Radnet
A health information exchange service that shares medical images and metadata through cloud-based sharing and real-time image reading
Role
UX Designer,
Project Manager
Team
1 GUI Designer,
3 Developer,
1 Service Designer
Duration
2015.06 - 2016.03 (9 mo)
It is a service that radiologists provide preliminary reads and final reads to hospitals and patients using cloud service. Health hub's cloud server extracts radiologic images from modalities such as CT and MRI, and protect data and allocate image read requests to the right radiologists.
What I Do
My key objectives was to enhance the usability of core products and to expand company's core service to embeded devices (MRI, CT, etc) and mobile applications.
To achieve these goals, I designed and participated in each core modules that connects end to end. I contributed each module at the following proportion.
- HPACS: 100%. I designed and managed the whole development
- HScan: 40%. I redesigned major modules of the mobile application
- HTelerad: 50%. I redesigned major modules and introduced new features to the platform
- HCase: 10%. I evaluated its usability and guided the next design direction of it
It takes too much time and efforts to share radiology images from one hospital to another.
Current tele-radiology service has too many steps and complicated to share radiology data and the radiological image reads.
Service Research
The first step to understand this field was to review all existed documents. And then, I moved on to the target user research. I prepared for user interview questions and usability test cases.
- Understanding Technician's Work Protocol
I followed the technician's daily work for a day and interviewed them to find break points in their routine.
Also, I could understand current workspace settings and how it blocked their work process.
After shadowing technicians and radiologists (I can't share radiologist's workplaces), I interviewed them about their routine and potential break points. Then, I heuristically evaluated the current service flow with Health Hub products. As a result, I was able to put together work flow of the radiology.
Competitive Research
I have checked other services and their strong/ weak points to find where Healthhub's products can fit in.
Technical Research
The service is based on platform as a service (PaaS), one type of cloud services. The service consists with three parts: parsing, communicating, and managing medical images and meta data. The parsing methods varies depend on devices' Operating System: iOS, web, Linux, and Windows. Endpoint data structure was SQL based, so parsing methods were in the SQL type: MySQL, MsSQL, and etc. And then transmission mixed non-SQL structure for parallel transmission. The OS for devices were also different so Object-C, Java/ Javascript, .Net were used to send data to the server. The data was transferred to server in SQL format. The Viewer in the HH service has its own communication pathway so image/ text delivery was fast.
Market Research
Tele-radiology was not popular in South Korea then. The company was a pioneer of this field. Because the competition between doctors and institutions in South Korea is intense and the accessibility to hospital is good, the telemedicine was not as popular as the telemedicine in the US. Thus, The most important point was developing/ convincing its needs.
Wireframe Key Products
DICOM Viewer
I was involved in the viewer's refining phase. Most of radiologist's needs were applied to the viewer product. I analyzed existing UIs and advised usability improvements.
Agent
The Agent should support all the environment regardless of the size of hospitals. So, I designed the Agent to have basically two functions: Automatic Request, Manual Request. For relatively big hospitals, The less displaying UI, The better program the Agent is for the users. In that case, automatic Agent is needed. On the other hand, the Agent’s UI is not excluded to users for the relatively small hospitals.
Web Application
The service is intertwined with three key functions: Agent, Web, Diagnosis tool. The wireframe was based on these requirements. I enhanced usability and functionality of the viewer's web request, statistics, print/ export reads UIs.
- Check the Agent successfully upload DICOM images.
- Check the Web server receive and store the images.
- Check the images are visible to Diagnosis tool.
- Check the Web server receive and update status of the image when the image diagnosed.
- Check the Agent updates the images’ status and receive diagnosis.
The service has been grown to take 60 percent of domestic market (South Korea) share in the Tele-radiology service. The service is now released as HTelerad for the teleradiology service, HPACS for the DICOM Image viewer, HScan and HCase for the web application. These products are for B2B except for HScan.
I learned how a cloud based service flow, what are needed to build the cloud based service, and what would be the key considerations to design user centered products.