Brain tumor automatic segmentation method through fusion of full convolutional neural network and conditional random field

Abstract

The present invention belongs to the computer-assisted medical field, and especially relates to a brain tumor automatic segmentation method through fusion of a full convolutional neural network and a conditional random field. The objective of the invention is to solve the problem that the depth learning technology cannot ensure the continuity of a segmentation result on the appearance and the space when the brain tumor segmentation is performed in the prior art. In order to solve the problem mentioned above, the method comprises the following steps: the step 1, employing a non-uniform offset correction and luminance regularization method to process the magnetic resonance image of the brain tumor image to generate a second magnetic resonance image; and the step 2, employing the neural network fusing the full convolutional neural network and the conditional random field to perform brain tumor segmentation of the second magnetic resonance image and output the brain tumor segmentation result. The brain tumor automatic segmentation method through fusion of the full convolutional neural network and the conditional random field can perform end-to-end brain tumor segmentation slice to slice and has higher operation efficiency.

Description

technical field [0001] The invention relates to an image segmentation method, in particular to an automatic brain tumor segmentation method that integrates a fully convolutional neural network and a conditional random field. Background technique [0002] Brain tumors have a high incidence rate, especially among the malignant lesions that children are susceptible to, brain tumors are second only to leukemia, ranking second. For brain tumors, whether benign or malignant, they can increase intracranial pressure, compress brain tissue, cause damage to the central nervous system, and endanger the lives of patients. [0003] The location of brain tumor lesion tissue and quantitative calculation (such as calculating tumor volume, diameter, etc.) are very important for the diagnosis of brain tumors, formulation of treatment plans, and efficacy monitoring. In the clinic, radiologists usually segment tumors manually from multimodal magnetic resonance images, which is a tedious and ti...

AI Technical Summary

Problems solved by technology

[0008] In order to solve the above-mentioned problems in the prior art, that is, in order to solve the problem that the current deep learning technology cannot guarantee the continuity of the segmentation results in ap

Images