Magnetic particle imaging reconstruction method based on attention mechanism

Abstract

The invention discloses a magnetic particle imaging reconstruction method based on an attention. According to the method, the strong computing power of a neural network model is utilized to match acquired large data volume, and effective information in signals is extracted by fusing self-attention mechanism learning, so that signal information loss caused by downsampling or truncation is reduced; and end-to-end reconstruction from a one-dimensional frequency domain signal to a two-dimensional image is realized.

Description

technical field [0001] The invention belongs to the field of magnetic particle imaging, and in particular relates to a magnetic particle imaging reconstruction method from a one-dimensional frequency domain signal to a two-dimensional image based on a deep learning self-attention mechanism. Background technique [0002] In clinical diagnosis and detection, how to accurately and objectively locate tumors and other lesions has always been an international research hotspot and challenging issue. Traditional medical imaging technologies such as CT, MRI, SPECT and other methods have problems such as high hazards, poor positioning, and low precision. In recent years, a new tracer-based imaging method, Magnetic Particle Imaging (MPI), has been proposed, which uses tomographic imaging to detect superparamagnetic iron oxide nanoparticles (SPIOs) that are harmless to the human body. The spatial concentration distribution of the target can be precisely positioned, and it has the chara...

AI Technical Summary

Problems solved by technology

[0004] However, there are obvious deficiencies in both traditional methods and deep learning methods, which are mainly reflected in the following aspects: (1) Due to the large amount of original collected data, existing methods, especially traditional methods, need to preprocess the data, which increases the complexity of reconstruction. (2) The existing preprocessing methods include downsampling or truncation, which are all artificial preprocessing methods, which will lead to the loss of sample information in the data, and eventually cause the loss of the quality of the reconstructed image; (3) due to different conditions The lengths of the collected signals are different, and there is a lack of a general processing method for extracting key information from different signals

Images