JP2023183408A - Quantitative evaluation method of human brain axon density based on magnetic resonance diffusion tensor imaging - Google Patents

Abstract

To disclose a quantitative evaluation method of a human brain axon density based on magnetic resonance diffusion tensor imaging.SOLUTION: The method includes the steps of: calculating diffusion tensor imaging fittings of a plurality of single diffusion sensitive factors by performing data processing on magnetic resonance diffusion-weighted image data obtained by scanning a human brain through magnetic resonance diffusion tensor imaging, and obtaining a plurality of fractional anisotropic exponential diagrams; and performing quantization processing of a human brain axon density on the basis of the plurality of fractional anisotropic exponential diagrams, and calculating a difference of a fractional anisotropic exponential percentage between different diffusion sensitive factors to obtain a quantitative diagram of the human brain axon density.EFFECT: The invention has high credibility, and an easy arithmetic process to be easily spread, realizes a clinically executable parameter fixed quantity relative to an axon density in a brain white matter, and can be used for detection of axon loss related diseases.SELECTED DRAWING: None

Description

The present invention relates to a magnetic resonance imaging data processing method in the field of magnetic resonance imaging technology, and particularly to a method for quantitatively evaluating human brain axon density based on magnetic resonance diffusion tensor imaging.

Diffusion-weighted imaging (DWI) uses specific magnetic resonance imaging sequences to generate images from the resulting data and uses water molecule diffusion to create contrast within the magnetic resonance image. , and the intensity of each voxel reflects the best estimate of water diffusivity at that location. Diffusion of water molecules within tissues is not free, but rather reflects interactions with many obstacles such as macromolecules, fibers and membranes. Therefore, water molecule diffusion patterns can reveal fine details of tissue structure, whether in normal or pathological conditions. It allows non-invasive in-vivo mapping of diffusion processes of molecules (mainly water) within biological tissues.

Diffusion tensor imaging (DTI) is a special magnetic resonance diffusion-weighted imaging technique that reflects the microscopic movement of water molecules and is used to study the pathophysiological state of diseases from the cellular and molecular level. , we track the running of nerve pathways using the direction information of anisotropic diffusion in the diffusion tensor field, and thereby obtain the running direction and three-dimensional morphology of nerve fibers and functional bundles in the brain white matter. The index fractional anisotropy (FA) is the ratio of the anisotropic part of the diffusion to the total value of the diffusion tensor and is used to evaluate the anisotropy of the dispersion. This is important when exploring brain white matter nerve axons.

Neurites are long, thin parts that extend from the cell bodies of nerve cells (neurons), and are divided into dendrites and axons, and are used for information transmission between cells. Quantifying axon morphology based on axon density and directional distribution provides an opportunity to understand the structural basis of brain function in normal and brain dysfunction populations.

Currently, there is no standard for a non-invasive quantitative measurement method of human brain axon density. Neurite orientation dispersion and density imaging (NODDI) is a development method based on emerging magnetic resonance diffusion-weighted imaging techniques to assess the complexity of neuronal axon and dendritic microstructure. Although this method can reflect the morphological information of nerve fibers, this method requires many diffusion directions and requires a long imaging time. The model calculation is complicated, which is disadvantageous for its widespread use in clinical practice.
Therefore, there is currently a need for a method for quantitatively evaluating human brain axon density that is non-invasive, reliable, and easily applicable to clinical practice.

The purpose of the present invention is to address the deficiencies of the prior art and provide a method for quantitatively evaluating human brain axon density based on magnetic resonance diffusion tensor imaging. The present invention can perform three-dimensional high-resolution imaging of axon density within a clinically acceptable time (12 minutes).

Axons are microstructures within human brain tissue with diameters ranging from 0.2 to 20 microns, and magnetic resonance sequences have difficulty directly capturing axon density due to resolution limitations. . The method of the present invention can solve the problem that it is difficult to perform quantitative evaluation by directly performing axonal imaging.

In order to achieve the above objectives, the present invention adopts the following technical solutions:
S1: Data processing is performed on magnetic resonance diffusion weighted image data obtained by scanning the human brain by magnetic resonance diffusion weighted imaging to calculate diffusion tensor imaging fitting of multiple single diffusion susceptibility factors b. Obtain the Fractional Anisotropy (FA).
S2: Quantify human brain axon density based on multiple partial anisotropy index diagrams, calculate the difference in partial anisotropy index percentage between different diffusion sensitivity factors b, and calculate human brain axon density. Obtain a quantitative figure.

The above magnetic resonance diffusion tensor imaging scan specifically performs three-dimensional scanning imaging of the whole brain.
Magnetic resonance diffusion weighted image data originates from three parts: intra-axonal space water, extra-axonal space water, and free water, which have different diffusion coefficients and fractional anisotropy indices. Relative to pure white matter signals, with increasing diffusion sensitivity factor b, the magnetic resonance signal of intra-axonal space water with higher diffusion coefficient decays faster, leading to an increase in the weight of extra-axonal space in diffusion-weighted imaging results. cause.

By setting the difference value of the anisotropy index percentage to be calculated, the present invention can quantitatively reflect changes in axon density and directly capture axon density. Accurately obtain the results of cord density and solve the problem that it is difficult to quantitatively evaluate by directly performing axonal imaging.

The method of the present invention adopts a conventional clinical magnetic resonance diffusion weighted sequence to realize quantitative evaluation of three-dimensional axon density, and image acquisition and reconstruction are based on magnetic resonance diffusion tensor imaging in a clinically feasible time. Achieves rapid and accurate three-dimensional quantitative evaluation of axon density in brain white matter.
The present invention is highly reliable, has a simple calculation process, is easily disseminated, realizes clinically practicable parameter quantification of axon density in brain white matter, and can be used to detect axon loss-related diseases.

Figure 1 shows the quantitative evaluation index of axon density according to the present invention, in which the percentage difference value pdFA of some anisotropy indices is measured in six regions of interest (ROI) of brain white matter of healthy subjects. This is an application example. The vertical axis is the human brain axon density quantitative map PDFA obtained using Scheme 3, and the horizontal axis is the axon density index (Neurite Density Index) obtained using neurite direction dispersion and density imaging (NODDI). Index, NDI) NDI-NODDI. The thin lines and dots in the figure are the data results with one averaging, and the black lines and dots are the data results with two averaging, with low and high signal-to-noise ratios (signal-to-noise ratios), respectively. Noise Ratio, SNR) condition. The six ROIs are the genu of the corpus callosum (GCC), the body of the corpus callosum (BCC), the splenium of the corpus callosum (SCC), and the anterior limb of the internal capsule. limb of the internal capsule (ALIC), posterior limb of the internal capsule (PLIC), and superior longitudinal fasciculus (SLF). FIG. 2 is a quantitative evaluation index of axon density according to the present invention, and the quantitative diagram pdFA of human brain axon density is an example of application in ex vivo human corpus callosum tissue. The vertical axis in Figure 2(a) is the pdFA of different sections of the corpus callosum of ex vivo human brain tissue obtained using the present invention, and the horizontal axis is the paraffin section nerve fibers using the Bielschowsky silver staining technique (Bielschowsky). This is an axon density index obtained by analysis using FIG. 2(b) is the corresponding position in the magnetic resonance image of the section. Figure 2(c) shows the microscopic imaging results of the section staining results.The colored parts are axons, and the ratio of the area of the colored parts to the total area is defined as NDI-Histology, which is the density of axons in ex vivo human brain. be used as the standard for quantitative evaluation. FIG. 3 is a Pearson correlation analysis of the human brain axon density quantitative diagram pdFA and the intra-axonal volume fraction INVF under different signal-to-noise ratio conditions in a numerical simulation experiment. The vertical axis is the human brain axon density quantitative map PDFA calculated using the magnetic resonance signal obtained based on the simulation signal formula, and the horizontal axis is the volume fraction within the axon set in the simulation experiment. . Table 1 is a biological brain data collection sequence parameter table. Table 2 is a parameter table of a combination solution for different diffusion sensitivity factors b. Table 3 is a statistical parameter table of the implementation results of the combination solution method for different diffusion sensitivity factors b. Table 4 is an ex-vivo human brain data acquisition sequence parameter table. Table 5 is a simulation experiment parameter setting table.

The implementation of the specific solution of the present invention will be further explained with reference to the drawings.
Examples of the present invention and their specific implementation situations are as follows:
In the first embodiment of the present invention, magnetic resonance diffusion weighted imaging was performed on 61 healthy subjects using a 3.0T magnetic resonance apparatus to obtain magnetic resonance diffusion weighted image data. The scan time of the sequence was 19 minutes, and the other parameters of the sequence are as shown in Table 1.
[Table 1]

表２に示すように、３種類の組み合わせ方式が存在する（スキーム１－３）。
[表２]

Using the "Diffusion Tensor Fitting" tool in FSL, we selected the weighted least squares method and determined the number of averages of 1 (relatively low signal-to-noise ratio) and 2 times (relatively high signal-to-noise ratio), respectively. Diffusion tensor imaging fitting is performed on the magnetic resonance diffusion weighted image data of one diffusion sensitivity factor b to obtain partial anisotropy index diagrams corresponding to three diffusion sensitivity factors b.
As a quantitative evaluation index of human brain axon density based on diffusion tensor imaging, the corresponding quantitative map of human brain axon density is calculated using the partial anisotropy index map obtained according to the following formula:
[Number 1]

As shown in Table 2, there are three types of combination schemes (Scheme 1-3).
[Table 2]

Model fitting between neurite orientation dispersion and density imaging (NODDI) was performed using the complete dataset collected with the sequences in Table 1, and the resulting axonal density index (NDI-NODDI) is calculated and used as the standard for evaluating living human brain axon density. Within the selected white matter region of interest, Pearson correlation analysis was performed on the human brain axon density quantitative map PDFA and the axon density index NDI-NODDI obtained above, and the Pearson correlation coefficient (CC) was calculated. The results are shown in Figure 1. The statistical parameter values corrected by multiple testing were calculated as corrected P values, and the results are shown in Table 3, where gray cells represent data sets with two averages. Figure 1 and Table 3 both show the human brain axon density quantitative map pdFA and the axon density index NDI-NODDI obtained using a data set with an average number of times of 2 (relatively high signal-to-noise ratio). This shows that the correlation is higher and can more accurately reflect human brain axon density.

As shown in Table 3, in the present invention, the combination of diffusion sensitivity factor b in Scheme 3 is the optimal option, and its quantitative diagram pdFA of human brain axon density has a correlation with the axon density index NDI-NODDI. highest. The time required for this scheme was 12 minutes.
[Table 3]

In a second embodiment of the present invention, magnetic resonance diffusion weighted imaging scans were performed on one ex vivo hemibrain fixed in formalin solution for 4 weeks using a 3.0T magnetic resonance apparatus, and the magnetic resonance diffusion weighted image data were collected. Obtained. Other parameters of the sequence are shown in Table 4.
[Table 4]

After completion of magnetic resonance scanning, ex vivo hemibrain preparations were cut into 5 mm thick coronal blocks. This process produces six slices corresponding to their positions in the sagittal magnetic resonance image, as shown in Figure 2(b). The corpus callosum tissue block was taken out and six corresponding sections were obtained after paraffin embedding, and the thickness of the sections was 6 μm. Neurite histochemical staining of tissue sections was performed by the paraffin section nerve fiber Bielschowski silver staining technique, as shown in Figure 2(c). The axonal density index (NDI-Histology) of the corpus callosum is quantitatively measured on the stained sections using Image-J software.

Pearson correlation analysis was performed on the axonal density quantitative evaluation index NDI-Histology obtained by human brain axon density quantitative map PDFA and histochemical staining of ex vivo human brain sections, and the results are shown in Figure 2 (a). , both have a high linear correlation. The present invention is also excellent for quantitative evaluation of axonal density in ex vivo human brain specimens.

In order to examine the principle by which a quantitative evaluation index for human brain axon density (human brain axon density quantitative map PDFA) quantitatively evaluates axon density, the present invention conducts a simulation experiment based on a white matter "standard model."

[表５]

[Table 5]

As shown in Figure 3, the simulation results show that within the signal-to-noise ratio range (20-40) where the biological experimental data lies, the volume fraction within the axon, INVF, represents a quantitative evaluation index of axon density. With the increase in , the value of the human brain axon density quantitative map pdFA, which is a quantitative evaluation index of axon density, also increases linearly, and the two have a high linear correlation. Simulation experiments show that the quantitative evaluation index of axon density, pdFA, of the present invention can sensitively detect changes in human brain axon density with data quality that can be achieved under clinical conditions.

Compared to the axonal density index (NDI-NODDI) obtained based on the conventional neurite direction dispersion and density imaging (NODDI) model fitting, the partial anisotropy index percentage based on the diffusion tensor imaging of the present invention The difference value pdFA requires less time for the calculation process. Taking data processing on a Linux workstation (2 x 2.80 GHz Intel (registered trademark) Xeon (registered trademark) 20 core processor, 252 GB memory) as an example, if the same data is input, Approximately 10 hours of computational time was required for quantitative evaluation of whole-brain axon density in the human brain by model fitting with density imaging. Quantitative assessment of axonal density using the magnetic resonance diffusion tensor imaging-based fractional anisotropy index percentage difference pdFA of the present invention requires only about 3 minutes of computational time.

Note that the above describes only the embodiments of the present invention and the technical principles for its operation. As will be understood by those skilled in the art, the present invention is not limited to the specific embodiments described above, and various obvious modifications and adjustments to the parameters may be made to those skilled in the art without departing from the scope of the invention. do not. Accordingly, although the embodiments described above may include many other equivalent embodiments without departing from the inventive concept, the scope of the invention is determined by the appended claims.

Claims (5)

S1: Data processing is performed on the magnetic resonance diffusion weighted image data obtained by scanning the human brain by magnetic resonance diffusion weighted imaging to calculate the diffusion tensor imaging fitting of multiple single diffusion susceptibility factors b. obtaining two partial anisotropy index diagrams;
S2: Quantify the human brain axon density based on the two partial anisotropy index diagrams, calculate the difference in the partial anisotropy index percentage between two different diffusion sensitivity factors b, and calculate the human brain axis density. obtaining a quantitative diagram of cord density;
A method for quantitatively evaluating human brain axon density based on magnetic resonance diffusion tensor imaging.