<b><i>Background/Aims:</i></b> The aims of study were (1) to determine the cut-off values of parameters related to auditory perceptual assessment (visual analog scale [VAS]) and acoustic analysis (cepstral peak prominence [CPP], low-frequency/high-frequency [L/H] ratio, CPP SD, L/H ratio SD, acoustic voice quality index [AVQI], and cepstral spectral index of dysphonia [CSID]) for predicting voice problems within a Korean population, and (2) to verify the discriminative power of these cut-off values. <b><i>Methods:</i></b> 1,113 voice samples were analyzed in this study. Perceptual assessments (VAS) were performed by 5 speech-language pathologists. For the acoustic analysis, cepstral parameters, CSID, and AVQI were calculated. The cut-off values of parameters predicting voice problems were obtained using receiver operating characteristic (ROC) analysis. Additionally, the sensitivity, specificity, and area under the ROC curve (AUC) were measured. <b><i>Results:</i></b> High reliabilities were observed for the perceptual assessments. The cut-off values of parameters had a high sensitivity, specificity, and AUC. Of these, CSID was the parameter with the highest AUC values. <b><i>Conclusion:</i></b> Each parameter demonstrated a high discriminative power for classifying the presence or absence of voice problems. The results of this study could be used as an objective criterion for screening voice problems.
Purpose:The increased incidence of asthma due to rising allergic diseases requires the prevention of worsening asthma. It is necessary to develop a patient-tailored asthma prediction model. Methods: We developed causative factors for the asthma forecast system: infant and young children (0-2 years), preschool children (3-6 years), school children and adolescents (7-18 years), adults (19-64 years), old aged adult (> 64 years). We used the Emergency Department code data which charged the short-acting bronchodilator (Salbutamol sulfate) from Health Insurance Review and Assessment Service for the development of asthma prediction models. Three kinds of statistical models (multiple regression models, logistic regression models, and decision tree models) were applied to 40 study groups (4 seasons, 2 sex, and 5 age groups) separately. Results: The 3 kinds of models were compared based on model assessment measures. Estimated logistic regression models or decision tree models were recommended as binary forecast models. To improve the predictability, a threshold was used to generate binary forecasts.
Conclusion:We suggest the binary forecast models as a patient-tailored asthma prediction system for this category. It may be needed the extended study duration and long-term data analysis for asthmatic patients for the further improvement of asthma prediction models.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.