# 睡眠中心

蕭嘉宏老師、Alfred Lee、莊舒雅、林yuwen

# System Architecture

Server Location: H707
- Content Server (for storing sleep signals)
- FHIR Server (for storing document indexing)
- Portal
- Client APP

## Content Server
For years, the sleep center has been collecting sleep signals from patients and storing them on a computer. A content server has been created to store these sleep signals. The sleep signals will be anonymized before being uploaded to the content server.

### De-Identification
Currently, biomedical signals or medical images files produced by medical instruments consist of two parts:
1. file header: patient data, medical encounter, modality parameter, etc)
2. the actual signal / image data

De-identification steps are as below:
1. Read biomedical signal or medical image file.
2. Get file header.
3. Extract information from file header, store the information separately (local).
4. Publish file (only contains data).
5. Generate file metadata (patient data, medical encounter, modality parameter, access URL, etc). Save separately.

Using the above de-identified data, an anonymous content repository and viewing system can be created.

## FHIR Server
The FHIR server is used to index the patient and the anonymized data.

## Portal
- organization member management

### Authentication and Authorization Mechanism
- ID token

## Client APP
Client applications for access sleep signals, such as:
- Signal viewer
- Signal analyze system

# Future Applications
## Sleep abnormality detection at home
- Using IoT for collecting vital signs
- Upload signs to cloud platform for storage