| abstract |
A traumatic coagulopathy prediction method provided by the present invention includes: obtaining the diagnosis and treatment information of trauma patients to be predicted; extracting the prediction feature information of the diagnosis and treatment information; inputting the prediction feature information into the pre-built trauma coagulation disease In the prediction model, to obtain the prediction result; wherein, the traumatic coagulopathy prediction model is constructed as follows: obtain the diagnosis and treatment information of the sample trauma patient; extract the sample characteristic information of the diagnosis and treatment information; according to the sample characteristic information and the The sample results corresponding to the sample feature information are trained to construct the traumatic coagulopathy prediction model. According to the prediction model of traumatic coagulopathy, the results of traumatic coagulopathy in trauma patients can be predicted, and doctors can diagnose high-risk patients early based on the results, so as to intervene as early as possible in the critical treatment window period, carry out early identification, and give clinical intervention in time, such as initiating blood transfusion procedures , supplementation of coagulation factors and early use of tranexamic acid, etc. |