Nazir Faisal, Gregucci Denise, Calabretta Maria Maddalena, Cambrea Caterina, Vahidi Peyman, Lavrnić Stevo, Toscano Attilio, Michelini Elisa
Department of Chemistry "Giacomo Ciamician", University of Bologna, Via P. Gobetti 85, 40129 Bologna, Italy.
Department of Agricultural and Food Sciences, University of Bologna, Viale Fanin 50, 40127 Bologna, Italy.
Anal Chem. 2025 Aug 26;97(33):18092-18100. doi: 10.1021/acs.analchem.5c02369. Epub 2025 Aug 12.
Several biosensors for water toxicity monitoring have been reported in the literature; however, none of them fully integrate both analytical and post-analytical steps that are required in a standard laboratory setting before reporting the result. To provide a workflow for smartphone biosensor developers, we implemented a novel procedure that was applied to the standard toxicity assay based on the bioluminescent bacteria . We addressed the main issues to turn this method into a sustainable all-in-one toxicity paper biosensor, i.e., the immobilization of bacteria, the integration of a calibration curve, and a customized artificial intelligence (AI) application that converts the smartphone picture into user-friendly quantitative information. The biosensor analytical performance was evaluated with different water contaminants and real water samples, showing promising results. A limit of detection of 0.23 ppb was obtained for the cyanotoxin microcystin-LR produced by harmful algal blooms. We also demonstrated for the first time that the inclusion of a calibration curve in a paper sensor, combined with an AI app, enables accurate analyses even when pictures are taken with smartphone models equipped with cameras with different resolutions. To the best of our knowledge, this is the first bioluminescence paper biosensor in which an AI algorithm enables to obtain quantitative results by interpolating the bioluminescent signals from an on-board calibration curve. We believe this novel biosensor will open new opportunities not only for water monitoring, but the same approach could be implemented in any optical smartphone biosensor for applications spanning from onsite analysis to citizen science.
文献中已报道了几种用于水毒性监测的生物传感器；然而，在报告结果之前，它们都没有完全整合标准实验室环境中所需的分析和分析后步骤。为了为智能手机生物传感器开发者提供一个工作流程，我们实施了一种新颖的程序，该程序应用于基于生物发光细菌的标准毒性测定。我们解决了将该方法转变为可持续的一体化毒性纸质生物传感器的主要问题，即细菌的固定化、校准曲线的整合以及将智能手机图片转换为用户友好的定量信息的定制人工智能（AI）应用。用不同的水污染物和实际水样评估了生物传感器的分析性能，结果令人满意。对于有害藻华产生的蓝藻毒素微囊藻毒素-LR，检测限为0.23 ppb。我们还首次证明，在纸质传感器中纳入校准曲线并结合AI应用程序，即使使用配备不同分辨率摄像头的智能手机型号拍摄图片，也能进行准确分析。据我们所知，这是第一种生物发光纸质生物传感器，其中AI算法能够通过对板载校准曲线的生物发光信号进行插值来获得定量结果。我们相信，这种新型生物传感器不仅将为水监测带来新机遇，而且相同的方法可以应用于任何光学智能手机生物传感器，应用范围从现场分析到公民科学。
