| 作品介紹 |
Acute Ischemic Stroke (AIS), a leading cause of mortality in Taiwan, is typically assessed in the emergency department using non-contrast computed tomography (NCCT) to locate affected regions. However, detecting ischemic changes on NCCT within the first 12 hours post-stroke is challenging due to low contrast and the variability in ischemic lesion size and location across different slices and patients. To address these challenges, this work introduces DLAIS—a deep learning software tool designed for the automatic identification of ischemic slices, precise localization of stroke lesions, and accurate delineation of infarction boundaries throughout the brain. DLAIS provides a comprehensive three-dimensional view of brain tissue, rendering healthy tissue in yellow and highlighting infarcted regions in red, thus enhancing visual clarity and improving clinicians' ability to perceive ischemic locations swiftly. |