SAR value acquisition method in MRI, SAR value acquisition system, electronic apparatus, and storage medium

A specific absorption rate (SAR) value acquisition method in magnetic resonance imaging (MRI), comprising:

• acquiring a mapping relation between distribution data of radio frequency (RF) coil transmission field of each channel of an RF transmitter coil and distribution data of RF electric field of each channel of the RF transmitter coil, and obtaining a mapping relation set;
• acquiring first distribution data of RF coil transmission field of each channel during an examination of a test person;
• determining first distribution data of RF electric field of each channel during the examination of the test person according to the first distribution data of RF coil transmission field of each channel and the mapping relation set;
• and
• adjusting a weight of each channel according to the first distribution data of RF electric field of each channel acquired during the examination of the test person, and determining a distribution of SAR values of the test person;
• wherein the mapping relation between the distribution data of RF coil transmission field of each channel of the RF transmitter coil and the distribution data of RF electric field of each channel of the RF transmitter coil is a trained RF electric field prediction model;
• wherein acquiring the trained RF electric field prediction model comprises:
• constructing a human-model data set;
• obtaining second distribution data of RF coil transmission field of each channel and second distribution data of RF electric field of each channel for each human model through simulation, based on the constructed human-model data set and a channel parameter of the RF transmitter coil, to obtain a training data set;
• and
• constructing and training an RF electric field prediction model according to the training data set until a preset condition for an end of training is met, and obtaining the trained RF electric field prediction model;
• wherein each human model in the human-model data set comprises electromagnetic parameters of human tissue;
• and
• the constructing the human-model data set comprises:
• obtaining electromagnetic parameters of human tissue of at least two typical human models selected from a human-model library;
• augmenting the electromagnetic parameters of human tissue of the at least two typical human models, constructing a new human model based on the augmented electromagnetic parameters, and adding the new human model in the typical human models;
• executing iteratively steps of:
• performing the augmentation processing on the electromagnetic parameters of human tissue of the at least two typical human models, constructing the new human model based on the augmented electromagnetic parameters, and adding the new human model in the typical human models, until a preset target for constructing human models is satisfied;
• and
• obtaining the human-model data set based on the typical human models.

A specific absorption rate (SAR) value acquisition method in magnetic resonance imaging (MRI), comprising:

• acquiring a mapping relation between distribution data of radio frequency (RF) coil transmission field of each channel of an RF transmitter coil and distribution data of RF electric field of each channel of the RF transmitter coil, and obtaining a mapping relation set;
• acquiring first distribution data of RF coil transmission field of each channel during an examination of a test person;
• determining first distribution data of RF electric field of each channel during the examination of the test person according to the first distribution data of RF coil transmission field of each channel and the mapping relation set;
• and
• adjusting a weight of each channel according to the first distribution data of RF electric field of each channel acquired during the examination of the test person, and determining a distribution of SAR values of the test person;
• wherein the mapping relation between the distribution data of RF coil transmission field of each channel of the RF transmitter coil and the distribution data of RF electric field of each channel of the RF transmitter coil is a trained RF electric field prediction model;
• wherein acquiring the trained RF electric field prediction model comprises:
• constructing a human-model data set;
• obtaining second distribution data of RF coil transmission field of each channel and second distribution data of RF electric field of each channel for each human model through simulation, based on the constructed human-model data set and a channel parameter of the RF transmitter coil, to obtain a training data set;
• and
• constructing and training an RF electric field prediction model according to the training data set until a preset condition for an end of training is met, and obtaining the trained RF electric field prediction model;
• wherein, the obtaining the second distribution data of RF coil transmission field of each channel and the second distribution data of RF electric field of each channel for each human model through simulation, based on the constructed human-model data set and the channel parameter of the RF transmitter coil, to obtain the training data set, comprises:
• determining the number of scan segments and a scan part corresponding to each of the scan segments of each human model according to dimensions of the RF transmitter coil;
• performing segmented scans on each human model according to the channel parameter of the RF transmitter coil, the number of the scan segments of each human model, and the scan part corresponding to each of the scan segments, and obtaining third distribution data of RF coil transmission field and third distribution data of RF electric field of the scan part corresponding to each scan segment of each human model;
• and
• obtaining the second distribution data of RF coil transmission field of each channel and the second distribution data of RF electric field of each channel for all tissues of each human model, based on the third distribution data of RF coil transmission field and the third distribution data of RF electric field of the scan part corresponding to each of the scan segments of each human model.