Publications by authors named "Ching-Tzu Tsai"
Article Synopsis
- Predicting 30-day hospital readmissions is essential for enhancing patient care and reducing costs, but prior studies on neurosurgery models lack practical implementation insights.
- The project involved training machine learning models using various methods on a dataset of neurosurgical admissions, aiming for high predictive accuracy and identifying potential interventions through expert interviews.
- The random forest model demonstrated the best performance with an AUROC of 0.89, and six targeted interventions were proposed, leading to a simulated reduction in readmission rates from 13.13% to 10.12%.
Background: Poor sleep hygiene persists in college students today, despite its heavy implications on adolescent development and academic performance. Although sleep patterns in undergraduates have been broadly investigated, no study has exclusively assessed the sleep patterns of premedical undergraduate students. A gap also exists in the knowledge of how students perceive their sleep patterns compared to their actual sleep patterns.
View Article and Find Full Text PDFInteractive data visualization can be a viable way to discover patterns in patient-generated health data and enable health behavior changes. However, very few studies have investigated the design and usability of such data visualization. The present study aimed to (1) explore user experiences with sleep data visualizations in the Fitbit app, and (2) focus on end users' perspectives to identify areas of improvement and potential solutions.
View Article and Find Full Text PDF