基于微阵列芯片与智能视觉的农药多残留检测研究及应用

Research and Application of Pesticide Multi-Residue Detection Based on Microarray Chip and Intelligent Vision

摘要: 目的：为解决传统检测在农药残留多重检测中集成度低、操作繁琐、判读主观等问题，本研究旨在开发一种基于高通量微阵列免疫分析与智能识别系统的16种农药残留同步检测方法。方法：基于免疫侧向层析原理，以非接触式点样仪将百草枯等16种农药抗原固定于硝酸纤维素膜，制得微阵列芯片；用胶体金或胶体碳标记特异性抗体，将芯片与微孔试剂组装为检测卡，结合微阵列读数仪，并基于机器视觉优化定量算法对豇豆等农产品样本中的16种农药残留进行检测分析。结果：该智能识别系统的图像采集重复性良好，各检测点灰度值的批内变异系数 < 1.59%，边缘定位精度达亚像素级。在最优条件下，16种农药在三种基质中的检出限范围为0.003~6 mg/kg，加标回收实验显示假阳性率与假阴性率均为0%，且与结构类似物无交叉反应，特异性强，整个检测流程仅需13 min。结论：本研究建立的高通量微阵列生物芯片检测系统，突破了传统方法在多残留检测中的通量限制，为农产品中农药多残留的现场快速筛查提供了高效、精准的解决方案。

Abstract: Aims: In order to solve the problems of low integration, cumbersome operation and subjective interpretation of traditional detection in multiple detection of pesticide residues, this study aims to develop a simultaneous detection method of 16 kinds of pesticide residues based on high-throughput microarray immunoassay and intelligent recognition system. Methods: Based on the principle of immune lateral chromatography, 16 kinds of pesticide antigens, such as paraquat, were fixed on nitrocellulose membrane by non-contact spotter, and the microarray chip was prepared; the specific antibodies were labeled with colloidal gold or colloidal carbon, and the chip and microporous reagent were assembled into a detection card. Combined with the microarray reader, 16 kinds of pesticide residues in agricultural products, such as cowpea, were detected and analyzed based on the machine vision optimization quantitative algorithm. Results: The image acquisition of this intelligent recognition system exhibited good repeatability, with intra-batch coefficient of variation (CV) for grayscale values at each detection point < 1.59%. The edge positioning accuracy reached sub-pixel level. Under optimal conditions, the detection limits for 16 pesticides in three matrices ranged from 0.003 to 6 mg/kg. The recovery test showed that the false positive rate and false negative rate were both 0%, and there was no cross reaction with structural analogues, with strong specificity. The whole detection process took only 13 minutes. Conclusion: The high-throughput microarray biochip detection system established in this study breaks through the flux limitation of traditional methods in multi-residue detection, and provides an efficient and accurate solution for the on-site rapid screening of pesticide residues in agricultural products.