JP6966786B2 - MRI examination method for degenerative disc disease - Google Patents

Description

A method for examining disc degeneration using magnetic resonance imaging (hereinafter referred to as MRI), particularly q-space imaging (hereinafter referred to as QSI) or diffusion tensor imaging. , Hereinafter referred to as DTI).

MRI is widely used in clinical practice along with computed tomography (CT) for disease testing. MRI is an inspection method in which a subject obtains information generated from hydrogen nuclei resonating with radio waves of a specific frequency in a strong magnetic field and uses it for inspection, and is used for inspection of diseases such as brain, spinal cord, and limbs.

In the field of orthopedics, MRI is used to test various diseases such as cervical spondylosis, hernia, and tumor. There are many analysis methods for MRI, and various classification methods have been proposed as diagnostic criteria for the severity of intervertebral disc degeneration and herniated disk even in the case of spinal examination (Non-Patent Document 1). There are many classification methods, although there are advances in analysis methods, but all classification methods are qualitative classification methods, and the degree of diagnostic agreement between observers is low, and more objective classification methods are available. The background is what is required.

There are various analysis methods for MRI, but roughly speaking, there are methods for analyzing the obtained signal as free diffusion of water molecules, and limited diffusion that occurs in a restricted space such as cells and organs. There is a diffusion MRI method to analyze as. In clinical practice, diffusion-weighted images have made rapid progress in the latter half of the 90's because hyperacute cerebral infarction can be visualized more sensitively than conventional MRI imaging methods and CT. In the brain, methods for diagnosing lesions such as cerebral infarction and brain tumors by visualizing them by several analytical methods are being established.

Even in the spinal cord, the MRI examination method is widely used as a criterion for determining the pathological condition. For spinal disc degeneration, it is common practice to analyze the disc using T2-weighted images and classify the degree of disc degeneration by the Pfirrmann classification method, which classifies the disc into five stages according to the signal intensity of the mid-sagittal section. (Non-Patent Document 2).

In the Ferman classification, the degree of degeneration is determined by classifying the intervertebral disc degeneration into 5 stages of 1 to 5 from the T2-weighted image of the intervertebral disc. FIG. 5 is an image showing the pathophysiology of typical degeneration of intervertebral discs classified by the Ferman classification. The Ferman classification is a method of qualitatively classifying the pathological condition as follows from the signal intensity of the obtained MRI image and the shape of the intervertebral disc.

Grade 1: uniform disc shape, normal disc height, uniform bright white image, Grade 2: non-uniform disc shape, normal disc height, horizontal gray lines Non-uniform bright white image, Grade 3: Disc shape is non-uniform and between the annulus fibrosus and nucleus pulposus is unclear, but disc height is normal, image intensity is moderate gray image, Grade 4: Disc shape is non-uniform, the boundary between the annulus fibrosus and nucleus pulposus is lost, and disc height is normal or slightly reduced. The signal is low intensity, dark gray image, grade 5: disc shape is non-uniform, the boundary between the annulus fibrosus and the nucleus pulposus is lost, and the disc space is lost. The signal is low intensity, black image.

As described above, since the Ferman classification is a qualitative evaluation method for judging from the signal intensity of the MRI image and the shape of the intervertebral disc, different judgments may occur depending on the observer. In particular, in the case of moderate degeneration, that is, grades 2 to 4, it is difficult to distinguish them, and the judgment by the observer varies. Furthermore, since it is a qualitative evaluation method, it is not possible to capture minute changes. Therefore, there has been a demand for a quantitative MRI analysis method that can capture the pathological condition in more detail.

Various analysis methods are known for diffusion MRI, but QSI has been attracting attention as a method that may be able to perform more detailed analysis. QSI is an MRI analysis method initially reported as an analysis method for porous substances (Non-Patent Document 3). In recent years, QSI analysis has begun to be applied to biometric information because it is possible to capture the diffusion phenomenon caused by water molecules more finely than the conventional method and to acquire micro-sized structural information (non-biological information). Patent Documents 4, 5).

Although an analysis method using QSI has been established in the brain, it has been reported that white matter degeneration and changes in the intervertebral disc can be detected in the spinal cord during the day (Non-Patent Documents 6 and 7). It has also been reported that QSI is applied to intervertebral disc degeneration and has a correlation with the classification by the Ferman classification method, which is a qualitative evaluation method (Non-Patent Document 8).

In addition, focusing on the directionality and speed of the diffusion phenomenon, DTI, which reflects the anisotropy of diffusion, enables qualitative evaluation of a highly unidirectional structure, so it is a unidirectional white matter fiber. Clinical application is progressing in the brain and spinal cord with high structure (Non-Patent Document 9). In tissues where fibers are oriented, such as the white matter and spinal cord, the anisotropy of diffusion is strong, and it is thought that this reflects minute changes that could not be obtained from conventional images by DTI analysis. .. A method for evaluating intervertebral disc degeneration by DTI analysis has also been reported in the spinal cord (Non-Patent Documents 10 and 11).

Yiping, L. et al., Clin. Orthop. Relat. Res. 2015, Vol.473, pp.1896-1902. Pfirrmann, C.W., et al., Spine, 2001, Vol.26 (17), pp.1873-1878. Callaghan, P.T., et al., Nature, 1991, Vol.351, pp.467-469. Yamada, I., et al., Magn.Reson. Med., 2015, Vol.73, pp.2262-2273. Fujiyoshi, K., et al., J. Neurosci., 2016, Vol.36 (9), pp.2796-2808. Assaf, Y., et al., Magn.Reson. Med., 2000, Vol. 44, pp.713-722. Katsura, M., et al., J. Magn.Reson. Imaging, 2014, Vol. 40, pp.1208-1214. Katsura, M., et al., Proc.Intl. Soc. Mag. Reson. Med. 24, 2016, http://www.ismrm.org/16/program_files/EP20.htm. Fujiyoshi, K. et al., Exp. Neurol., 2013, Vol.242, pp.74-82. Antoniou, J. et al., J. Magn. Reson.Imaging, 2013, Vol.38, pp.1402-1414. Uda, T. et al. Spine (Phila Pa 1976), 2013, Vol.38, pp.407-414. Nakashima D., et al., ORS 2017 Annual Meeting Paper No.0047.

However, the currently reported analysis method using QSI has a problem that it is not possible to capture minute changes at the stage of attempting quantification. Therefore, it is only classified in the same way as the Ferman classification, and it has not been possible to capture changes in the spine in more detail.

The present inventors have succeeded in capturing minute changes in intervertebral disc degeneration by performing QSI analysis using a rat intervertebral disc degeneration model (Non-Patent Document 12). However, even if the same analysis method was applied to human disc degeneration, it was not possible to capture quantitative changes as in the case of rats. Further, the analysis method using DTI has not been applied to the extent that quantitative analysis of intervertebral disc degeneration and minute changes can be captured.

An object of the present invention is to provide a method for quantitatively evaluating the degree of human intervertebral disc degeneration, grasping it in more detail, and analyzing minute changes. The subject is to objectively show the pathophysiology of intervertebral disc degeneration by grasping changes in more detail than the Ferman classification currently used as an index of intervertebral disc degeneration, and to use it for the treatment of diseases.

The present invention relates to methods and programs for examining human intervertebral discs using QSI or DTI.
(1) A method for inspecting a human intervertebral disc using diffusion MRI, which uses Q-space imaging (QSI) analysis or diffusion tensor image (DTI) analysis, and Kurtosis (KT) Fractional Anisotropy (FA) by QSI analysis. KT FA), probability at zero diffusion FA (ZP FA), full width at half maximum FA (FWHM FA), FA by DTI analysis, λ ₁ mapping, _{at least one of (λ 2} + λ ₃ ) / 2 mapping. An inspection method characterized by analysis by.
(2) The inspection method according to (1) is characterized in that the intervertebral disc degeneration is quantified.
(3) The b value, axis information, and diffusion time are extracted from the spectral data of each voxel of the target intervertebral disc obtained by the MRI device, and KT FA, ZP FA, FWHM FA by QSI analysis, FA by DTI analysis, λ. _{A program that causes a computer to perform a step of performing an operation by at least one of 1} mapping and (λ ₂ + λ ₃ ) / 2 mapping, and a step of quantitatively expressing the state of the target intervertebral disc based on the calculation result.
(4) The program according to (3), which causes a computer to execute a step of displaying the calculation result as an image.

By obtaining a quantitative method for examining intervertebral disc degeneration due to QSI or DTI, it is possible to more objectively distinguish the pathological condition of a patient. It is the first time that the present inventors have found that minute changes can be quantitatively captured by the above parameters in the analysis by DTI and QSI. According to the method of the present invention, changes in the pathological condition can be grasped in detail and objectively, so that small changes in the pathological condition such as improvement of the disease by treatment can be quickly grasped, which can be useful for treatment.

The flowchart explaining the operation of the program which concerns on embodiment of this invention. The figure which shows the comparison between the Kurtosis Fractional Anisotropy (KT FA) and the conventional method. The figure which shows the comparison between the analysis method of this invention, and the T2-weighted image which is a conventional method. The figure which shows the comparison between the analysis method by QSI of this invention, and the conventional method. The figure which shows the typical image of the intervertebral disc degeneration by the Ferman classification which is a conventional method.

The present inventors performed QSI analysis using a rat intervertebral disc degeneration model, and in rats, degenerated discs using probability at zero disorder (ZP), Kurtosis (K), and full width at staff maximum (FWHM). It has been reported that normal intervertebral discs can be significantly distinguished (Non-Patent Document 12). However, in human degenerative disc disease, the degree of degeneration could not be significantly distinguished even by analysis using these parameters. Therefore, we investigated a method that can quantitatively measure the degree of intervertebral disc degeneration in humans.

In addition, when the same method as the report (Non-Patent Documents 6 to 8) in which human intervertebral disc degeneration is analyzed by QSI is used, it is possible to classify to the same extent as the Ferman classification, but it is possible to distinguish more clearly. There wasn't. Therefore, when the analysis was performed by changing the parameters, in the analysis by QSI, Kurtosis Fractional Anisotropy (KT FA), probability at zero dispersion FA (ZP FA), full width at half maximum FA (Full width at half maximum FA) It became clear that 1 to 5 can be clearly and quantitatively distinguished. Furthermore, in the DTI analysis, it was found that FA, λ ₁ mapping, and (λ ₂ + λ ₃ ) / 2 mapping can clearly and quantitatively distinguish grades 1 to 5 of the Ferman classification.

Based on the data acquired by MRI, all the steps of analysis by the QSI method and the DTI method are executed by a computer program. Specifically, b-value data, axis information, and nucleic acid time are extracted from the data in each voxel of the target intervertebral disc obtained by the MRI apparatus. In the case of QSI analysis, KT FA, ZP FA, FWHM FA, and in the case of DTI analysis, FA, λ ₁ mapping, and (λ ₂ + λ ₃ ) / 2 mapping are calculated by a computer. Furthermore, the state of the target intervertebral disc can be quantitatively shown as a numerical value based on the calculation result, or the position of the intervertebral disc can be visualized by a method such as color mapping (FIG. 1).

[Example 1]
Hereinafter, the data will be described with reference to the data. In 21 patients who gave informed consent in the Department of Orthopedic Surgery, Keio University School of Medicine (84 discs for which high-resolution images were obtained among the 5 discs between lumbar vertebrae 1 / 2-5 / sacrum) MRI images were obtained from the extraction (average age 53 years, 15 males, 6 females).

For the analysis, MAGNETOM Skyra 3T (manufactured by Siemens Healthcare) was used, and TR / TE (ms): 4000/89, FOV (mm) was used using Echo Planar Imaging (EPI) under the condition that the δ of the above device was the shortest. ): 576 × 576, Matrix: 192 × 192, Tickness (mm): 3, gap (mm): 0, slice (sheets): 7, voxel size (mm): 3 × 3 × 6, Δ / δ (ms) ): 43.1 / 31.7, b value (s / mm ² ): 0, 50, 200, 450, 800, 1250, 1800, 2400, 3150, 4000 were set as imaging conditions. The setting for applying the dipole magnetic field gradient (MPG) was 6 axes.

The outline is described below, but the calculation method of each value is based on the general method used in this field. In the QSI method, a signal attenuation curve is derived from the signal values for several steps in each vector direction and subjected to inverse Fourier transform to create a probability variation probability distribution. Then, Kurtosis (K), ZP, and FWHM are calculated, and tensors are calculated to calculate λ ₁ , λ ₂ , and λ ₃ , which are eigenvalue values of each tensor of K, ZP, and FWHM. _{From each of λ 1} , λ ₂ , and λ _{3 of} K, ZP, and FWHM, each <D> is obtained by the following formula (I).

異方性比率（Ｆｒａｃｔｉｏｎａｌ Ａｎｉｓｏｔｒｏｐｙ、ＦＡ）は下記式（ＩＩ）によりＫｕｒｔｏｓｉｓ、ＺＰ、ＦＷＨＭより算出した各λ_１、λ_２、λ_３より求める。

_{The anisotropy (FA) is obtained from λ 1} , λ ₂ , and λ ₃ calculated from Kurtosis, ZP, and FWHM by the following formula (II).

That is, the KT FA can be obtained as the FA of K, the ZP FA as the FA of ZP, and the FWHM FA as the FA of FWHM. Further, in the DTI analysis, the apparent diffusion constant (ADC) was calculated with a b value of 800, and FA, λ ₁ , (λ ₂ + λ ₃ ) / 2 were calculated and displayed.

For intervertebral discs classified into Ferman classification grades 3 and 4 from T2-weighted images (an example of an MRI image is shown below in FIG. 2), the obtained signals are signal strength by T2 mapping, ADC, and QSI KT FA. Was calculated and made into a graph. The Ferman classification is diagnosed by a skilled spine surgeon with 10 years of experience from T2-weighted images. The horizontal axes are T2 mapping [ms], ADC [* 10 ^-3 mm ^{2 /} s], and KT FA [a. u. ], And the vertical axis is the frequency of each.

As shown in FIG. 2, in T2 mapping and ADC, the signals obtained from the grade 3 and grade 4 intervertebral discs of the Ferman classification overlap each other, and it is difficult to judge the degree of intervertebral disc degeneration from the signal intensity. In particular, in the signal obtained by ADC, the values of grades 3 and 4 overlap and cannot be separated. Further, in T2 mapping, the signal tends to be low in grade 3 and high in grade 4, but the median frequencies of both tend to be close to each other, and the degree of disc degeneration should be judged from the obtained signal values. It is difficult. On the other hand, in QSI KT FA, the signal values of grades 3 and 4 are separated, and the degree of disc degeneration can be determined from the signal values.

As mentioned above, since the conventional Ferman classification is a qualitative evaluation, it is difficult to distinguish grades 2 to 4 which are moderately denatured. In addition, the degree of matching of the diagnosis is low because of the subjectivity. On the other hand, as shown in FIG. 2, the analysis by KT FA is a quantitative analysis, and moderate degeneration such as grades 3 and 4 can be clearly distinguished. Although the data are shown here for grades 3 and 4, differences such as grades 2 and 3 or grades 2 and 4 can also be clearly distinguished.

Next, the intervertebral discs classified from Grade 1 to Grade 5 of the Ferman classification for KT FA, ZP FA, FWHM FA by QSI analysis, FA of ADC by DTI analysis, λ ₁ mapping, and (λ ₂ + λ _{3) / 2 mapping.} The signal strength was calculated by each method and summarized in a graph (Fig. 3). The horizontal axis shows grades 1 to 5 of the Ferman classification. The Ferman classification is based on the diagnosis of a skilled spine surgeon with 10 years of experience from T2-weighted images in the same manner as above.

Grade of Farman classification by using KT FA, FWHM FA, ZP FA by QSI analysis, FA (ADC) by DTI analysis, λ ₁ mapping, (λ ₂ + λ _{3) / 2 mapping compared to T2 mapping.} 1 to 5 can be classified very well (Fig. 3). In particular, in T2 mapping, there is almost no quantitative difference between grades 3 to 5 of the Ferman classification, but a quantitatively significant difference can be observed in the six analytical methods of the present invention. Thus, the present invention is the first to be able to perform a quantitative analysis of the degree of degeneration in the intervertebral disc to the extent that the grades of the Ferman classification can be completely separated.

Since the Ferman classification is a qualitative method for making a complex judgment based on the signal intensity of the T2-weighted image and the shape of the intervertebral disc, the degree of agreement by the observer has been a problem. In particular, moderate disc degeneration of grades 2-4 of the Ferman classification may be judged differently by doctors. On the other hand, since the method of the present invention is a quantitative and unprecedentedly sensitive method, there is no discrepancy in judgment by the observer. That is, it is possible to objectively judge the disease without the subjectivity of the observer. In addition, it becomes possible to grasp subtle differences such as improvement of diseases by medication as numerical values. As a result, it becomes possible to obtain an objective index of follow-up observation, which has been difficult to obtain so far, and it becomes possible to use it for treatment.

[Example 2]
Next, the results of a cross-sectional study with an increased number of samples are shown. With the permission of the Institutional Review Board at Nagura Orthopedic Surgery, 44 Japanese volunteers (average age 63.5 years (standard deviation 9.2), 28 men, women) obtained informed consent. In 23 people), MRI images (n = 220) of the intervertebral discs of the lumbar vertebrae (5 intervertebral discs between lumbar vertebrae 1/2 to 5 / sacrum) were acquired and analyzed.

For the analysis, a MAGNETOM Skyra fit 3T (manufactured by Siemens Healthineers) equipped with a 30-channel body coil and a 30-channel spine coil was used. Using the T2-weighted image in the axial direction as a reference, analysis was performed by the T2-weighted image of the median sagittal plane, T2 mapping, and QSI. The imaging protocol is shown in Table 1.

The abbreviations in the table are as follows.
T2WI: T2-weightted MRI
T2map: T2 mapping MRI
RARE: rapid acquisition with relaxation enhancement
Δ: the time beween the leading edge of the diffusion gradient
δ: gradient lens, aprx: aproximately

Of the 220 spinal cords, 25 were excluded because of low signal intensity and 31 was a disease other than disc degeneration, and MRI images obtained from 164 spinal cords were used for the following analysis. The extracted intervertebral discs were classified according to the Ferman classification. As a result of classification, grade 1 was classified into 1, grade 2 was classified into 19, grade 3 was classified into 59, grade 4 was classified into 54, and grade 5 was classified into 31 discs. Since there was only one disc classified as Grade 1, it was excluded from the following analysis.

Analysis was performed by T2 mapping, ADC map, and QSI using a computer. Theoretically, the optimum b value should be within the reciprocal of the average value of the ADC that gives the optimum signal intensity to noise ratio in the tissue to be analyzed, so a b value of 450 s / mm ² is used. used. In addition, an ADC map based on the conventional single exponential model was calculated from a part of the QSI data (b value: 0, 450 [s / mm ² ]). Table 2 and FIG. 4 show the results of each analysis method.

As shown in FIG. 4, a significant difference was observed between Farman grades II and III in all analysis methods except ADC (p <0.05). In addition, all QSI analysis (mean Kurtosis, mean Probability at zero display, mean FWHM, FA of Kurtosis (KT FA), FA of Probability at zero FLA A significant difference was observed between Farman Grades II and III as compared to T2 mapping and ADC, which are MRI analysis methods.

Significant differences were found in all analytical methods in Farman Grades III and IV (p <0.05). In particular, KT FA, ZP FA, and FA FWHM can be more clearly distinguished from Farman Grade III and IV than other analytical methods (p <1 * ^10-10 ). In particular, FA FWHM was able to most clearly distinguish between Farman Grade III and IV of all analytical methods.

For Farman Grade V, all analytical methods showed the highest standard deviation compared to other Farman grades. Further, in KT FA, ZP FA, and FA FWHM, the value increased from Farman Grade II to IV, while the value decreased from IV to V. However, in the analysis method using QSI, it is possible to quantitatively classify IV from Farman Grade II, which is liable to vary depending on the observer and has been extremely difficult to classify.

KT FA, FWHM FA, ZP FA by QSI analysis, FA (ADC) by DTI analysis, λ ₁ mapping, (λ ₂ + λ ₃ ) / 2 mapping are sensitive to moderate disc degeneration of Ferman classification 2-4. Since it is possible to distinguish them, any of them may be used for the inspection. Analysis by the DTI method has the advantage that it can be performed even if it is not the latest MRI equipment because it can be analyzed in a short time, and diagnosis can be performed even in a clinic. In addition, since the analysis by the QSI method is considered to be more quantitative, detailed results may be obtained.

In the method of the present invention, it is possible to obtain a quantitative value that could not be obtained from the T2-weighted image that has been mainly used in the spinal cord, so that an accurate diagnosis can be made. In addition, since it is possible to capture minute changes in the patient over time, it is possible not only to capture the minute changes in the patient with respect to the treatment, but also to objectively judge the effect of the therapeutic agent.

Claims (4)

A method for inspecting human intervertebral discs using diffusion MRI.
Using Q-space imaging (QSI) analysis or diffusion tensor imaging (DTI) analysis
Kurtosis (KT) Fractional Anisotropy (FA) (KT FA), probability at zero displacement FA (ZP FA), full width at half maximum FA (FWH) by QSI analysis
Lambda ₁ mapping Ru good to DTI analysis, inspection method characterized by analyzing the at least one of _{(λ 2 + λ 3) /} 2 mapping. The inspection method according to claim 1 is
An inspection method characterized by quantifying intervertebral disc degeneration. From the spectral data of each voxel of the target intervertebral disc obtained by the MRI device,
Extract b value, axis information, diffusion time,
KT FA by QSI analysis, ZP FA, FWHM FA, lambda ₁ mapping Ru good to DTI analysis, and performing an operation by at least one of _{(λ 2 + λ 3) /} 2 mapping,
A step that quantitatively expresses the state of the target intervertebral disc based on the calculation result,
A program that causes a computer to run. The step of displaying the calculation result as an image
The program according to claim 3, which is executed by a computer.

Images