模式识别在中药质量评价中的应用进展

Application of Pattern Recognition to Quality Assessment of the Traditional Chinese Medicine

目的 :对模式识别在中药质量评价中研究进展进行综述。方法 :总结有关化学模式识别、显微图像模式识别及人工神经网络在中药质量评价中应用的研究文献 ,综述不同方法在该领域中的应用。结果 :到目前为止 ,红外、紫外、裂解 高分辨气相色谱、GC Mass总叠加质谱等被应用于中药的化学总特征的表现 ;基于体视学和计算机图像测试技术的图像定量分析 ,以及图形生成理论和计算机图形学的三维重建和显示技术被应用于中药组织的体视学参数的确定 ;误差反传等类型人工神经网络在中药质量评价中也有重要的应用。结论 :化学模式识别、显微图像模式识别、人工神经网络在中药质量评价中起到了重要的作用 ,具有广阔的发展前景。

Objective: To review the application of the pattern recognition to quality assessment of the Chinese Traditional Medicine. Method: Papers of application of chemical pattern recognition, pattern recognition of stereoscopic features and artificial neural networks to quality assessment of the Chinese Traditional Medicine were collected and the application of different method was reviewed. Result: IR, UV, GC, GC-Mass and the sum mass spectra were used to evaluate the chemical character of different samples, and the microscopic features of the epidermis were scored quantitatively by stereology and image analysis on the basis of classical taxonomy,. BP networks were also widely used. Conclusion: Chemical pattern recognition, pattern recognition of stereoscopic features and artificial neural networks are of important use in quality assessment of material medica.