| 李文静,毕束清,任夏,等.基于紫外光谱数据的植物类海洋中药寒热药性模式识别评价研究[J].中国海洋药物,2026,(1):-. |
| 基于紫外光谱数据的植物类海洋中药寒热药性模式识别评价研究 |
| Study on pattern recognition of plant marine traditional Chinese medicine with cold-heat nature based on Ultraviolet spectrum data |
| 投稿时间：2024-12-20 修订日期：2025-02-26 |
| DOI： |
| 中文关键词: 紫外光谱 海洋中药 寒热药性 模式识别技术 |
| English Keywords:Ultraviolet spectrum marine traditional Chinese medicine cold-heat nature pattern recognition technology |
| Fund Project: |
|
| 摘要点击次数: 146 |
| 全文下载次数: 0 |
| 中文摘要: |
| 目的 基于紫外光谱数据结合模式识别技术,探讨植物类海洋中药寒热药性物质基础表征规律,为海洋中药寒热药性快速、客观评价提供创新方法和技术支撑,促进我国海洋中药新资源的开发和利用。方法 选取20味寒热药性明确的植物类海洋中药,采用水、无水乙醇、乙酸乙酯、三氯甲烷和石油醚为提取剂,平行制备3份样本,通过紫外可见吸收光谱法获得300组光谱数据。应用支持向量机、决策树、随机森林、朴素贝叶斯和K近邻5种模式识别技术,构建分类判别模型,并采用K折交叉验证与留一法进行模型验证与评估。结果 寒热药性植物类海洋中药在紫外光谱特征上表现出显著差异。寒性药在紫外光谱中通常表现出更多、更丰富的吸收峰,尤其在290 nm、340 nm和410 nm附近,而热性药的吸收峰较少且分布范围较窄。留一法验证显示,基于SG平滑的二阶导数预处理的紫外光谱数据建立的随机森林模型表现最佳,是植物类海洋中药药性分类辨识的最优模型。结论 通过紫外光谱特征吸收峰的位置结合随机森林等模式识别技术,能够有效揭示植物类海洋中药的寒热药性差异。本研究为构建海洋中药药性综合评价体系提供了科学参考,有助于扩大海洋中药的临床应用,为我国海洋中药资源的深入开发提供新思路。 |
| English Summary: |
| Objective To investigate the substance basis characterization of plant marine traditional Chinese medicine with cold-heat nature using Ultraviolet (UV) spectrum data combined with pattern recognition technology, to develop innovative methods and provide technical support for rapid and objective evaluation of cold-heat nature of marine traditional Chinese medicine, and further promote the development and utilization of new resources of marine traditional Chinese medicine. Methods In this study, plant marine traditional Chinese medicine with well-defined cold-heat nature was selected as the research object, and water, absolute alcohol, ethyl acetate, chloroform, and petroleum ether were used as extract solvents, 3 parallel samples were prepared. A total of 300 sets of spectrum data were obtained by UV-Visible absorption spectrum. Five pattern recognition technologies, including support vector machine, decision tree, random forest, Naive Bayes, and K-nearest neighbor, were used to construct the discrimination models. The K-fold cross-validation and leave-one method were employed to verify and evaluate the accuracy and performance of constructed models. Results There are significant differences in the UV spectral characteristics of plant marine traditional Chinese medicine with cold-heat nature. The medicines with cold nature usually show more and more abundant absorption peaks in the UV spectrum, particularly around 290 nm, 340 nm, and 410 nm, while medicines with heat nature have fewer absorption peaks and narrow distribution range. By using the leave-one-out method to verify, it is found that the random forest model established with the UV spectrum data preprocessed by the second derivative after Savitzky-Golay smoothing was the optimal model for the classification and identification of plant marine traditional Chinese medicine with cold-heat nature. Conclusion The study establishes that by leveraging the locations of characteristic absorption peaks in the UV spectrum and applying random forest and other pattern recognition models, a significant difference between cold and heat nature of plant marine traditional Chinese medicine can be effectively revealed. This study provides a scientific reference for constructing a comprehensive evaluation system about the cold-heat nature of plant marine traditional Chinese medicine, helps to enlarge the clinical application of marine traditional Chinese medicine, and provides a new idea for the in-depth development of marine traditional Chinese medicine resources. |
| 查看/发表评论 下载PDF阅读器 |
|
| 关闭 |
|
|
|