Severe acute respiratory distress syndrome early warning method and system

Abstract

The embodiment of the invention provides a severe acute respiratory distress syndrome early warning method and system. The method comprises the following steps: obtaining to-be-predicted physiologicalsign data of an ICU critical patient; performing data preprocessing, feature extraction, feature standardization and feature selection on the to-be-predicted physiological sign data to obtain physiological feature time series data; and inputting the physiological feature time series data into a pre-trained monitoring and early warning model to obtain a monitoring and early warning result output by the monitoring and early warning model. According to the embodiment of the invention, four typical physiological characteristic data of a critical patient are extracted, and linear and nonlinear characteristic selection is carried out to construct a monitoring and early warning model, so that the possible acute respiratory distress syndrome of a patient is identified and early warned, and medical personnel are assisted to diagnose and treat the acute respiratory distress syndrome of the patient in time.

Description

technical field [0001] The invention relates to the technical fields of clinical medicine and artificial intelligence, in particular to a method and system for early warning of severe acute respiratory distress syndrome. Background technique [0002] Acute Respiratory Distress Syndrome (ARDS) is a clinical syndrome with a series of pathophysiological reactions caused by acute pulmonary inflammatory response caused by various reasons. Mortality diseases, the incidence of which increases as the population ages. Due to the lack of effective treatment methods, once acute respiratory distress syndrome occurs, the mortality rate is extremely high, and the treatment cost is expensive. Therefore, it is of great clinical significance to give early warning to patients who may develop acute respiratory distress syndrome and take effective treatment in time to prevent its occurrence. [0003] At present, there are generally two early warning methods for acute respiratory distress synd...

AI Technical Summary

Problems solved by technology

[0003] At present, there are generally two early warning methods for acute respiratory distress syndrome as follows: the first is the traditional lung injury prediction score (LIPS), which predicts acute respiratory distress syndrome by using the patient's condition record information after admission, but this The method cannot realize dynamic real-time monitoring and early warning; the second is to use artificial intelligence and machine learning techno

Images