融合电子化临床指南的心力衰竭治疗决策模型

Electronic Clinical Guideline Embedded Heart Failure Treatment Model

为有效实施心力衰竭疾病的个性化治疗,本研究提出融合电子化心力衰竭临床治疗指南的治疗模型方案:以电子化临床指南为基础,对心衰状况进行多目标评估并制定个性化治疗策略。建模过程筛选特异性参数并聚类分组,分别为:心结构、运动耐量、体液潴留、心功能、神经活性和电生理等一级指标值,根据指标所指示的六方面状况严重程度给予相应的药物治疗,并采用线性加权法获得患者心衰综合指标。对75例临床病例的初步测试结果表明:心力衰竭疾病治疗模型能全面评估心衰状况,并依据患者具体病症做出建议治疗方案,建议方案中单一用药的平均正确率达到91%,该模型是一种有效的个性化治疗决策辅助工具。

In this paper, a heart failure （HF） treatment model based on electronic clinical guideline was proposed to realize personalized therapy. Integrated with clinical guideline of HF, a multi-objective optimization algorithm was employed to assess diseases severity and then produce individual treatment strategies. HF risk related parameters were screened and cataloged in six subclasses, i.e. cardiac structure, exercise tolerance, fluid overload, cardiac function, neural activity and electronic physiology. Accordingly, medication plans for individuals were determined. By linear weighting methods, a comprehensive index was generated to assess HF status of patients. The model was validated on 75 subjects. And the preliminary results demonstrated that the model was able to assess HF status and give treatment suggestions. In the term of agreement with the prescriptions given by physicians, the average accuracy of single drug suggested by the model was 91%. This work offered a helpful tool for facilitating individual HF treatment.