Segmentation algorithm of intracranial hemorrhage based on mu-net applied to CT images

Abstract

The invention discloses an intracranial hemorrhage lesion segmentation algorithm applied to CT images based on MU‑Net. The intracranial hemorrhage lesion segmentation algorithm proposes a new segmentation structure MU‑Net based on U‑Net, and uses it Applied to the intracranial hemorrhage segmentation task. In the encoder module, the network module of Res2Net is introduced. Such a design can extract finer multi-scale features and increase the receptive field of feature maps. In order to reduce the semantic gap existing between the corresponding layers of the encoding layer and the decoding layer, a multi-encoding information fusion module (MIF) is proposed, which effectively compensates for the global information lost by the decoder through information fusion of features. Besides, in order to further reduce the semantic gap between encoder and decoder and gather multi-scale information, the present invention proposes a new decoder module (MDB).

Description

technical field [0001] The invention belongs to the field of intelligent medical image processing, and relates to an intracranial hemorrhage lesion segmentation algorithm applied to CT images based on MU-Net. Background technique [0002] Intracranial hemorrhage (ICH) is a bleeding disorder that occurs in the ventricle or brain tissue. Intracranial hemorrhage includes: intraventricular hemorrhage (IVH), parenchymal hemorrhage (IPH), subarachnoid hemorrhage (SAH), epidural hemorrhage (EDH) and subdural hematoma (SDH). Diagnosing intracranial hemorrhage is a major challenge in medicine. Determining the site of intracranial hemorrhage is key to treating the patient. The current method of diagnosing intracranial hemorrhage is a CT scan of the brain. Because tissues such as brain, blood, muscle, and bone absorb X-rays differently, CT scans produce high-contrast images that are then viewed by doctors. The research on the segmentation of intracranial hemorrhage focuses on the C...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem of poor U-Net segmentation performance in the field of intracranial hemorrhage lesion segmentation, the present invention provides an intracranial hemorrhage lesion segmentation algorithm based on MU-Net applied to CT images

Images