关键词: 临床科研一体化; 专病库; 神经介入; 病历内涵质控; 风险预测

Abstract:

Neurointervention is an important method for the treatment of cerebrovascular disease. In order to improve the level of clinical and scientific research more effectively, Beijing Tian Tan Hospital Affiliated to Capital Medical University established an integrated research model of clinical and research with the specialized disease database of neurointervention as the core. Based on artificial intelligence technologies such as natural language processing, machine learning and deep learning, through automatic collection, processing and modeling of the data of 379 patients with anterior circulation cerebral infarction who admitted to the intervention center for thrombectomy from May 2012 to June 2019, neurointervention center built a specialized disease database of emergency thrombectomy for anterior circulation cerebral infarction. After establishing the integrated research model, on the one hand, it formed a display view of multidimensional data relationship with extremely fine granularity, which provides more possible directions for clinical research; on the other hand, the whole-process diagnosis and treatment timeline data and intelligent prediction engine can be used to predict the risk of intracranial hemorrhage after thrombectomy, which is of great significance for the adjustment and optimization of clinical treatment decisions. Therefore, the integrated research model based on specialized disease database will effectively drive the output and transformation of scientific research achievements, and realize the goal of taking reality diagnosis and treatment data as the basis and research results as the aid to clinical decision-making, which has a broad development prospect.

Key words: Integration of clinical research; Specialized disease database; Neurointervention;uality control of medical record; Risk prediction