Asthma is a chronic respiratory disease that significantly impacts children's quality of life.Timely pre-diction and accurate diagnosis are crucial to the health of children with asthma.However,children in the stable stage of asthma do not exhibit wheezing or other characteristic sounds during an asthma attack.Therefore,there is no significant difference in the breath sounds of children in the stable stage of asthma and those of healthy children,making it challenging for healthcare professionals to diagnose asthma using traditional auscultation methods.This study utilized a support vector machine(SVM)algorithm in machine learning to predict the presence of asthma in children.The results indicate that SVM performed well in classifying the breath sounds of asthmatic and healthy children.The accuracy of the SVM's prediction for the inspiratory phase was 96.53％,while for the expiratory phase it was 91.66％.This demonstrates that the SVM method is highly feasible for diagnosing childhood asthma,can improve the accuracy and efficiency of diagnosis,and can provide a reliable diagnostic tool for this field.
维普
万方数据