Study On Semantic-Based Knowledge Base For Clinical Decision Support

With the deepening of the medical and health system reform in our country,the healthcare industry is faced with many challenges such as an explosion of data,a lack of knowledge,an over-load of demand,and a shortage of resources.Clinical decision support systems(CDSSs)can pro-vide healthcare decision makers with automatic recommendations on the prevention,diagnosis,and treatment of diseases,thus holding great values in improving the efficiency and quality of health-care services,optimizing the allocation of healthcare resources,and promoting population health.Knowledge base is the core component of CDSSs,however,the traditional method of knowledge base construction is not sufficient to express the diversity,complexity and variability of health-care information,and often lacks standardization;In addition,the ability of knowledge sharing and personalization was limited.In this paper,we presented a semantic-based approach to knowledge base construction for CDSSs,practical and effective solutions to address the above mentioned prob-lems in terms of data,application,and knowledge were proposed.The ultimate goal is healthcare knowledge standardization and sharing.The main innovation points are as follows:A decision-making knowledge base model that supports semantic interoperability was devel-oped,solving the problem of insufficient interaction between CDSS and the data layer,integrated analysis and intelligent decisions on multi-domain,multi-dimensional and multi-level knowledge and data were realized.A four-phase knowledge engineering cycle for the acquisition,representa-tion,application and evaluation of healthcare knowledge was proposed,enabling continuous update and improvement of the knowledge being modeled;ontology and semantic rules based on standard healthcare information model were designed for the explicit,formal and normalized representation healthcare domain knowledge;a CDSS architecture based on semantic middleware was put for-ward,an information highway between CDSS and hospital electronic medical records(EMRs)was established through semantic mapping,decision support for personalized diagnosis and treatment were achieved.A modularization method of decision-making knowledge base construction and application was proposed,solving the problem of insufficient adaptability of the CDSS in the application lay-er,seamless integration of CDSS and clinical workflow was realized.Independent,reconfigurable knowledge modules were developed,serving the demands for knowledge at different levels of com-plexity under different healthcare circumstances;these knowledge modules were encapsulated and released as semantic services,rich semantic specifications on the content,information payload and transport protocol were provided,facilitating automated discovery and flexible utilization by ap-plications,reducing the efforts and cost of CDSS development,implementation and maintenance,meanwhile improving applicability and sharability.A CDSS for personalized chronic disease follow-ups was developed,solving the problem of insufficient comprehensiveness of the CDSS on the knowledge layer,scaling of CDSS appli-cation from diagnostic and treatment auxiliary to lifelong health management was realized.A semantic EMR was designed to satisfy the long-term and multi-factor characteristics of chronic diseases,enabling complete and accurate recording of patient data;a domain knowledge base to support symptom identification,observation interpretation and rehabilitation guidance was built;a decision-making knowledge base to support patient assessment,disease control and health man-agement was built;a CDSS for chronic disease follow-ups was developed to improve the efficiency and quality of patient follow-ups by healthcare service providers,meanwhile improving the self-management ability of patients.This paper introduces semantic technologies to decision-making knowledge base construc-tion,approaches to standard representation,cross-platform sharing and personalized application of decision-making knowledge were put forward from the perspective of the data layer,the application layer and the knowledge layer,providing technical support for improving the quality of healthcare services in our country.