Image reconstruction method, device and system in magnetic resonance imaging and storage medium

Abstract

The invention provides an image reconstruction method, device and system in magnetic resonance imaging and a computer storage medium. The method comprises the steps: for each to-be-executed image reconstruction task, enabling the magnetic resonance system to calculate a calculation capability demand value of the task; judging whether the computing power demand value of the task is greater than a preset threshold value or not, and if so, sending the task to a shared image reconstruction device, so as to enable the shared image reconstruction device to execute the task; otherwise, sending the task to a local image reconstruction device, so as to enable the local image reconstruction device to execute the task, wherein the computing performance of the shared image reconstruction device is higher than that of the local image reconstruction device, and the shared image reconstruction device is shared by the plurality of magnetic resonance systems. According to the invention, the hardware cost required for image reconstruction in MRI is reduced.

Description

technical field [0001] The present invention relates to the technical field of MRI (Magnetic Resonance Imaging, magnetic resonance imaging), in particular to an image reconstruction method, device, system and storage medium in MRI. Background technique [0002] MRI technology has become a very useful tool in medical diagnosis. The MRI hardware system mainly includes: magnet subsystem, magnetic field gradient (referred to as gradient) subsystem, radio frequency subsystem, spectrometer subsystem, and main computer. Among them, the radio frequency subsystem mainly includes: transmitting coil and receiving coil, and the spectrometer subsystem mainly includes: pulse sequence generator, gradient waveform generator, transmitter and receiver, etc. [0003] During the imaging process, under the control of the pulse sequence generator, the transmitter outputs a radio frequency pulse signal to the radio frequency transmitting coil to generate a radio frequency field for exciting the h...

AI Technical Summary

Problems solved by technology

And powerful computing performance of CPU and GPU means high cost

[0005] On the other hand, the powerful computer performance in the current system has not been fully utilized

For example: during a 3-hour volunteer scanning process, the actual CPU usage in the MRI image reconstruction machine is: the time slices with CPU usage below 5% accounted for 98% of the entire scanning time period, which means high cost The cost of hardware is actually only fully utilized less than 2% of the time, and for customers with multiple systems, the overall idleness of hardware resources will be more obvious

Images