An image reconstruction method suitable for low-field magnetic resonance

By introducing multiple auxiliary coils into a low-field magnetic resonance system, acquiring and processing the distribution of environmental noise, and combining the transfer function and prior information model, the imaging quality problem caused by the variability of noise sources in open low-field magnetic resonance was solved, and higher quality image reconstruction was achieved.

CN117169797BActive Publication Date: 2026-11-17SHANGHAI SOUNDWISE TECHNOLOGY CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202310989759.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-07
Publication Date
2026-11-17
Estimated Expiration
2043-08-07

AI Technical Summary

Technical Problem

In existing technologies, the location and intensity of noise sources in open low-field magnetic resonance imaging are easily variable in complex electromagnetic environments, and traditional noise cancellation methods are ineffective, leading to deterioration of imaging quality.

Method used

Multiple auxiliary coils are set up in the low-field magnetic resonance system. The overall environmental noise distribution is generated by collecting auxiliary environmental noise. Noise estimation and image reconstruction are performed by combining the transfer function and prior information model.

Benefits of technology

It achieves accurate noise estimation in complex scenes, thus improving image reconstruction quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117169797B_ABST
    Figure CN117169797B_ABST
Patent Text Reader

Abstract

This invention relates to the field of magnetic resonance imaging (MRI) technology, specifically to an image reconstruction method suitable for low-field MRI, comprising: step S1: acquiring auxiliary environmental noise on the auxiliary coils respectively, and generating an overall environmental noise distribution based on the auxiliary environmental noise; step S2: generating estimated noise corresponding to the main coil based on the overall environmental noise distribution; step S3: acquiring the main echo sequence using the main coil, and performing image reconstruction based on the estimated noise and the main echo sequence. The beneficial effect is that, addressing the problem of poor performance of existing noise estimation schemes in open low-field MRI scenarios, this solution introduces multiple auxiliary coils into the MRI system, obtains the overall environmental noise distribution through signal acquisition and processing based on the auxiliary coils, and then estimates the noise of the main coil based on the overall environmental noise distribution, achieving better noise estimation results in complex scenarios and contributing to image reconstruction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of magnetic resonance imaging technology, and more specifically to an image reconstruction method suitable for low-field magnetic resonance. Background Technology

[0002] MRI is a commonly used medical computed tomography method that uses the magnetic resonance phenomenon to obtain electromagnetic signals from the human body and reconstruct human information. Compared with high-field MRI, low-field MRI has advantages such as lower examination costs and easier accessibility. Open-type MRI, which does not require electromagnetic shielding, can provide patients with a comfortable scanning experience. However, open-type MRI is susceptible to external electromagnetic interference, which can affect image quality.

[0003] In the prior art, there are many solutions for estimating and removing noise to achieve better image reconstruction results in response to environmental noise problems. For example, Chinese patent CN201910649552.8 discloses a method for removing black stripes in low-field magnetic resonance linear scanning diffusion imaging. This solution eliminates the influence of environmental noise on imaging by pre-acquiring a background noise image before scanning and performing image reconstruction based on a subtraction method.

[0004] However, during implementation, the inventors discovered that since open-field low-field magnetic resonance imaging is often used in scenarios such as ambulances and wards, the electromagnetic environment is relatively complex. The location and intensity of noise sources are prone to change during the scanning process compared to fixed scenarios. Traditional noise cancellation methods usually cannot solve this problem well, which can easily lead to image quality degradation. Summary of the Invention

[0005] To address the aforementioned problems in existing technologies, an image reconstruction method suitable for low-field magnetic resonance is provided.

[0006] The specific technical solution is as follows:

[0007] An image reconstruction method suitable for low-field magnetic resonance imaging (MRI) is provided, applicable to a low-field MRI system, wherein the low-field MRI system is equipped with a main coil and multiple auxiliary coils, and the image reconstruction method includes:

[0008] Step S1: Collect auxiliary ambient noise on the auxiliary coil respectively, and generate the overall ambient noise distribution according to the auxiliary ambient noise;

[0009] Step S2: Generate an estimated noise corresponding to the main coil based on the overall environmental noise distribution;

[0010] Step S3: Acquire the main echo sequence using the main coil, and reconstruct the image based on the estimated noise and the main echo sequence.

[0011] On the other hand, prior to step S1, a coil coordinate system was pre-constructed for the main coil, and the auxiliary coil was calibrated in the coil coordinate system;

[0012] Step S1 includes:

[0013] Step S11: Collect the auxiliary ambient noise for each of the auxiliary coils;

[0014] Step S12: Map each group of auxiliary environmental noise according to the position of the auxiliary coil in the coil coordinate system;

[0015] Step S13: Plot the power spectral density in the coil coordinate system according to all the auxiliary environmental noises to obtain the overall environmental noise distribution.

[0016] On the other hand, step S2 includes:

[0017] Step S21: Determine the intermediate estimated noise corresponding to the main coil based on the overall environmental noise distribution;

[0018] Step S22: Perform frequency correction on the intermediate estimated noise based on the transfer function to obtain the estimated noise.

[0019] On the other hand, step S21 includes:

[0020] Step S211: Extract the noise corresponding to the center sampling point of the main coil and multiple center region noises associated with the main coil from the overall environmental noise distribution;

[0021] Step S212: Construct a signal matrix based on the noise in the central region, and construct an offset matrix based on the coordinates of the auxiliary sampling points corresponding to the noise in the central region;

[0022] Step S213: Correct the noise at the center sampling point based on the signal matrix and the offset matrix to obtain the intermediate estimated noise.

[0023] On the other hand, in step S211, multiple sets of noise from the central sampling point and noise from the central region are collected respectively to form a noise set;

[0024] In step S213, corrections are performed on each of the noise sets to obtain set-corrected noise, and the intermediate estimated noise is generated based on all the set-corrected noise.

[0025] The auxiliary sampling points selected in each noise set are different.

[0026] On the other hand, before performing step S1, the first frequency band parameter of the main coil and the second frequency band parameter of each of the auxiliary coils are measured in advance, and the transfer function is constructed according to the first frequency band parameter and the second frequency band parameter.

[0027] On the other hand, step S3 includes:

[0028] Step S31: Determine the environmental noise level by performing a threshold value determination based on the estimated noise, and acquire the main echo sequence using the main coil;

[0029] Step S32: Call the corresponding prior information model according to the environmental noise level;

[0030] Step S33: Input the main echo sequence into the prior information model for image reconstruction.

[0031] On the other hand, before performing step S1, prior noise information of multiple noise sources at different locations is collected under different environmental noise levels, and a prior information model under the current environmental noise level is constructed based on the prior noise information.

[0032] On the other hand, in step S3, the scanning frequency range of the main coil is adjusted according to the noise frequency range of the estimated noise, and the main echo sequence is acquired according to the scanning frequency range.

[0033] The above technical solution has the following advantages or beneficial effects:

[0034] To address the issue that existing noise estimation schemes are ineffective in open low-field magnetic resonance imaging (MRI) scenarios, this solution introduces multiple auxiliary coils into the MRI system. Based on the auxiliary coils, signal acquisition and processing are performed to obtain the overall environmental noise distribution. Then, based on the overall environmental noise distribution, noise estimation is performed on the main coil, achieving better noise estimation results in complex scenarios, which is helpful for image reconstruction. Attached Figure Description

[0035] Embodiments of the invention will be described more fully with reference to the accompanying drawings. However, the drawings are for illustration and explanation only and do not constitute a limitation on the scope of the invention.

[0036] Figure 1 This is an overall schematic diagram of an embodiment of the present invention;

[0037] Figure 2 This is a schematic diagram of sub-step S1 in an embodiment of the present invention;

[0038] Figure 3 This is a schematic diagram of sub-step S2 in an embodiment of the present invention;

[0039] Figure 4 This is a schematic diagram of sub-step S21 in an embodiment of the present invention;

[0040] Figure 5 This is a schematic diagram of step S3 in an embodiment of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0043] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention.

[0044] This invention includes:

[0045] An image reconstruction method suitable for low-field magnetic resonance imaging (MRI) is proposed, applicable to low-field MRI systems, which include a main coil and multiple auxiliary coils, such as... Figure 1 As shown, image reconstruction methods include:

[0046] Step S1: Collect auxiliary ambient noise on the auxiliary coil respectively, and generate the overall ambient noise distribution based on the auxiliary ambient noise;

[0047] Step S2: Generate estimated noise corresponding to the main coil based on the overall environmental noise distribution;

[0048] Step S3: Acquire the main echo sequence using the main coil, and reconstruct the image based on the estimated noise and the main echo sequence.

[0049] To address the issue of poor performance of existing noise estimation schemes in open low-field MRI scenarios, this embodiment incorporates multiple auxiliary coils associated with the main coil in the low-field MRI system. Before the actual scan begins, auxiliary environmental noise is pre-collected at various locations using the auxiliary coils, and an overall environmental noise distribution corresponding to a large area is generated based on this auxiliary environmental noise. This overall environmental noise distribution can effectively characterize the electromagnetic noise situation over a large area. Based on this overall environmental noise distribution, the environmental noise at the main coil location can be more accurately estimated, yielding the estimated noise. The main coil noise is then used for searching, the main echo sequence of the patient is acquired, and a high-quality image reconstruction process is performed by combining the estimated noise.

[0050] In implementation, the aforementioned image reconstruction method is primarily used as a software implementation in a low-field magnetic resonance imaging (MRI) system to acquire and process input signals for image reconstruction. Auxiliary coils refer to coils deployed at multiple locations and directions within the MRI system, primarily used for acquiring environmental noise signals. These coils have different dimensions from the main coil, and their frequency ranges may also deviate depending on the coil design. Auxiliary environmental noise refers to the environmental noise signal sequence acquired at each location using each auxiliary coil. The overall environmental noise distribution is noise distribution data fitted based on the auxiliary environmental noise. This data can characterize the strength, location, and frequency range of noise signals in the environment over a relatively large range, thereby fitting the estimated noise at the location of the main coil. The main coil is used for scanning and imaging the patient, and the main echo sequence is the echo signal acquired by this main coil.

[0051] In one embodiment, before step S1, a coil coordinate system is pre-constructed for the main coil, and the auxiliary coil is calibrated in the coil coordinate system;

[0052] like Figure 2 As shown, step S1 includes:

[0053] Step S11: Collect the auxiliary ambient noise for each auxiliary coil;

[0054] Step S12: Map each group of auxiliary environmental noise according to the position of the auxiliary coil in the coil coordinate system;

[0055] Step S13: Plot the power spectral density in the coil coordinate system according to all auxiliary environmental noises to obtain the overall environmental noise distribution.

[0056] Specifically, to achieve a more accurate noise distribution fitting effect, in this embodiment, a coil coordinate system is established in advance based on the spatial positional relationship between the main coil and the auxiliary coil, and the auxiliary coil is calibrated in the coil coordinate system to obtain the position of the auxiliary coil mapped in the coil coordinate system. Based on this, after collecting auxiliary environmental noise, the auxiliary environmental noise can be mapped to the coil coordinate system according to the position of the auxiliary coil. Subsequently, based on multiple sets of auxiliary environmental noise, the power spectral density at each point in the coil coordinate system can be fitted, and then combined to obtain the overall environmental noise distribution, thus achieving a characterization of the environmental noise.

[0057] In one embodiment, such as Figure 3 As shown, step S2 includes:

[0058] Step S21: Determine the intermediate estimated noise corresponding to the main coil based on the overall environmental noise distribution;

[0059] Step S22: Perform frequency correction on the intermediate estimated noise based on the transfer function to obtain the estimated noise.

[0060] Specifically, considering that the frequency range of the auxiliary coil may deviate from that of the main coil due to the coil design, in this embodiment, after estimating the intermediate estimated noise based on the overall environmental noise, the intermediate estimated noise is further corrected based on the transfer function and transformed to the frequency range corresponding to the main coil to obtain the actual estimated noise, thereby making the subsequent noise removal and image reconstruction process more accurate.

[0061] In one embodiment, such as Figure 4 As shown, step S21 includes:

[0062] Step S211: Extract the noise corresponding to the center sampling point of the main coil and multiple center region noises associated with the main coil from the overall environmental noise distribution;

[0063] Step S212: Construct a signal matrix based on the noise in the central region, and construct an offset matrix based on the coordinates of the auxiliary sampling points corresponding to the noise in the central region;

[0064] Step S213: Correct the noise at the center sampling point based on the signal matrix and the offset matrix to obtain the intermediate estimated noise.

[0065] Specifically, to achieve a more accurate noise estimation, in this embodiment, a central sampling point is selected in the coil coordinate system for the main coil, and multiple auxiliary sampling points associated with the central sampling point are collected in any direction according to a specific distance interval outside the central sampling point. Based on the central sampling point and the auxiliary sampling points, the corresponding central sampling point noise and central region noise can be extracted from the overall noise distribution. Subsequently, a corresponding signal matrix is ​​constructed based on the central region noise. This signal matrix can correct the central sampling point noise obtained by direct fitting. At the same time, considering that the centroid of the point set composed of auxiliary sampling points may be offset from the coordinates of the central region, an offset matrix is ​​also introduced for correction, thereby achieving a more accurate noise correction process.

[0066] In one embodiment, in step S211, multiple sets of noise from the center sampling points and noise from the center region are collected respectively to form a noise set;

[0067] In step S213, corrections are performed on each noise set to obtain set-corrected noise, and intermediate estimated noise is generated based on all set-corrected noise.

[0068] The auxiliary sampling points selected in each noise set are different.

[0069] Specifically, to achieve better noise estimation results, in this embodiment, auxiliary sampling points are selected multiple times, and noise sets are constructed for each set. During the noise estimation process, estimation can be performed separately for each noise set to obtain set-corrected noise. The actual intermediate estimated noise is then calculated based on the set-corrected noise, thereby improving the accuracy of the fit.

[0070] In one embodiment, before performing step S1, the first frequency band parameters of the main coil and the second frequency band parameters of each auxiliary coil are measured in advance, and the transfer function is constructed according to the first frequency band parameters and the second frequency band parameters.

[0071] Specifically, in order to achieve a more accurate noise estimation effect in the frequency domain, in this embodiment, before the acquisition begins, the first frequency band parameters and the second frequency band parameters of the main coil and the auxiliary coil are acquired in advance. Then, a transfer function is constructed according to the first frequency band parameters and the second frequency band parameters to achieve a more accurate transformation process of noise in the frequency domain.

[0072] In one embodiment, before executing step S1, prior noise information of multiple noise sources at different locations is collected under different environmental noise levels, and a prior information model under the current environmental noise level is constructed based on the prior noise information.

[0073] like Figure 5 As shown, step S3 includes:

[0074] Step S31: Determine the environmental noise level based on the estimated noise threshold, and acquire the main echo sequence using the main coil;

[0075] Step S32: Call the corresponding prior information model according to the environmental noise level;

[0076] Step S33: Input the main echo sequence into the prior information model for image reconstruction.

[0077] Specifically, to achieve better image reconstruction results, in this embodiment, multiple prior information models are pre-constructed for different environmental noise levels before the acquisition begins. These prior information models can, based on prior information about noise sources in different locations, combine noise signals removed from the actual input estimated noise auto-echo sequence before image reconstruction. Based on this setup, after obtaining the estimated noise, a threshold value can be determined to assess the current environmental noise level; subsequently, the appropriate prior information model is selected for processing according to the environmental noise level, thereby achieving a better image reconstruction process.

[0078] In one embodiment, in step S3, the scanning frequency range of the main coil is adjusted according to the noise frequency range of the estimated noise, and the main echo sequence is acquired according to the scanning frequency range.

[0079] Specifically, in order to achieve better scanning results, in this embodiment, after obtaining the estimated noise, the scanning frequency range of the main coil is further adjusted according to the noise frequency range where the estimated noise is mainly distributed, so that the scanning frequency of the main coil deviates from the noise frequency, thereby achieving better scanning results.

[0080] The above are merely preferred embodiments of the present invention and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should recognize that any equivalent substitutions and obvious changes made based on the description and illustrations of the present invention should be included within the protection scope of the present invention.

Claims

1. An image reconstruction method suitable for low-field magnetic resonance, characterized in that, Applicable to low-field magnetic resonance systems, wherein the low-field magnetic resonance system is equipped with a main coil and multiple auxiliary coils, the image reconstruction method includes: Step S1: Collect auxiliary ambient noise on the auxiliary coil respectively, and generate the overall ambient noise distribution according to the auxiliary ambient noise; Step S2: Generate an estimated noise corresponding to the main coil based on the overall environmental noise distribution; Step S3: The main coil is used to acquire the main echo sequence, and image reconstruction is performed based on the estimated noise and the main echo sequence; before step S1, a coil coordinate system is pre-constructed for the main coil, and the auxiliary coil is calibrated in the coil coordinate system; Step S1 includes: Step S11: Collect the auxiliary ambient noise for each of the auxiliary coils; Step S12: Map each group of auxiliary environmental noise according to the position of the auxiliary coil in the coil coordinate system; Step S13: Plot the power spectral density in the coil coordinate system according to all the auxiliary environmental noises to obtain the overall environmental noise distribution; Step S2 includes: Step S21: Determine the intermediate estimated noise corresponding to the main coil based on the overall environmental noise distribution; Step S22: Perform frequency correction on the intermediate estimated noise based on the transfer function to obtain the estimated noise; Step S21 includes: Step S211: Extract the noise corresponding to the center sampling point of the main coil and multiple center region noises associated with the main coil from the overall environmental noise distribution; Step S212: Construct a signal matrix based on the noise in the central region, and construct an offset matrix based on the coordinates of the auxiliary sampling points corresponding to the noise in the central region; Step S213: Correct the noise at the center sampling point based on the signal matrix and the offset matrix to obtain the intermediate estimated noise.

2. The image reconstruction method of claim 1, wherein, In step S211, multiple sets of noise from the central sampling point and noise from the central region are collected respectively to form a noise set; In step S213, corrections are performed on each of the noise sets to obtain set-corrected noise, and the intermediate estimated noise is generated based on all the set-corrected noise. The auxiliary sampling points selected in each noise set are different.

3. The image reconstruction method of claim 1, wherein, Before performing step S1, the first frequency band parameters of the main coil and the second frequency band parameters of each of the auxiliary coils are measured in advance, and the transfer function is constructed according to the first frequency band parameters and the second frequency band parameters.

4. The image reconstruction method of claim 1, wherein, Step S3 includes: Step S31: Determine the environmental noise level by performing a threshold value determination based on the estimated noise, and acquire the main echo sequence using the main coil; Step S32: Call the corresponding prior information model according to the environmental noise level; Step S33: Input the main echo sequence into the prior information model for image reconstruction.

5. The image reconstruction method of claim 4, wherein, Before performing step S1, prior noise information of multiple noise sources at different locations is collected under different environmental noise levels, and a prior information model under the current environmental noise level is constructed based on the prior noise information.

6. The image reconstruction method of claim 1, wherein, In step S3, the scanning frequency range of the main coil is adjusted according to the noise frequency range of the estimated noise, and the main echo sequence is acquired according to the scanning frequency range.

Citation Information

Patent Citations

  • Method for removing black stripes by low-field magnetic resonance line scan diffusion imaging

    CN110367985A

  • Noise suppression methods and apparatus

    CN107110930A

  • Internal ultrasound assisted local therapeutic delivery

    CN112584774A