基于中风病医案证候涌现的意象诊疗模式研究

As far as we know, dialectical treatment is called into question currently. And syndrome research methodology also gets into trouble. The key of these questions is the scientificalness of Chinese medicine methodology. On the monism, TCM syndrome is the law of subjective and objective one that imagery-thinking identifies. The scientific expression of the imagery-thinking is the bottleneck of the ideation and manifestation mode of TCM diagnosis and treatment. This study draws on the method of the emergence in the complex systems, and that portrays self-organization emerge of the imagery-thinking in the syndromes emerged and image diagnosis and treatment path. In this research, we will build the ideation and manifestation of TCM diagnosis and treatment of neural network system what is on the basis of massive data of stroke medical records. Through improving the emergence of dynamic self-organizing map networks、 algorithm and Visual analysis method, we will study of stroke syndromes in the evolution of the law and the characteristics of syndrome that is "dynamic time and space，interior-excess and exterior-deficiency and multi-dimensional interface". By the design approach of multi-agent systems to the emergence of the micro - macro mechanism, we will study of the relationship between emergence of heat-phlegm and sthenic-fu syndrome and disease prognosis. Then the imagery-thinking is portrayed in the emergence of syndrome and ideation and manifestation in the diagnosis and treatment path. Through combining ideation and manifestation、blending subject and object and human-computer interaction, stroke ideation and manifestation mode of TCM diagnosis and treatment system will be formed and optimized.The system interprets ideation and manifestation mode of TCM diagnosis and treatment as "Taking Phenomenon as factor， factor as phenotype，phenotype as syndrome, according to the testimony of disease, square card corresponding " . That is the adaptive scientific connotation of imagery treatment model. Significance: From the multi-agent systems microscopic self-organization to generate macro "Emerging", the research will reveal an objective understanding law of imagery-thinking. The combination of emergency theory and holism will provide a revelation for the intermediation of imagery and ontology.

辨证论治遭受质疑根源于对中医方法论的科学性认识问题。一元论下的中医证候是象思维辨识的主客一体的规律。突破象思维客观表达瓶颈是中医诊疗模式研究的关键。本研究借鉴复杂系统涌现思想与方法，在证候涌现与意象诊疗路径中刻画象思维自组织涌现。依托名医工作站中风病医案海量数据，构建集成了神经网络的意象诊疗系统。通过改良涌现自组织映射网络及改进算法与可视化分析方法，研究中风病证候演变规律及证候 "内实外虚、动态时空、多维界面" 的涌现特征。采用多主体系统涌现微-宏观机制设计方法，研究痰热腑实证涌现与疾病预后的关系。经意象结合、人机互动的反馈，实现主客交融中风病意象诊疗系统的完善优化，诠释"以象为素，以素为候，以候为证，据证言病、方证相应"的意象诊疗模式的自适应科学内涵。意义：从多主体系统微观自组织生成宏观"涌现"的过程揭示象思维客观认识规律。为象体融通提供了从整体论与涌现论综合研究的方法与思路。

本项目依托王永炎院士名医传承工作站中对医案整理的前期工作，基于王永炎院士在中风病疾病与证候诊断标准化规范化研究成果和化痰通腑治疗痰热腑实证的研究成果，以王永炎院士在东直门医院主管病房及查房时的中风病案为研究对象，采用深度学习方法，进行了“以象为素、以素为候、以候为证”意象诊疗过程的象思维及痰热腑实证涌现的刻画，从而形成意象诊疗系统辅助决策的软件开发研究。并进一步根据王永炎院士及脑病团队完成的中风病诊疗的研究成果构建意象诊疗系统。.本研究对东直门医院1975-2005年30年间共2000多个中风病案进行了筛查、系统调查、扫描、病历信息数据的整理和归纳，构建了经过扫描纸质版原始病案形成王永炎院士诊疗中风病原始病案数据库，在此基础上经过反复调研完善形成了结构化的中风病住院诊疗信息数据库。.最终收录符合研究标准的原始中风病案1803份进入中风病结构化数据库。其中中经络1519例（痰热腑实490例），中脏腑313例（痰热腑实238例）。采用深度学习算法对结构化数据库进行学习和演练，形成的意象诊疗系统的辅助决策显示：证候的平均预测准确率为70.94%，数值越高对证候的判断越准确。查全率（敏感性）最高的是痰热腑实，为0.9960，除此之外，大于0.9的还有风痰瘀血和痰热瘀血，数值最低的是气虚血瘀，为0.36054。其次是风痰淤血0.9741，0.8到0.9之间的有的有元气败脱神明散乱、肝阳暴亢、气阴两虚，0.7到0.8之间的有痰热内闭清窍、阴虚风动。通过神经网络（深度学习）学习病案信息形成的对中医证候的识别准确率，痰热腑实证的识别率最高，其次是风痰瘀血等，符合既往的经验认识，验证了中医辨证论治的科学性。并将进一步对证候“内实外虚，动态时空，多维界面”进行可视化呈现。.新增的研究工作及项目：增加了以隐马尔科夫转移矩阵概率测算的中风病证候演变研究；新增并完成1882份中风病原始病案扫描构建了王永炎院士诊疗中风病的原始医案库；新增完成的《王永炎院士神经内科病证实验录》，已获出版号于中国中医药出版社2017年出版。