Tumor brain region positioning method and device based on consistency loss

Abstract

The invention relates to a tumor brain region positioning method and device based on consistency loss. The method comprises the steps: removing a focus region according to position information of the focus region in a to-be-processed image, and obtaining a first image; performing region completion on the first image through a generative network to obtain a second image; performing region segmentation on the second image through a partition network to obtain a partition map; and according to the position information of the focus region in the to-be-processed image and the partition map, determining a partition to which the focus region belongs. According to the tumor brain region positioning method based on the consistency loss, the second image without a focus can be simulated by complementing the first image after the focus region is removed, and region segmentation can be performed on the second image without the focus, so that the accuracy of partitioning is improved. Therefore, the partition to which the position of the focus region belongs can be determined, the accuracy of determining the partition to which the focus belongs is improved, and a basis can be provided for diagnosis of the focus.

Description

technical field [0001] The present disclosure relates to the technical field of image processing, and in particular to a method and device for locating tumor brain regions based on consistency loss. Background technique [0002] Brain tumors mostly grow in the cranial cavity, also known as intracranial tumors and brain cancers, which can originate from the brain, meninges, nerves, blood vessels, and brain appendages, or form by metastasizing from other tissues or organs of the body into the brain. [0003] Brain tumors are classified according to different locations, shapes and other factors. They are mainly divided into intracerebral tumors and extracerebral tumors. Brain here is the abbreviation of brain parenchyma. The intraparenchymal refers to the cerebral cortex and the cerebral medulla; the extraparenchymal refers to the ventricles, subarachnoid space, dura mater tissue and even the skull. If it is determined that the tumor is an extracerebral tumor, the tumors that ...

AI Technical Summary

Problems solved by technology

[0005] MRI (Magnetic Resonance Imaging, Magnetic Resonance Imaging) is one of the imaging examinations for the evaluation of brain tumors. Different from common scenes in daily life, medical images will have low contrast, low signal-to-noise ratio, and light intensity due to the influence of image acquisition equipment. Low-level problems; the movement of organs during the scanning process will also cause image quality problems, so it brings certain difficulties to tumor location. In traditional tumor location, doctors generally make judgments, but people will be affected by factors such as energy Influenced by factors such as restrictions and mood swings, the accuracy of judgment is unstable

Images