Method of combining imaging with serum GDNF index to judge the degree of cognitive impairment in Parkinson's disease

Abstract

The present invention discloses a method for judging the degree of cognitive impairment of Parkinson's disease by combining imaging with serum GDNF index, and relates to the technical field of judging the degree of cognitive impairment of Parkinson's disease, specifically a method for judging the degree of cognitive impairment of Parkinson's disease by combining imaging with serum GDNF index, comprising the following steps: S1, performing GE3.0TMRI scanning on the included subjects; S2, transferring the original imaging image file from the MR scanner to the computer and saving it using eFilmWorkstation. The present invention detects and analyzes the network connectivity index in the imaging of three groups of patients, and finds that the degree center values ​​of the right frontal lobe and parietal lobe have changed, and can well predict whether the disease is present after combining the GDNF level, and can well predict the degree of cognitive impairment after combining the PD course and medication status-LED, and in addition, after measuring the cerebral cortical thickness, it is found that the thickness of the cortex of the left hemisphere middle frontal gyrus is most correlated with the scores of each cognitive domain, and can effectively distinguish the mild and moderate cognitive impairment of PD after combining the serum GDNF index, PD course, and education level.

Description

Technical Field

[0001] The present invention relates to the technical field of judging the degree of cognitive impairment in Parkinson's disease, and in particular to a method for judging the degree of cognitive impairment in Parkinson's disease by combining imaging with serum GDNF index. Background Art

[0002] Parkinson's disease with cognitive impairment is very common and prevalent, and evidence shows that cognitive impairment often occurs before motor symptoms. At all stages of the PD process, the presence of cognitive impairment will lead to a significant increase in mortality and affect major daily activities. So, for PD with cognitive impairment, or in the prodromal stage of clinical PD, can we find biomarkers that can reflect the neuronal degeneration process early and are easy to obtain, so as to improve the prediction of PD progression, and may serve as future mechanism-targeted treatment, or as a risk factor for identifying PD. This will help in the early detection and differential diagnosis of PD and monitor disease progression.

[0003] Glial cell line-derived neurotrophic factor is an effective protective survival factor for central dopaminergic neurons. Abnormal GDNF also has an important impact on the occurrence of cognitive impairment in PD. There is evidence that after the death of PD patients, there are changes in the expression and dysfunction of GDNF in the skull, and synchronous changes in GDNF in the circulation level have been confirmed. Its impact on PD with cognitive dysfunction and its correlation with clinical status need further study. In current brain science research, functional magnetic resonance imaging is the most commonly used method to study brain function and morphology. Based on resting-state image analysis, it was found that patients with severe cognitive impairment in PD had changes in cortical thickness and network connectivity in the cortical and subcortical areas.

[0004] One of the purposes of the present invention is to explore whether serum GDNF levels are related to the occurrence of PD-CI, and what type of cognitive impairment is most closely related to it based on a clinical case-control study; further to find and combine objective imaging markers to explore whether the connectivity of the neural network of PD patients with different serum GDNF levels changes, and the network node areas that are changed. Finally, by establishing multimodal biological diagnostic markers for serum GDNF, differential biological indicators in imaging, basic medical history (including medication status) information, etc., it is possible to make early predictions of PD and the degree of cognitive dysfunction. Summary of the invention

[0005] In view of the deficiencies of the prior art, the present invention provides a method for determining the degree of cognitive impairment in Parkinson's disease by combining imaging with serum GDNF index, which solves the problems raised in the above-mentioned background technology.

[0006] To achieve the above objectives, the present invention is implemented by the following technical scheme: A method for judging the degree of cognitive impairment in Parkinson's disease by combining imaging with serum GDNF index, comprising the following steps:

[0007] S1. Perform GE3.0TMRI scan on the included subjects;

[0008] S2, transfer the original imaging image files from the MR scanner to the computer and save them using eFilmWorkstation;

[0009] S3, use MATLAB for data analysis;

[0010] S4. Determine the degree center value of the difference cluster and combine it with the serum GDNF index to participate in the prediction analysis of cognitive outcome indicators;

[0011] S5. Determine the cortical area with the highest correlation with cognitive domain scores, and use the cortical thickness combined with serum GDNF index to predict the outcome of cognitive impairment;

[0012] S6. Calculate the cortical thickness of the region of interest on the cortical surface;

[0013] S7. By comparing the cortical thickness of each brain region in each group of subjects, we can obtain the cortical regions with significant differences, and compare the results with the cognitive results.

[0014] S8. Combine serum GDNF index to participate in the prediction analysis of cognitive status.

[0015] Optionally, the subjects in step S1 are divided into three groups, and based on the K-means clustering results and feature comparison of serum GDNF, the groups are named PD-high-GDNF group and PD-low-GDNF group, and HC is used as a normal control group.

[0016] Optionally, the step S1 specifically includes the following steps:

[0017] S101. Each subject underwent non-contrast-enhanced MRITWI2 sequence and 3DT1WI brain volume sequence scans as a baseline;

[0018] S102, using three-dimensional positioning, axial T1W1 scanning, 192 slices, covering the whole brain;

[0019] S103, using a diffusion-weighted echo-planar sequence, with the front-back joint line as the scanning baseline, and performing an oblique axial DTI scan;

[0020] S104. Finally, collect BOLD images, with a total of 36 layers.

[0021] Optionally, the step S3, using MATLAB to perform data analysis, further includes the following steps: based on the functional network of each voxel, the degree centrality is calculated by calculating the number of suprathreshold correlations for each subject.

[0022] The present invention provides a method for judging the degree of cognitive impairment in Parkinson's disease by combining imaging with serum GDNF index, which has the following beneficial effects:

[0023] In the method of combining imaging with serum GDNF index to judge the degree of cognitive impairment in Parkinson's disease, the imaging network connectivity index of three groups of patients, PD-high-GDNF group, PD-low-GDNF group and control group, was detected and analyzed, and it was found that the degree center values ​​of the right frontal lobe and parietal lobe changed. The DC values ​​of two clusters (Cluster 1, 2) (located in the frontal lobe and parietal lobe) combined with serum GDNF content, disease course, and LED index can be used to identify whether PD is present and to judge the degree of cognitive impairment; the PD group showed a downward trend in the ability to separate and integrate information, among which the PD-low-GDNF group decreased more; 3. In the cortical thickness analysis, the reduction in the cortical thickness of the middle frontal gyrus in the PD group was closely related to the reduction in cognitive function. The combination of PD disease course, serum GDNF content, and education level indicators can distinguish PD with mild and moderate cognitive impairment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 , three Schematic diagram of brain regions with differences in group degree centrality (cluster 1) and their correlation with clinical cognitive characteristics of all subjects;

[0025] (A) Rendering of cluster 1 (B) axial slice view; color bar indicates f value; coordinate area is mainly located in the right frontal lobe; DC clustering was performed at a threshold of P < 0.05, and multiple comparison correction was performed using the false discovery rate (FDR) method; (C) degree centrality value of group differences in cluster 1; *P < 0.05; ****P < 0.0001. (D, E) DC1 scatter plot of cluster 1 and its correlation with MMSE and MoCA scores;

[0026] Figure 2 , three Schematic diagram of brain regions with differences in group degree centrality (cluster 2) and their correlation with clinical cognitive characteristics of all subjects;

[0027] Among them, (A) rendering of cluster 2 (B) axial slice view. The color bar indicates the f value. The coordinate area is mainly distributed in the right parietal lobe; DC clustering was performed at a threshold of P < 0.05, and the false discovery rate (FDR) method was used for multiple comparison correction; (C) The degree centrality value of the difference between groups in cluster 1. *P < 0.05; ****P < 0.0001, ns, the difference is not statistically significant; (D, E) The DC of cluster 2 scatter plot and the linear regression of the significant correlation between DC and cognitive function (MMSE, MoCA) in all subjects;

[0028] Figure 3 , three Dimensional scatter plot correlation profile diagram;

[0029] The x-axis, y-axis, and z-axis represent serum GDNF, degree centrality of clusters (cluster 1: A, C, cluster 2: B, D), and cognitive scores (mmse: A, B, moca: C, D); the colored points represent different groups; the individual clusters within the same group are closely related, and the difference between the PD group and the HC group is the most obvious;

[0030] Figure 4 , Schematic diagram of the receiver operating characteristic (ROC) curve obtained when predicting Parkinson's disease and different degrees of cognitive impairment based on the single index and comprehensive index of degree centrality;

[0031] Among them, (A) ROC curve for validation of Parkinson's disease diagnosis. (B) ROC curve for validation of moderate cognitive impairment based on MMSE score in PD subgroup; (C) ROC curve for validation of moderate cognitive impairment based on MoCA score in PD subgroup; Cluster 1: blue line, 2: green line; Combination 1 (brown line): DC and serum GDNF of 1 and 2; Combination 2 (red line): duration, LEDD, serum GDNF and DC of 1 and 2;

[0032] Figure 5 , Schematic diagram of comparison of cortical thickness differences;

[0033] Among them, (A) Cortical structure, with statistically significant changes in different colors to distinguish the thickness of each cortical area; The analysis results of cortical thickness showed that compared with the healthy control group, the cortical thickness of PD patients (PD-high-GDNF & PD-low-GDNF) was thinner, showing significant different areas of the standardized brain (average of all subjects); In particular, the cortical thickness of 9 brain regions in the PD-low-GDNF group was significantly reduced; (B) Statistical comparative analysis was performed, based on the one-way analysis of variance, P<0.05, and the above 9 regions were statistically different;

[0034] Figure 6 , Schematic diagram of the receiver operating characteristic (ROC) curve of left midfrontal thickness as a single index and a composite index to predict Parkinson's disease and different degrees of cognitive impairment;

[0035] Among them, (A) ROC curve for verifying the diagnosis of Parkinson's disease based on L_caudal mid-frontal thickness; (B) ROC curve for verifying moderate cognitive impairment based on MMSE score in PD subgroup; (C) ROC curve for verifying moderate cognitive impairment based on MoCA score in PD subgroup; L_caudal mid-frontal thickness: blue line; L_caudal mid-frontal thickness and serum GDNF: green line; L_caudal mid-frontal thickness, serum GDNF and duration: yellow line; L_caudal mid-frontal thickness, serum GDNF and duration and education time: red line.

[0036] Figure 7 , Schematic diagram of the analysis of serum GDNF index involved in the prediction of cognitive status. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0038] As attached Figure 1-7 As shown in FIG. 1 , the method of combining imaging with serum GDNF index to determine the degree of cognitive impairment in Parkinson's disease includes the following steps:

[0039] S1. Perform GE3.0TMRI scans on the included subjects: each subject undergoes non-contrast enhanced MRITWI2 sequence and 3DT1WI brain volume sequence scans as baseline, uses three-dimensional positioning, performs axial T1W1 scans, 192 layers, and covers the whole brain, and uses diffusion-weighted-echo-planar sequence, uses the front-to-back joint line as the scanning baseline, performs oblique axial DTI scans, and finally acquires BOLD images, a total of 36 layers. Among them, the subjects in step S1 were divided into three groups, and according to the K-means clustering results and feature comparisons, they were named and grouped into PD-high-GDNF group and PD-low-GDNF group, and HC was used as the normal control group;

[0040] S2, transfer the original imaging image files from the MR scanner to the computer and save them using eFilmWorkstation;

[0041] S3. Data analysis was performed using MATLAB: degree centrality was calculated by counting the number of suprathreshold correlations for each subject based on the functional network of each voxel;

[0042] S4. Determine the degree center value of the difference cluster and combine it with the serum GDNF index to participate in the prediction analysis of cognitive outcome indicators;

[0043] S5. Determine the cortical area with the highest correlation with cognitive domain scores, and use the cortical thickness combined with serum GDNF index to predict the outcome of cognitive impairment;

[0044] S6. Calculate the cortical thickness of the region of interest on the cortical surface;

[0045] S7. By comparing the cortical thickness of each brain region in each group of subjects, we can obtain the cortical regions with significant differences, and compare the results with the cognitive results.

[0046] S8. Combine serum GDNF index to participate in the prediction analysis of cognitive status.

[0047] result

[0048] 1. Comparison of degree centrality (DC) differences and cognitive correlation analysis of resting-state imaging indicators among groups

[0049] In the comparison of the three groups, after voxel-based clusterp<0.05 FDR (False Discovery Rate) correction, it was found that there were two clusters with significant differences in DC values ​​(see Table 1 below, areas with significant differences in degree centrality among the groups), one located in the right frontal lobe (cluster 1) and the other in the right parietal lobe (cluster 2); the DC value of cluster 1 showed an "inverted U-shaped" relationship with cognitive scores (MMSE and MoCA); while the DC value of cluster 2 showed a negative correlation with cognitive assessment scores.

[0050] Table 1. Regions with significant differences in degree centrality among groups

[0051]

[0052] In the right frontal region of the PD group, cluster 1 was mainly located in the inferior frontal gyrus and extended to the middle frontal gyrus, brodmannarea 46 and area 10 ( Figure 1 B); The DC of Cluster 1 increased significantly compared with the healthy control group, and the increase in the PD-high-GDNF group was more obvious, and the difference was statistically significant ( Figure 6 C), and further compared the PD-high-GDNF and PD-low-GDNF groups. There were also differences between the two groups. The DC value of cluster 1 was reduced in the PD-low-GDNF group ( Figure 1 C, *P<0.05), and analyzed the correlation between the DC value of the cluster and the clinical cognitive assessment score of the subjects. The scatter plot showed that among all the subjects, the DC value of the cluster and the cognitive scores (MMSE and MoCA) showed an "inverted U" correlation ( Figure 1 D, E), P < 0.05, see Table 2 for details. Pearson correlations between degree centrality and cognitive scores and GDNF levels were performed for all subjects and the PD cohort.

[0053] In the right parietal area of ​​the PD group ( Figure 2 A), the cluster 2 centrality DC of the cluster was also increased compared with the normal population. After further differentiation of GDNF content, it was found that the cluster centrality value of the PD-low-GDNF group increased significantly, and the difference was statistically significant ( Figure 2 C); however, there was no difference between the PD-high-GDNF group and the control group ( Figure 2 C, ns, P>0.05); Further three-dimensional analysis of the mass showed that it extended to the postcentral gyrus, brodmannarea 7 and 5 areas ( Figure 2 B); Correlation analysis between the DC value of cluster2 and cognitive function scores showed a negative correlation with cognitive scores (MMSE and MoCA) ( Figure 2 D, E), the differences were statistically significant, as shown in Table 2. Pearson correlations between degree centrality and cognitive scores and GDNF levels were performed for all subjects and the PD cohort, respectively.

[0054] In order to explore the relationship between serum GDNF, degree centrality of difference clusters, and cognitive assessment scores, a three-dimensional scatter plot was performed on the research data set of all subjects. The HC normal reconciliation group, PD-high-GDNF, and PD-low-GDNF groups were clearly distinguished ( Figure 3 ), and the distinction between PD and HC was obvious. Further combined with the person correlation analysis of the two factors and the results of the first part of the study, in all PD subjects, the DC of Cluster1 and 2 was correlated with clinical cognitive assessment (MMSE, MoCA), and was statistically significant (Table 2, all subjects, PD cohorts were analyzed for Pearson correlation between degree centrality and cognitive scores, GDNF levels); the previous part of the study showed that there was a great difference in cognitive assessment between the PD-high-GDNF and PD-low-GDNF groups, and the negative correlation of the DC value of Cluster2 was also statistically significant in all subject groups; however, there was no linear relationship between the DC value of Cluster1 (Table 2, all subjects, PD cohorts were analyzed for Pearson correlation between degree centrality and cognitive scores, GDNF levels).

[0055] Table 2 Pearson correlation analysis between degree centrality and cognitive scores and GDNF levels for all subjects and PD cohort

[0056]

[0057] 2. ROC analysis of degree centrality (DC) for judging PD and mild to moderate cognitive impairment

[0058] We further studied the ability of Cluster 1 and 2 to distinguish clinical PD and cognitive impairment, and performed ROC analysis on the data ( Figure 4 ); In the judgment analysis of Cluster 1 and 2 for PD and HC, it was found that the DC analysis results of the two clusters could effectively distinguish whether PD was present, specifically, the AUC area of ​​cluster 1 was 0.615, and the AUC area of ​​cluster 2 was 0.676. Only cluster 2 could effectively distinguish PD and HC populations (P=0.0.27). The specific results are shown in Table 3. The ROC analysis results for Parkinson's disease were evaluated by combining the serum GDNF test results. The serum results were combined with the two cluster results of the imaging analysis, and the diagnostic efficiency was increased, with an AUC of 0.956, a sensitivity of 93.3%, and a specificity of 87.5%, P<0.000.

[0059] Table 3. ROC analysis results for Parkinson's disease

[0060]

[0061] Combination: 1:2 DC, serum GDNF

[0062] Based on the scores and judgment criteria of clinical MMSE and MoCA, the cognitive impairment of the PD population was mainly mild cognitive impairment and moderate cognitive impairment. The analysis showed that for the MMSE results, clusters 1 and 2 could not be used alone as judgment indicators to distinguish the degree of cognitive impairment. However, after combining the blood GDNF results and comprehensive prediction indicators (combination 1), its AUC reached 0.760, with a sensitivity and specificity of 80% and 76%, respectively. The result was not statistically significant (P=0.071). If combined with the course of disease and medication, the AUC reached 0.808, which was statistically significant (P= 0.032) See Table 4 for details, ROC analysis of MMSE in judging the cognitive status of PD patients; for MoCA results, clusters 1 and 2 alone cannot determine the degree of cognitive impairment, and even combined with GDNF results, they cannot be accurately distinguished, with no statistical significance. However, if combined with the course of disease, the LED volume information index is included as a comprehensive prediction index (combination 2), which can effectively distinguish the type of cognitive impairment under MoCA scores, AUC = 0.7965, sensitivity of 75%, specificity of 77.3%, see Table 5 for details, ROC analysis of MoCA in judging the cognitive status of PD patients.

[0063] Table 4. ROC analysis of MMSE in judging cognitive status of PD patients

[0064]

[0065] Combination 1: 1 and 2 clusters DC, serum GDNF

[0066] Combination 2: Duration of clusters 1 and 2, LEDD, serum GDNF, and DC

[0067] Table 5. ROC analysis of MoCA in judging cognitive status of PD patients

[0068]

[0069] Combination 1: 1 and 2 clusters DC, serum GDNF

[0070] Combination 2: Duration of clusters 1 and 2, LEDD, serum GDNF, and DC

[0071] 3. There are major differences in the thickness of the frontal and temporal cortices

[0072] In the cortical thickness analysis, compared with HC, PD patients had decreased cortical thickness in multiple brain regions, and the cortical regions with statistical differences included ( Figure 5 A): The caudal part of the middle frontal gyrus, fusiform gyrus, inferior temporal lobe, triangular area of ​​the inferior frontal gyrus, superior frontal gyrus, and superior temporal gyrus in the left hemisphere; the fusiform gyrus, middle temporal gyrus, and superior temporal gyrus in the right hemisphere; after statistical analysis, compared with the healthy control group, the thickness of the above 9 cortical areas in the PD population showed a thinning trend, among which the cortical thickness of the PD-low-GDNF group decreased most significantly ( Figure 5 B), for specific comparison information, see Table 6, Differences and comparisons of cortical thickness between groups.

[0073] Table 6. Differences and comparisons of cortical thickness among groups

[0074]

[0075] After comparing 68 brain regions, the main differential brain regions were frontal and temporal regions. The cortical thickness was further correlated with the cognitive assessment results of the subjects. The results showed that not all differential brain regions were correlated with cognitive scores. Only 6 cortical regions were positively correlated with cognitive results, that is, the lower the cognitive score, the more the cortical thickness decreased. See Table 7 for details, Correlation analysis between cognitive function scores and cortical thickness.

[0076] Table 7. Correlation analysis between cognitive function scores and cortical thickness

[0077]

[0078] It is known that the prefrontal cortex is involved in the regulation of multiple cognitive functions, including emotion regulation and cognitive function coordination, including attention, execution and memory. Therefore, we further conducted spermman correlation analysis on the six cortices and sub-cognitive domains. The results showed that the left hemisphere middle frontal gyrus was positively correlated with orientation, memory, attention and execution, with statistical differences (P<0.05, Table 8, Correlation analysis of cognitive domain scores and different cortical thicknesses); and there was a marginal statistical difference in correlation with naming function (P=0.051); in addition, the superior frontal gyrus, superior temporal gyrus, and fusiform gyrus were related to the results of some cognitive domain assessments, although some were only marginally significant, but this did not affect the judgment that the frontal cortex was closely related to cognitive function. It can be seen that the reduction in the thickness of the middle frontal gyrus cortex in PD patients is closely related to the reduction of cognitive function.

[0079] Table 8. Correlation analysis between cognitive domain scores and different cortical thicknesses

[0080]

[0081] Given the importance of the middle frontal gyrus in the left hemisphere, further exploration was conducted through subject regression curve analysis ( Figure 6 ), and found that the area under the curve of the middle frontal gyrus cortical thickness in distinguishing Parkinson's patients from healthy people was 0.745 (95% CI: 0.586, 0.905), P = 0.012; when the cutoff value was 2.374mm, it could have a high sensitivity (100%) and specificity (50%). However, in the PD population, the single middle frontal gyrus thickness index could not effectively distinguish the mild and moderate cognitive impairment of MMSE and MoCA scores. Therefore, we tried to combine other parameter indicators for differential diagnosis and found that the middle frontal gyrus thickness combined with serum GDNF and PD course and years of education can be valuable for MMSE identification of mild and moderate cognitive impairment. =0.902 (95%CI: 0.719, 1.000), P=0.030, sensitivity=100% and specificity=70.6%; if the mild and moderate cognitive impairment were judged based on MoCA, the middle frontal gyrus thickness and serum GDNF combined with the PD course could effectively distinguish the mild and moderate cognitive impairment, AUC=0.938 (95%CI: 0.828, 1.000), P=0.008, sensitivity=100% and specificity=87.5%; in summary, the left middle frontal gyrus thickness index, PD course (years), serum GDNF content and education level comprehensive index can effectively distinguish the mild and moderate cognitive impairment of PD.

[0082] The above are only preferred specific implementation modes of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solutions and inventive concepts of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. The method of combining imaging with serum GDNF index to determine the degree of cognitive impairment in Parkinson's disease includes the following steps: S1. Perform GE3.0TMRI scan on the included subjects; S2, transfer the original imaging image files from the MR scanner to the computer and save them using eFilmWorkstation; S3, use MATLAB for data analysis; S4. Determine the degree center value of the difference cluster and combine it with the serum GDNF index to participate in the prediction analysis of cognitive outcome indicators; S5. Determine the cortical area with the highest correlation with cognitive domain scores, and use the cortical thickness combined with serum GDNF index to predict the outcome of cognitive impairment; S6. Calculate the cortical thickness of the region of interest on the cortical surface; S7. By comparing the cortical thickness of each brain region in each group of subjects, we can obtain the cortical regions with significant differences, and compare the results with the cognitive results. S8. Combine serum GDNF index to participate in cognitive status prediction analysis; The step S3, using MATLAB to perform data analysis, further includes the following steps: calculating the degree centrality by calculating the number of suprathreshold correlations for each subject based on the functional network of each voxel; and measuring the cortical thickness after the network topological properties change.

2. The method for judging the degree of cognitive impairment in Parkinson's disease by combining imaging with serum GDNF index according to claim 1, characterized in that: The subjects in step S1 were divided into three groups, and based on the K-means clustering results and feature comparison of serum GDNF results, the groups were named PD-high-GDNF group and PD-low-GDNF group, and HC was used as the normal control group.

3. The method for judging the degree of cognitive impairment in Parkinson's disease by combining imaging with serum GDNF index according to claim 1, characterized in that: The step S1 specifically includes the following steps: S101. Each subject underwent non-contrast-enhanced MRITWI2 sequence and 3DT1WI brain volume sequence scans as a baseline; S102, using three-dimensional positioning, axial T1W1 scanning, 192 slices, covering the whole brain; S103, using a diffusion-weighted echo-planar sequence, with the front-back joint line as the scanning baseline, and performing an oblique axial DTI scan; S104. Finally, collect BOLD images, with a total of 36 layers.

Images