公开(公告)号：US11816767B1

公开(公告)日：2023-11-14

申请号：US18218115

申请日：2023-07-05

Applicant: INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES

Inventor： Jie Tian ,  Zechen Wei ,  Hui Hui ,  Xin Yang ,  Huiling Peng

IPC: G06K9/00 ,  G06T11/00 ,  G06T5/00 ,  G06T5/10 ,  A61B5/0515

CPC classification number: G06T11/006 ,  A61B5/0515 ,  G06T5/002 ,  G06T5/10 ,  G06T2207/10072 ,  G06T2207/20081 ,  G06T2207/20084

Abstract: A method and system for reconstructing a magnetic particle distribution model based on time-frequency spectrum enhancement are provided. The method includes: scanning, by a magnetic particle imaging (MPI) device, a scan target to acquire a one-dimensional time-domain signal of the scan target; performing short-time Fourier transform to acquire a time-frequency spectrum; acquiring, by a deep neural network (DNN) fused with a self-attention mechanism, a denoised time-frequency spectrum; acquiring a high-quality magnetic particle time-domain signal; and reconstructing a magnetic particle distribution model. The method learns global and local information in the time-frequency spectrum through the DNN fused with the self-attention mechanism, thereby learning a relationship between different harmonics to distinguish between a particle signal and a noise signal. The method combines the global and local information to complete denoising of the time-frequency spectrum, thereby acquiring the high-quality magnetic particle time-domain signal.