Page 82 - My FlipBook
P. 82
Collaborative Projects所
內
合
作
計
畫
Figure 1 : System diagram of 3D object segmentation.
Figure 2 : AUC for microbleed detection models.
Figure 3 : Detection of microbleeds by a 3D object detection model (top row=full image; bottom
row=enlargements). Red represents detected results; green is ground-truth results; orange is
colocalization of model-detected and ground-truth microbleeds. The numbers in the lower left
corner of the upper row of images represent di erent imaging aspects of the same patient.
3.2 Development of a stroke app
There are two objectives for this project; rst to develop a clear end-product and, second, to discover suitable
collaborative mechanisms. The project has three principle components, i.e., content, system implementation,
and a consent mechanism. National Yang Ming University is leading the content development aspect,
National Chiao Tung University is responsible for system implementation, and IIS is dealing with the consent
mechanism. The IIS team is leveraging previously developed authorization service design to implement a
dynamic consent mechanism and is currently engaged in recruiting patients to use this app and provide
health-related information for further research. The core idea of dynamic consent is to treat data providers
as active and continuous participants. Currently, participants tend to sign an informed consent form at the
80
內
合
作
計
畫
Figure 1 : System diagram of 3D object segmentation.
Figure 2 : AUC for microbleed detection models.
Figure 3 : Detection of microbleeds by a 3D object detection model (top row=full image; bottom
row=enlargements). Red represents detected results; green is ground-truth results; orange is
colocalization of model-detected and ground-truth microbleeds. The numbers in the lower left
corner of the upper row of images represent di erent imaging aspects of the same patient.
3.2 Development of a stroke app
There are two objectives for this project; rst to develop a clear end-product and, second, to discover suitable
collaborative mechanisms. The project has three principle components, i.e., content, system implementation,
and a consent mechanism. National Yang Ming University is leading the content development aspect,
National Chiao Tung University is responsible for system implementation, and IIS is dealing with the consent
mechanism. The IIS team is leveraging previously developed authorization service design to implement a
dynamic consent mechanism and is currently engaged in recruiting patients to use this app and provide
health-related information for further research. The core idea of dynamic consent is to treat data providers
as active and continuous participants. Currently, participants tend to sign an informed consent form at the
80
