School of Physics and Optoelectronic Engineering, Yangtze University, Jingzhou 434023, China.
Shenzhen Key Laboratory of Marine IntelliSense and Computation, Institute for Ocean Engineering, Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China.
Biosensors (Basel). 2024 Mar 28;14(4):160. doi: 10.3390/bios14040160.
Harmful algal blooms (HABs) pose a global threat to the biodiversity and stability of local aquatic ecosystems. Rapid and accurate classification of microalgae and cyanobacteria in water is increasingly desired for monitoring complex water environments. In this paper, we propose a pulse feature-enhanced classification (PFEC) method as a potential solution. Equipped with a rapid measurement prototype that simultaneously detects polarized light scattering and fluorescence signals of individual particles, PFEC allows for the extraction of 38 pulse features to improve the classification accuracy of microalgae, cyanobacteria, and other suspended particulate matter (SPM) to 89.03%. Compared with microscopic observation, PFEC reveals three phyla proportions in aquaculture samples with an average error of less than 14%. In this paper, PFEC is found to be more accurate than the pulse-average classification method, which is interpreted as pulse features carrying more detailed information about particles. The high consistency of the dominant and common species between PFEC and microscopy in all field samples also demonstrates the flexibility and robustness of the former. Moreover, the high Pearson correlation coefficient accounting for 0.958 between the cyanobacterial proportion obtained by PFEC and the cyanobacterial density given by microscopy implies that PFEC serves as a promising early warning tool for cyanobacterial blooms. The results of this work suggest that PFEC holds great potential for the rapid and accurate classification of microalgae and cyanobacteria in aquatic environment monitoring.
有害藻华（HABs）对当地水生生态系统的生物多样性和稳定性构成了全球性威胁。为了监测复杂的水环境，人们越来越希望能够快速、准确地对水中的微藻和蓝藻进行分类。在本文中，我们提出了一种脉冲特征增强分类（PFEC）方法，作为一种潜在的解决方案。PFEC 配备了一个快速测量原型，该原型可以同时检测单个粒子的偏振光散射和荧光信号，从而可以提取 38 个脉冲特征，将微藻、蓝藻和其他悬浮颗粒物（SPM）的分类精度提高到 89.03%。与显微镜观察相比，PFEC 可以揭示水产养殖样本中三个门的比例，平均误差小于 14%。本文发现，PFEC 比脉冲平均分类方法更准确，这可以解释为脉冲特征携带了关于粒子的更详细信息。在所有现场样本中，PFEC 与显微镜在优势种和常见种之间具有很高的一致性，这也证明了前者的灵活性和鲁棒性。此外，PFEC 获得的蓝藻比例与显微镜给出的蓝藻密度之间的高 Pearson 相关系数为 0.958，这意味着 PFEC 可以作为蓝藻水华的一种有前途的预警工具。这项工作的结果表明，PFEC 在水环境保护监测中对微藻和蓝藻的快速、准确分类具有很大的潜力。
