Black plastid-1 sign automatic identification method and system of multi-task neural network based on multi-echo GRE sequence

Abstract

The invention discloses a black plastil-1 sign automatic identification method and system of a multi-task neural network based on a multi-echo GRE sequence. The method comprises the steps of magnetic resonance data preprocessing, brain nucleus segmentation, black plastil-1 sign identification key area positioning and black plastil-1 sign automatic identification. The system utilizes a multi-echo GRE magnetic resonance sequence to reconstruct an image clearly displaying brain nigra, utilizes a deep learning technology to segment a nigra nucleus, and carries out automatic identification and positioning of a nigra-plastid-1 sign on a segmented nigra area; the reconstructed image for clearly displaying the brain nigra comprises a trueSWI image, a T2 * image, an R2 * image, a magnetic sensitive weighted image and a quantitative magnetic susceptibility image. According to the invention, priori knowledge of clinicians is integrated in the deep learning method, automatic positioning and accurate discrimination of the black plastil-1 sign recognition effective area are realized, and the efficiency and accuracy of black plastil-1 sign interpretation are greatly improved.

Description

technical field [0001] The invention belongs to the field of image processing and medical technology, and relates to an automatic identification method and system for identifying substantia nigra-1 signs in magnetic resonance images using deep learning technology. Background technique [0002] Parkinson's disease (PD) is a chronic progressive neurodegenerative disease with a prevalence rate of 1.7% among people over 65 years old. It has a long disease cycle and a high disability rate, which is harmful to the aging society. [0003] Magnetic resonance imaging (MRI) has the characteristics of no radiation, high tissue resolution, multi-sequence and multi-parameter imaging, and is the most commonly used means of auxiliary examination for PD. However, the structural changes shown by conventional brain MRI have little diagnostic value for Parkinson's disease. The increase of iron deposition in the substantia nigra plays a very important role in the occurrence and development of ...

AI Technical Summary

Problems solved by technology

[0005] However, the current identification of "Nigral body-1 sign" is mainly based on the manual interpretation of imaging doctors, and its subjectivity is unavoidable

At the same time, manual recognition can only make a rough judgment on the presence or absence of this sign, and it is difficult to carry out more accurate quantitative measurement, which limits its application in the study of the occurrence and development of PD diseases

Images