Decision Support Systems for Predicting Erectile Dysfunction

Abstract

Abstract. Men with ED generally complain failure to attain or maintain an adequate erection during sexual activity. The prevalence of ED is strongly correlated with age, affecting about 40% of men at age 40 and nearly 70% at aged 70. A variety of chronic diseases, including diabetes, hypertension, cardiovascular disease (CVD), chronic renal failure, depression, sleep disorder, and gout were shown to be associated with ED. In this study, the data used for designing the clinical decision support system (CDSS) was retrieved from a subset of the National Health Insurance Research Database (NHIRD) of Taiwan. The positive cases were male patients with age ranging from 20 to 65 years old who were diagnosed with ED (ICD-9-CM Code 607.84) between Jan. 2000 and Dec. 2010 confirmed by at least 3 outpatient visits, while the negative cases were randomly selected from the database with no history of ED and 1:1 frequency-and age-matched with the ED patients. Experiments of 1 cross validation (10 folds) and 2 independent training and testing (ITT) were conducted to verify the effectiveness of the predictive models. The results show that the sensitivity, specificity, and accuracy of tenfold cross validation achieved 69.45%, 69.45%, and 69.54%, respectively. For the ITT experiments, the sensitivity, specificity, accuracy, and the area under ROC curve were 70.28%, 72.48%, 71.32%, and 0.7226, respectively in the first experiment, and 69.42%, 70.74%, 70.06%, and 0.7143, respectively, in the second experiment. Future works will focus on designing the CDSSs with ensemble classifiers consisting of multiple SVM models by adopting the laboratory data to improve the predictive performance for ED prediction.

Design of a Clinical Decision Support System for Fracture Prediction Using Imbalanced Dataset

“…Recently, we have conducted a prospective study to verify the effectiveness of a designed CDSS in ventilator weaning [ 19 ] and designed a predictive model for predicting erectile dysfunction using the Taiwan National Health Insurance Database [ 86 ]. For predicting successful ventilator weaning, a predictive sensitivity of 87.7% has been achieved by using CDSS, which is significantly higher ( p < 0.01) than the weaning determined by physicians (sensitivity: 61.4%).…”