Research On Clinical Decision Support Methods Based On Knowledge-based Reasoning For Traditional Chinese Medicine Diagnosis And Treatment

As a complex decision process, clinical diagnosis and treatment (CDT) is always a challenging process of evolution and innovation because of complication of diseases and individuality of patients. Versus modern medicine, in view of its greater creative and decision-making, traditional Chinese medicine (TCM) CDT is an extremely complex system. Moreover, clinical ability and level varies among TCM physician. Therefore, it is an important topic of investigation on and succession of TCM clinical that how to improve the average effect of CDT and to quickly promote young physicians’competence. Construction of common knowledge about TCM diagnosis and treatment and accumulation of clinical case data furnish CDT decision which is assisted with computer technology with feasibility and opportunities. In this context, this paper research TCM clinical diagnosis and treatment decisions based on knowledge reasoning. Combining inductive logic programming (ILP) and Markov logic network (MLN), we achieve diagnosis decision support, and further realize prescribed treatment decision support ground on case-based reasoning. The concrete research works in the paper is the following aspects:1. Detecting TCM syndrome diagnosis decision-making using the ILP-based algorithm framework, combining with the characteristics of TCM CDT, considering the similarities among syndromes, improving the diagnostic rules method and forming TCM clinical syndrome diagnosis rules knowledge base. The improved method generates less rules, and each rules cover more positive cases, and owns higher accurate prediction.2. Through weight learning with the qualitative rule that is obtained by applying MLN weight training algorithm to ILP, and then reasoning with MC-SAT slice sampling algorithm, deriving possible TCM syndrome diagnosis based on the symptoms and signs of patients, we achieve diagnostic decisions supporting. Reasoning based the similarities among ontology-based semantics can get more accurate results.3. Employing clinical case data, joining with case-based reasoning, attaining the similarity cases from related cases through similarity calculation method, consuming the prescriptions in the similarity cases as recommendation prescriptions and finally triumphing TCM prescriptions treatment decisions supporting. We also research and develop o knowledge-based reasoning TCM clinical treatment decision method. We experiment and test in the insomnia cases and get better treatment decision-making effect.