Analysis Of The Prognostic Factors Of Patients With Bullous Pemphigoid Based On Random Forest Model

Objective:To research the prognostic elements of patients with bullous pemphigoid using random forest model and Logistic regression model,so as to provide a basis for treatment and prognosis judgment.Methods:Inpatients who were diagnosed with bullous pemphigoid in the dermatology ward of the first hospital of Shanxi medical university from January 2015 to April 2021 were selected as the study object.The random forest model and Logistic regression model were constructed to research the prognostic elements of patients with bullous pemphigoid and the predictive efficacy of the two models was compared.Results:The results of random forest model showed that the top five priorities were age,mucous membrane involvement,neurological disease,blood calcium,and local skin infection.Logistic regression model indicated that advanced age,hypocalcemia,and patient with neurological diseases were risk factors for poor prognosis in BP patients and Patients with mucosal involvement had lower mortality within 1 year.The parameters for random forest prediction model in the training set were as follow:AUC,1.000,accuracy,1.000,senstivity,1.000,specificity,1.000,F1 score,1.00.In test set,AUC was 0.763,accuracy was 0.8148,sensitivity was 0.6667,specificity was 0.8571,F1 score was 0.62.The parameters for Logistic regression model in the training set were as follow:AUC,0.923,accuracy,0.8788,senstivity,0.7500,specificity,0.9074,F1 score,0.69.In test set,AUC was 0.8118,accuracy was 0.8148,sensitivity was 0.7778,specificity was0.5556,F1 score was 0.63.The difference between the training set and the test set of the Logistic regression model was less than that of the random forest model and the model was more stable.After the union of Logistic regression model and random forest model,the accuracy,sensitivity and specificity of the training set are 100%.The accuracy of the test set was 85.2%,the sensitivity was 71.42%,and the specificity was 90%,which were higher than the two models alone.Conclusion:Age,mucous membrane involvement,neurological diseases,blood calcium and local skin infection are the most significant factors influencing the prognosis of patients with bullous pemphigoid.Random forest and Logistic regression have their own advantages and disadvantages.Combining the two models to predict the prognosis of patients with bullous pemphigoid is of more practical significance.