基于深度学习的穴位敏化客观化方法研究  

Research on Objectification Method of Acupoint Sensitization Based on Neural Network Depth Learning

作　　者：温川飙[1] 陈菊[1] 贾伟[1] 谭惠娟[1] 吴曦[1] 黄玲[1] 梁繁荣[1] 

机构地区：[1]成都中医药大学,四川成都610072

出　　处：《辽宁中医杂志》2017年第8期1723-1725,共3页Liaoning Journal of Traditional Chinese Medicine

基　　金：国家重点基础研究计划(973计划)(2012CB518501);国家自然科学基金资助项目穴位的敏化研究(81590950)

摘　　要：目的:通过穴位敏化的客观化规律及现状分析,提出基于深度学习的穴位解决方法。方法:针对穴位敏化研究缺乏从宏观角度囊括不同敏化形式的大样本穴位敏化综合性的问题,提出采用深度学习技术深入开展穴位敏化的客观化研究的方法。结果:从提取穴位敏化客观动态数据特征、实现穴位敏化特征预处理的深度神经网络设计、建立穴位敏化识别策略权重估值网络的角度,探讨穴位敏化的客观化研究的潜在切入点。结论:通过对潜在切入点的研究,表面中医的整体思维模式与大数据条件下的深度学习技术的结合,必然产生一种不同于实验和实证研究的全新的中医研究方法。Objective： To provide the key technical support for the further research on acupoint sensitization through the research of objectification and acupoint sensitization. Methods： This paper analyzed acupoint sensitization and objectification and found out it is a comprehensive problem which is lack of a large sample including different sensitizations and it proposed using the neural network depth learning techniques to carry out the possibility of objectification research on the acupoint sensitization. Results： By extracting from the objectiive data characteristics of acupoint sensitization,realizing the artificial neural network design and establishing the acupoint sensitization method,it identified the potential entry points. Conclusion： Through the study on the potential entry points,it believes the whole thinking mode of TCM and big data under the condition of the depth of artificial intelligence learning techniques is bound to have a difference from experimental and empirical research of the new medicine research methods.

关 键 词：深度学习 穴位敏化 客观化研究 

分 类 号：R245.9[医药卫生—针灸推拿学]