Multi-b-value diffusion weighted MRI image processing method of IVIM and DKI combined model

Abstract

The invention discloses an IVIM and DKI combined model multi-b-value diffusion weighted MRI image processing method, and belongs to the field of MRI image processing. The method comprises the following steps: acquiring a human body image by using a multi-b-value diffusion weighted sequence, and fitting through an IVIM and DKI combined model to obtain a true diffusion coefficient distribution diagram, a pseudo diffusion coefficient distribution diagram, a perfusion fraction distribution diagram and a diffusion kurtosis diagram; delineating an ROI (Region of Interest) on the apparent diffusion coefficient image; based on the delineated ROI, lesion and peripheral regions are segmented on a true diffusion coefficient distribution diagram, a pseudo diffusion coefficient distribution diagram, a perfusion fraction distribution diagram and a diffusion kurtosis diagram respectively, and incoherent motion and diffusion kurtosis imaging parameter quantitative values in voxels of the lesion and the peripheral environment are calculated. Therefore, fitting, segmentation and quantification of the IVIM and DKI combined model are efficiently completed, the IVIM and DKI data processing flow is simplified, and meanwhile errors caused by IVIM and DKI image fractional acquisition and ROI sketching are avoided.

Description

Technical Field

[0001] The invention belongs to the field of magnetic resonance image (MRI) processing, and in particular relates to an IVIM and DKI joint model multi-b value diffusion weighted MRI image processing method. Background Art

[0002] At present, the intravoxel incoherent motion (IVIM) and diffusion kurtosis imaging (DKI) model sequences need to be collected separately or collected once, and then the diffusion-weighted images with specific b values ​​are extracted through third-party software, and analyzed on the MRI (magnetic resonance imaging) workstation through the software provided by the supplier, or analyzed offline using open source software (such as the open source software MitkDiffusion from the German Cancer Research Center). The main disadvantages are cumbersome and time-consuming processes, unsatisfactory fitting results, errors caused by multiple acquisitions and delineation of ROIs (Region of Interest), etc. Therefore, a multi-b-value diffusion-weighted MRI image processing method based on the combined IVIM and DKI model is proposed. Summary of the invention

[0003] In view of the shortcomings of the prior art, the purpose of the present invention is to provide an MRI image processing method of multi-b-value diffusion weighting based on the IVIM and DKI joint model, which solves the problems in the prior art.

[0004] The purpose of the present invention can be achieved through the following technical solutions:

[0005] A multi-b-value diffusion weighted MRI image processing method combining IVIM and DKI models comprises the following steps:

[0006] The body tissue images were acquired by multi-b-value diffusion-weighted sequence acquisition, and the following were obtained by fitting the IVIM and DKI joint model: true diffusion coefficient distribution map, pseudo diffusion coefficient distribution map, perfusion fraction distribution map and diffusion kurtosis map;

[0007] Draw the ROI area on the apparent diffusion coefficient image;

[0008] Based on the delineated ROI, the lesion and surrounding area were segmented on the true diffusion coefficient distribution map, pseudo diffusion coefficient distribution map, perfusion fraction distribution map and diffusion kurtosis map, and the average values ​​of intravoxel incoherent motion and diffusion kurtosis imaging parameters of the lesion and surrounding environment were calculated.

[0009] Furthermore, when acquiring images of human tissues and organs through a multi-b value diffusion weighted sequence, 3-13 different b values ​​need to be selected, including multiple b values ​​less than 200 S / mm 2 The b value of .

[0010] Furthermore, the steps of acquiring a human tissue image by using a multi-b-value diffusion weighted sequence are as follows:

[0011] 1) The diffusion weighted imaging sequence was a single-shot spin-echo planar sequence, and the volume acquisition parameters were set;

[0012] 2) The body was traced using a monopolar gradient mode with 3 or 4 scans and the number of directions in which the diffusion-sensitive gradient was applied was 3 or 4;

[0013] 3) When using monopole gradient, it needs to be used in conjunction with dynamic field calibration.

[0014] Furthermore, the fitting formula of the IVIM and DKI combined model is:

[0015]

[0016] Where S is the signal intensity, S0 is the signal intensity when b=0, f is the perfusion fraction, D* is the pseudo-diffusion coefficient, D is the true diffusion coefficient, and K is the diffusion kurtosis coefficient.

[0017] Furthermore, ROIs were drawn on the apparent diffusion coefficient images using 3D-Slicer or ITK-SNAP software.

[0018] Furthermore, the intra-voxel incoherent motion and diffusion kurtosis imaging parameters include: pseudo diffusion coefficient, true diffusion coefficient, perfusion fraction and kurtosis coefficient.

[0019] The IVIM and DKI combined model multi-b-value diffusion weighted MRI image processing system includes:

[0020] Image acquisition and processing module: human tissue images are acquired through multi-b-value diffusion weighted sequence acquisition, and the following are obtained through IVIM and DKI joint model fitting: true diffusion coefficient distribution map, perfusion fraction distribution map, pseudo diffusion coefficient distribution map and diffusion kurtosis map;

[0021] ROI delineation module: delineate the ROI area on the apparent diffusion coefficient image;

[0022] And, image segmentation module: based on the delineated ROI, the lesion and the surrounding area are segmented on the true diffusion coefficient distribution map, pseudo diffusion coefficient distribution map, perfusion fraction distribution map and diffusion kurtosis map, and the intravoxel incoherent motion and diffusion kurtosis imaging parameters of the lesion and the surrounding environment are calculated.

[0023] A computer storage medium stores a readable program, which can execute the above-mentioned IVIM and DKI joint model multi-b value diffusion weighted MRI image processing method when the program is running.

[0024] An electronic device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0025] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method.

[0026] A computer program product includes computer instructions, wherein the computer instructions instruct a computing device to execute operations corresponding to the above-mentioned IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method.

[0027] Beneficial effects of the present invention:

[0028] 1. The image processing method of the present invention automatically and efficiently completes the fitting and segmentation of the IVIM and DKI joint model, imposes constraints on the DKI parameters, simplifies the IVIM and DKI data processing flow, and avoids the errors caused by the multiple acquisition of IVIM and DKI images and the delineation of ROI.

[0029] 2. The present invention calculates the relevant parameters such as perfusion, diffusion and kurtosis coefficient through the combined model of IVIM and DKI, which can non-invasively predict MVI and WHO pathological grade before surgery, which is of great significance for guiding the clinic to make reasonable diagnosis and treatment strategies, and helps clinicians determine the optimal resection margin width for patients suitable for surgery, which may be an important supplement for evaluating the biological characteristics of tumors. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0031] Figure 1 is a flow chart of the MRI image processing method of the present invention;

[0032] Figure 2 It is the IVIM and DKI joint model fitting curve diagram of the present invention;

[0033] Figure 3 The lesion and the lesion surrounding area ROI are delineated by the open source software 3D Slicer in the present invention;

[0034] Figure 4 is the pseudo diffusion coefficient (D *), true diffusion coefficient (D), perfusion fraction (f), and kurtosis (K) plots;

[0035] Figure 5 The true diffusion coefficient (D) and pseudo diffusion coefficient (D) of the lesion and the surrounding 5 mm * ), perfusion fraction (f) and diffusion kurtosis (K) value statistics table. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] Example 1

[0038] like Figure 1 As shown, the MRI image processing method of multi-b-value diffusion weighting of the IVIM and DKI joint model includes the following steps:

[0039] S1, human tissue images were acquired by multi-b value diffusion weighted sequence acquisition, and various parameter maps were obtained by IVIM and DKI joint model fitting, including true diffusion coefficient distribution map (D map), perfusion fraction distribution map (f map), pseudo diffusion coefficient distribution map (D map), and so on. * Figure) and diffusion kurtosis plot (K-plot);

[0040] When acquiring a human tissue image through a multi-b value diffusion weighted sequence, 3-13 different b values ​​need to be selected. The steps for acquiring a human tissue image through a multi-b value diffusion weighted sequence are as follows:

[0041] 1) The diffusion-weighted imaging execution sequence is the SE-SS-EPI sequence. In the multi-b-value diffusion-weighted sequence, the volume acquisition parameters are set, such as the diffusion gradient combination mode, the number of diffusion-sensitive gradient application directions, and the diffusion weight. The diffusion weight can set the desired number of b-values ​​and the number of repeated scans.

[0042] 2) The body scan was performed in monopolar gradient mode, with 3 or 4 scans for tracking, and the number of directions in which the diffusion-sensitive gradient was applied was 3 or 4;

[0043] 3) When using monopole gradient, it needs to be used in conjunction with dynamic field calibration.

[0044] The fitting formula of the IVIM and DKI combined model is:

[0045]

[0046] Where S is the signal intensity, S0 is the signal intensity when b=0, f is the perfusion fraction, D* is the pseudo-diffusion coefficient, D is the true diffusion coefficient, and K is the diffusion kurtosis coefficient.

[0047] By performing a single fit on the multi-b value diffusion weighted graph, we can obtain the D graph, f graph, and D * Figure and K-chart.

[0048] S2, outline the ROI area on the apparent diffusion coefficient (ADC) image, i.e., the lesion and surrounding area;

[0049] In this embodiment, the ROI can be drawn on the ADC map by using the open source software 3D Slicer and saved in Nifti format.

[0050] S3, based on the ROIs drawn in S2, in D, f, and D * The lesion and surrounding area were segmented on the A-map and K-map, and the intravoxel incoherent motion and diffusion kurtosis imaging parameters of the lesion and surrounding environment were calculated;

[0051] Intravoxel incoherent motion and diffusion kurtosis imaging parameters include: pseudo-diffusion coefficient (D * ), true diffusion coefficient (D), perfusion fraction (f), and diffusion kurtosis coefficient (K) for mesoscopic structural analysis of the lesion and surrounding microenvironment; calculation of intravoxel incoherent motion and diffusion kurtosis imaging parameters (D * The specific steps of , D, f and K) are:

[0052] S31, data preprocessing, using open source software such as 3D Slicer, ITK-SNAP, and MRIcroGL to convert high-order diffusion-weighted MRI images with multiple b values ​​(maximum b value 2500) into Nifti format;

[0053] S32, the diffusion-weighted image was processed using the joint model, and the curves of the intravoxel incoherent motion (IVIM) and diffusion kurtosis imaging (DKI) models were fitted ( Figure 2 ) and the pseudo-diffusion coefficient (D * ), true diffusion coefficient (D), perfusion fraction (f) and kurtosis coefficient, where the b value needs to be selected from 3 to 13;

[0054] S33, delineate ROI, using the open source software 3D-Slicer or ITK-SNAP to delineate the lesion and surrounding area on the ADC image;

[0055] S34, using ROI segmentation lesion and surrounding area parameter map, the pseudo diffusion coefficient (D *), true diffusion coefficient (D), perfusion fraction (f) and diffusion kurtosis coefficient (K).

[0056] It should be noted that the IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method provided by the present invention processes magnetic resonance images and cannot directly obtain diagnostic results. It is also necessary to combine the analysis of T2WI, T1WI, CE-T1WI (T1WI enhancement) and other images to obtain diagnostic results.

[0057] Based on similar inventive concepts, an embodiment of the present invention further provides a computer storage medium storing a readable program, which can execute the above-mentioned IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method when the program is running.

[0058] Based on similar inventive concepts, an embodiment of the present invention provides an electronic device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0059] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method.

[0060] Based on similar inventive concepts, an embodiment of the present invention further provides a computer program product, including computer instructions, which instruct a computing device to execute operations corresponding to the above-mentioned IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method.

[0061] Example 2

[0062] In this embodiment, the MRI image processing method of the present invention is described by using a specific example;

[0063] like Figure 1 As shown, the MRI image processing method of multi-b-value diffusion weighting of the IVIM and DKI joint model includes the following steps:

[0064] Step 1: The subjects' kidneys were collected using a high-order diffusion-weighted sequence in monopolar gradient mode under free breathing, with b values ​​of 0, 10, 30, 50, 100, 150, 200, 400, 800, 1000, 1500, 2000, and 2500 s / mm 2 .

[0065] Step 2: Convert multi-b-value high-order diffusion-weighted MRI images into Nifti format using the open source software 3D Slicer.

[0066] Step 3: According to the IVIM and DKI joint model, the curves of the intravoxel incoherent motion (IVIM) and diffusion kurtosis imaging (DKI) models are fitted ( Figure 2 ) and the pseudo-diffusion coefficient (D * ), true diffusion coefficient (D), perfusion fraction (f) and kurtosis coefficient ( Figure 3 ).

[0067] Figure 2 In the figure, the fitting curve of the combined model of IVIM and DKI is steep in the low b value part (less than 200), and decreases very rapidly with the increase of b value, which is basically consistent with the IVIM low b value fitting curve, mainly reflecting the contribution of microvascular perfusion to diffusion; in the medium b value part (greater than 200, less than 1500), the curve decreases slowly, which is basically consistent with the trend of the IVIM medium b value part (greater than 200, less than 1500) fitting curve, mainly reflecting the contribution of diffusion to diffusion; in the high b value part (greater than 1500), the fitting curve of the combined model is gentle, which mainly reflects the heterogeneity of diffusion and the complexity of microstructure.

[0068] Figure 3 The parameter distribution diagram of IVIM and DKI joint model fitting is given, where 301 corresponds to the true diffusion coefficient (D) distribution diagram, and 302 corresponds to the perfusion coefficient (D * ), 303 corresponds to the distribution of the perfusion fraction (f), and 304 corresponds to the distribution of the diffusion kurtosis coefficient (K).

[0069] Step 4: Use the open source software ITK Snap to outline the lesion area and surrounding area ROI on the ADC map (such as Figure 4 The left kidney abscess, the area surrounding the abscess, and the rest of the left kidney were all delineated with ROI.

[0070] Step 5: Segment the lesion and surrounding area parameter map based on the ROI. According to the above steps, Figure 5 The parameter distribution map of the IVIM-DKI joint model of the left renal lesion (renal abscess) was segmented to obtain the intravoxel incoherent motion and diffusion kurtosis imaging parameters of the left renal lesion (renal abscess) and the microenvironment around the lesion: the pseudo diffusion coefficient (D * ), true diffusion coefficient (D), perfusion fraction (f) and diffusion kurtosis coefficient (K).

[0071] It should be noted that, in this embodiment, the specific process of the MRI image processing method is introduced by taking kidney abscess as an example. The introduced process and method are not limited to the kidney as a single organ and abscess as a single disease. More generally, the data acquisition and processing method introduced in the present invention is also applicable to other organs of the body and brain imaging, and is applicable to various tumors and peritumoral areas. Note: When applicable to brain imaging, the high-order diffusion weighted acquisition sequence parameters need to be modified.

[0072] Microvascular invasion (MVI) is an important factor for tumor recurrence and poor prognosis. Theoretically, MVI is at the edge of the tumor, and it is of great significance to explore the peritumoral situation. The present invention calculates the relevant parameters such as perfusion, diffusion and kurtosis coefficient through the combined model of IVIM and DKI, and non-invasively predicts MVI and WHO pathological grade before surgery, which is of great significance for guiding the clinic to make reasonable diagnosis and treatment strategies, and helps clinicians determine the optimal resection margin width for patients suitable for surgery, which may be an important supplement for evaluating the biological characteristics of tumors.

[0073] Example 3

[0074] In this embodiment, an MRI image processing system for multi-b-value diffusion weighted imaging of IVIM and DKI joint model is proposed, which specifically includes:

[0075] Image acquisition and fitting module: human tissue images are acquired through multi-b-value diffusion weighted sequence acquisition, and the following are obtained through IVIM and DKI joint model fitting: true diffusion coefficient distribution map, perfusion fraction distribution map, pseudo diffusion coefficient distribution map and diffusion kurtosis map;

[0076] ROI delineation module: delineate the ROI area on the apparent diffusion coefficient image;

[0077] And, the image segmentation module: based on the delineated ROI, the lesion and the surrounding area are segmented out on the true diffusion coefficient distribution map, the perfusion fraction distribution map, the pseudo diffusion coefficient distribution map and the diffusion kurtosis map, and the intravoxel incoherent motion and diffusion kurtosis imaging parameters of the lesion and the surrounding environment are calculated.

[0078] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0079] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A multi-b-value diffusion weighted MRI image processing method based on the IVIM and DKI joint model, characterized in that: The following steps are involved: Human body tissue images were acquired by multi-b-value diffusion-weighted sequence acquisition, and the following were obtained by fitting the IVIM and DKI joint model: true diffusion coefficient distribution map, pseudo diffusion coefficient distribution map, perfusion fraction distribution map and diffusion kurtosis map; Draw the ROI area on the apparent diffusion coefficient image; Based on the delineated ROI, the lesion and surrounding area were segmented on the true diffusion coefficient distribution map, pseudo diffusion coefficient distribution map, perfusion fraction distribution map and diffusion kurtosis map, and the average values ​​of intravoxel incoherent motion and diffusion kurtosis imaging parameters of the lesion and surrounding environment were calculated.

2. The MRI image processing method of the IVIM and DKI joint model multi-b-value diffusion weighting according to claim 1, characterized in that: When acquiring images of human tissues and organs using a multi-b value diffusion weighted sequence, 3-13 different b values ​​need to be selected, including multiple values ​​less than 200 s / mm 2 The b value of .

3. The MRI image processing method of the IVIM and DKI joint model with multi-b-value diffusion weighting according to claim 2, characterized in that: The steps of acquiring human tissue images through multi-b value diffusion weighted sequence acquisition are: 1) The diffusion weighted imaging sequence was a single-shot spin-echo planar sequence, and the volume acquisition parameters were set; 2) The body was traced using a monopolar gradient mode with 3 or 4 scans and the number of directions in which the diffusion-sensitive gradient was applied was 3 or 4; 3) When using monopole gradient, it needs to be used in conjunction with dynamic field calibration.

4. The MRI image processing method of the IVIM and DKI joint model with multi-b-value diffusion weighting according to claim 1, characterized in that: The fitting formula of the IVIM and DKI combined model is: Where S is the signal intensity, S0 is the signal intensity when b=0, f is the perfusion fraction, D* is the pseudo-diffusion coefficient, D is the true diffusion coefficient, and K is the diffusion kurtosis coefficient.

5. The MRI image processing method of IVIM and DKI joint model multi-b value diffusion weighting according to claim 1, characterized in that: ROIs were drawn on the apparent diffusion coefficient images using 3D-Slicer or ITK-SNAP software.

6. The MRI image processing method of the IVIM and DKI joint model with multi-b-value diffusion weighting according to claim 1, characterized in that: The intra-voxel incoherent motion and diffusion kurtosis imaging parameters include: pseudo diffusion coefficient, true diffusion coefficient, perfusion fraction and kurtosis coefficient. 7.IVIM and DKI joint model multi-b value diffusion weighted MRI image processing system, characterized in that, include: Image acquisition and processing module: human tissue images are acquired through multi-b-value diffusion weighted sequence acquisition, and the following are obtained through IVIM and DKI joint model fitting: true diffusion coefficient distribution map, perfusion fraction distribution map, pseudo diffusion coefficient distribution map and diffusion kurtosis map; ROI delineation module: delineate the ROI area on the apparent diffusion coefficient image; And, the image segmentation module: based on the delineated ROI, the lesion and the surrounding area are segmented out on the true diffusion coefficient distribution map, the pseudo diffusion coefficient distribution map, the perfusion fraction distribution map and the diffusion kurtosis map, and the intravoxel incoherent motion and diffusion kurtosis imaging parameters of the lesion and the surrounding environment are calculated.

8. A computer storage medium storing a readable program, characterized in that: When the program is running, it can execute the IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method described in any one of claims 1-6.

9. An electronic device, characterized in that: include: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method as described in any one of claims 1-6.

10. A computer program product comprising computer instructions, characterized in that: The computer instructions instruct the computing device to perform operations corresponding to the IVIM and DKI joint model multi-b-value diffusion weighted MRI image processing method as described in any one of claims 1-6.

Images