Taiwan Integrated Database for Intelligent Sleep
(TIDIS)

TIDIS Meeting photos

Three research core axes

Project background and purpose

(一) Test subject

Application of artificial intelligence in sleep apnea

(二) Background and research purpose

Artificial Intelligence (AI) is a comprehensive expression of various data science technologies, which includes diverse knowledge ranging from theoretical to practical ones. Well-known applications of AI include Watson, Echo/Alexa, or self-driving car technology. Among various ingredients leading to the success of AI, a large amount of data with accurate experts’ annotation plays the central role. However, obtaining high quality data with accurate annotation is still challenging in many fields, particularly medicine. Due to the widespread existence of wearable devices, it is now easier to obtain a large amount of data in the medical society, but accuracy is still a challenge. The sleep center is equipped with accurate monitoring systems and managed by abundant sleep experts, which renders it a suitable environment to collect high quality dataset with accurate annotation. The purpose of this research is unifying several medical centers in Taiwan, under the umbrella of National Center of Theoretical Science, to establish a large database with accurate annotation and foster an artificial intelligence system for sleep medicine. 

(三) Research and design

Implementation period: expected from April 2018 to December 2022.

Number of samples: This database construction project is retrospective. Following the case estimation of participating medical centers across the country (e.g., Linkou Chang Gung has 500 patients per year), about 5000 patients can be collected each fiscal year. 

What are the measurements, time required, and risks? This study is a retrospective analysis, mainly analyzing medical records of completed whole night sleep examinations.

Statistical analysis method: It depends on the research topic.