Invention Grant
- Patent Title: Automatic brain infarction detection system on magnetic resonance imaging and operation method thereof
-
Application No.: US16884070Application Date: 2020-05-27
-
Publication No.: US11042983B1Publication Date: 2021-06-22
- Inventor: Jang-Zern Tsai , Syu-Jyun Peng , Yu-Wei Chen , Meng-Zong Tsai , Kuo-Wei Wang , Yeh-Lin Kuo
- Applicant: National Central University , Huan-Cheng Chang , Taipei Medical University (TMU)
- Applicant Address: TW Taoyuan; TW Taoyuan; TW Taipei
- Assignee: National Central University,Huan-Cheng Chang,Taipei Medical University (TMU)
- Current Assignee: National Central University,Huan-Cheng Chang,Taipei Medical University (TMU)
- Current Assignee Address: TW Taoyuan; TW Taoyuan; TW Taipei
- Agency: CKC & Partners Co., LLC
- Priority: TW109112042 20200409
- Main IPC: G06K9/00
- IPC: G06K9/00 ; G06T7/00 ; G06T5/00

Abstract:
The present disclosure provides an operating method of an automatic brain infarction detection system on magnetic resonance imaging (MRI), which includes steps as follows. Images corresponding to different slices of a brain of a subject are received from the MRI machine. The image mask process is performed on first and second images of the images. It is determined whether the cerebellum image intensity and the brain image intensity in the first image are matched. When the cerebellum image intensity and the brain image intensity are not matched, the cerebellar image intensity in the first image is adjusted. The first image is processed through a nonlinear regression to obtain a third image. A neural network identify an infarct region by using the first, second and third images that are cut.
Information query