazine was synthesized for propazine detection. b(Blue)-MIP-QDs were prepared using ZnCdS/ZnS QDs via reverse micro-emulsion, whereas r(red)-MIP-QDs were synthesized using CdSe/ZnS QDs. By utilizing graphene quantum dots (GQDs) as a stable fluorescence intensity reference, the wide-gamut fluorescence test paper was constructed on the basis of mixing b-MIP-QDs, r-MIP-QDs, and GQDs under the optimal ratio. When analyzing spiked propazine in fish and seawater samples using a test paper, satisfactory recoveries of 104.0%–114.6% and 92.0%–96.4% were obtained, with corresponding limits of detection of 5.0μg/kg and 1.0μg/L, respectively. The RGB extractor was utilized to extract the actual fluorescence color and construct a dataset consisting of R, G, and B values, as well as concentration data from 400 samples. The SVR model of Python 3.9.7 was used to obtain and analyze the concentration and feature data. After optimization, the constructed model achieved a correlation coefficient of 0.98 and an RMSE of only 1.81, indicating high prediction accuracy and excellent generalization ability that meet quenching prediction requirements. As an intelligent and rapid detection method, this model holds significant practical significance.
更多
查看译文
关键词
Propazine,Molecularly imprinted,Fluorescence sensor,Visual detection,Machine learning
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要