Abstract:
|
Ischemic stroke is the third most frequent cause of death and a major cause of disability in industrial countries. In clinical practice, Diffusion weighted images (DWI), T1-weighted (T1W), T2-weighted (T2W) and fluid attenuated inversion recovery (FLAIR) images are often acquired to diagnose and monitor disease progression of strokes. However, manual segmentation of stroke lesions from these brain images is often a challenging and time consuming task and can only be performed by trained clinicians. In this paper, we propose an automated method to locate, segment and quantify stroke lesion areas using a new bias correction algorithm and a multi-scale 3D spatial point process clustering model. We evaluate the performance of the proposed model using the Ischemic Stroke Lesion Segmentation Challenge data.
|