A method for localizing and classifying hypothalamic nuclei using spatial metabolomics mass spectrometry imaging
By performing continuous slicing of brain tissue and combining spatial registration of histological images with mass spectrometry imaging data, the problem of subjective judgment by the operator in the localization and division of hypothalamic nuclei was solved, achieving accurate localization of nuclei and reliability and repeatability of analysis results.
Patent Information
- Application Number
- CN202611585151.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-14
- Publication Date
- 2026-11-17
AI Technical Summary
The lack of standardized operating procedures in current technologies leads to subjective judgment by the operator during the localization and division of hypothalamic nuclei, affecting the accuracy and repeatability of the analysis results.
By obtaining continuous slices of brain tissue and combining spatial registration of histological images with mass spectrometry imaging data, a spatial coordinate correspondence between histological images and mass spectrometry imaging data is established, enabling accurate localization and division of hypothalamic nuclei.
It improves the accuracy and repeatability of hypothalamic nucleus localization, reduces human bias, and enhances the reliability and traceability of analysis results.
Smart Images

Figure CN123118718A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of spatial metabolomics mass spectrometry imaging technology, specifically relating to a method for locating and classifying hypothalamic nuclei in spatial metabolomics mass spectrometry imaging. Background Technology
[0002] Mass spectrometry imaging (MSI) is a novel molecular imaging technique that can directly obtain the structure, abundance, and spatial distribution of a large number of known or unknown endogenous metabolites and exogenous drugs from biological tissues. This technique enables qualitative, quantitative, and localization analysis of thousands of metabolites in biological tissues. Combined with bioinformatics analysis, it can identify differentially expressed metabolites in situ and elucidate their biological functions, offering significant advantages such as high specificity, high throughput, and preservation of spatial information.
[0003] The hypothalamus is a crucial brain region regulating appetite, energy expenditure, and endocrine homeostasis, playing a vital role in the development of metabolic diseases such as obesity. However, the hypothalamus is composed of numerous small, adjacent, and functionally distinct nuclei, making accurate identification of these functionally differentiated nuclei essential for spatial metabolomics research. Mass spectrometry imaging primarily reflects the distribution and intensity of ions, and its histological boundaries are typically less clear than those of hematologic and epithelial (H&E) images. Furthermore, since histological sections are usually adjacent to mass spectrometry imaging sections, differences in orientation, size, and local morphology can easily arise during preparation. Manually selecting regions of interest solely based on mass spectrometry ion maps can easily lead to the inclusion of adjacent nuclei or periventricular tissue in the target area, affecting the accuracy of the analytical results.
[0004] Currently, the entire process, from brain slice localization, atlas calibration, and ion imaging overlay to region division, node adjustment, and ROI migration between positive and negative ion modes, lacks unified operational standards, and each step may be affected by the operator's subjective judgment. Therefore, a standardized method is needed to unify brain atlas-H&E image localization with image overlay, irregular region selection, ROI export, and reuse in mass spectrometry imaging software, in order to improve the reproducibility and reliability of hypothalamic nucleus spatial metabolomics research. Summary of the Invention
[0005] The purpose of this invention is to overcome the shortcomings of the prior art and to provide a method for locating and classifying hypothalamic nuclei using spatial metabolomics mass spectrometry imaging.
[0006] The specific technical solution adopted in this invention is as follows:
[0007] In a first aspect, the present invention provides a method for locating and classifying hypothalamic nuclei using spatial metabolomics mass spectrometry imaging, the specific steps of which are as follows:
[0008] S1: Obtain isolated brain tissue containing the hypothalamus and its surrounding complete anatomical structures, and perform serial slicing on the brain tissue sample to obtain adjacent slices as histological slices and mass spectrometry imaging slices, respectively.
[0009] S2: Perform histological staining and digital imaging on the histological sections to obtain histological images;
[0010] S3: Based on the slice direction, stereoscopic coordinates, and anatomical landmarks of the histological image, select a reference atlas slice image that matches the histological image from the corresponding animal model brain stereoscopic atlas;
[0011] S4: Spatial registration is performed between the reference atlas section image and the histological image. The boundaries of the hypothalamic nuclei in the reference atlas section image are mapped onto the histological image. The boundaries of the nuclei are corrected according to the histological structures shown in the histological image to obtain a superimposed image containing the histological structures and the boundaries of the nuclei. Based on the boundaries of the nuclei in the superimposed image, one or more target hypothalamic nuclei are determined to form a region of interest.
[0012] S5: Perform mass spectrometry imaging scans on the corresponding regions of interest in the mass spectrometry imaging slices to obtain mass spectrometry imaging data; convert the mass spectrometry imaging data into a standardized format into a reference ion map reflecting the spatial contour of the tissue in the mass spectrometry imaging slices using mass spectrometry imaging software.
[0013] S6: Overlay and spatially register the superimposed image obtained in step S4 and the reference ion map obtained in step S5 to match the tissue contours in the two images and establish the spatial coordinate correspondence between the histological image and the mass spectrometry imaging data.
[0014] S7: Based on the spatial coordinate correspondence, obtain the mass spectrometry imaging data corresponding to the target hypothalamic nucleus in the superimposed image, and finally obtain the signal of the target metabolite in the different target hypothalamic nuclei.
[0015] Preferably, the brain tissue is rat brain tissue or mouse brain tissue, and a stereotactic brain atlas matching the animal model is selected.
[0016] Preferably, in step S1, the thickness of the continuous slices is controlled at 8~12 μm.
[0017] Preferably, in step S1, the histological sections are flattened and attached to a glass slide, and placed in a 37°C constant temperature oven for later use; the mass spectrometry imaging sections are flattened and attached to a positive charge desorption plate, and stored in a -80°C refrigerator for later use.
[0018] Preferably, the histological sections in step S2 are stained using hematoxylin-eosin staining; the digital imaging is performed using a digital slide scanner.
[0019] Preferably, the anatomical landmarks mentioned in step S3 are selected from at least three of the following: optic chiasm, fornix, mammillary body, pituitary stalk, hippocampus, and outer edge of brain tissue.
[0020] Preferably, in step S4, spatial registration is performed using the image processing software ImageJ, specifically as follows: the histological image is set as the target image, and the reference atlas section image is set as the moving image; several corresponding anatomical landmark structures are marked on the histological image and the reference atlas section image respectively; the reference atlas section image is spatially transformed according to the marked anatomical landmark structures, so that the external shape and internal anatomical structure of the brain tissue in the reference atlas section image match the corresponding tissue structure in the histological image, and the boundary of the hypothalamic nucleus in the reference atlas section image is mapped onto the histological image.
[0021] Preferably, the standardized format in step S5 is the imzML format; the mass spectrometry imaging software uses MSiReader software, which displays the two-dimensional spatial coordinates of each scanned pixel and its corresponding ion intensity information.
[0022] Preferably, the target hypothalamic nucleus includes one or more of the arcuate nucleus, paraventricular nucleus, ventromedial nucleus, dorsomedial nucleus, and posterior nucleus.
[0023] Secondly, the present invention provides an application of the hypothalamic nucleus localization and segmentation method using spatial metabolomics mass spectrometry imaging as described in the first aspect, to obtain signals of metabolites in target hypothalamic nuclei in brain tissue samples under different experimental conditions, and to analyze the differences in metabolic characteristics of the target hypothalamic nucleus under different experimental conditions based on the metabolite signals.
[0024] Compared with the prior art, the present invention has the following advantages:
[0025] (1) This invention uses continuous slicing of brain tissue samples and selects adjacent slices as histological slices and mass spectrometry imaging slices, respectively, so that the histological slices and mass spectrometry imaging slices have a high degree of consistency in the spatial location of the tissue, providing a reliable spatial correspondence basis for subsequent use of histological information to locate and classify the hypothalamic nuclei in mass spectrometry imaging data.
[0026] (2) This invention achieves accurate localization of hypothalamic nuclei by spatial registration of reference atlas cross-sectional images and histological images, and by correcting the boundaries of nuclei in combination with actual histological structures; further, by spatial registration of histological images and reference ion maps, a spatial coordinate correspondence between histological images and mass spectrometry imaging data is established, thereby mapping the target hypothalamic nuclei to the corresponding region in the mass spectrometry imaging data, improving the accuracy of nucleus localization and regional metabolite signal extraction.
[0027] (3) Based on the region of interest and spatial location information of the target hypothalamic nucleus, the present invention extracts the corresponding mass spectrometry imaging data and can verify and reuse the spatial location of the target nucleus under different ion acquisition modes, reducing the human bias that may be introduced by repeated selection of nucleus regions under different ion modes, which is conducive to improving the consistency of nucleus region division between different ion modes and different samples.
[0028] (4) By preserving the location and serial number information of brain tissue slices, image spatial registration information, region of interest information and its spatial location information, this invention enables the analysis process of hypothalamic nuclei from histological localization, map boundary mapping, mass spectrometry imaging spatial matching to regional metabolite signal extraction to have good traceability and reproducibility, which is conducive to improving the reliability of hypothalamic nuclei analysis results in spatial metabolomics mass spectrometry imaging. Attached Figure Description
[0029] Figure 1 The image shown is a superimposed image of rat histological structures and nucleus boundaries, as shown in Example 1.
[0030] Figure 2 This is a schematic diagram showing the division of the target hypothalamic nucleus region in rat brain tissue in Example 1;
[0031] Figure 3 The image shows the principal component analysis results of the metabolic profile of the target hypothalamic nuclei in the normal group and the obese group rats in Example 1, where (a), (b) and (c) represent the arcuate nucleus, ventromedial nucleus and paraventricular nucleus, respectively. Detailed Implementation
[0032] The present invention will be further described and illustrated below with reference to the accompanying drawings and specific embodiments. The technical features of each embodiment of the present invention can be combined accordingly, provided that there is no mutual conflict.
[0033] Example 1
[0034] This embodiment provides a method for locating and classifying hypothalamic nuclei using spatial metabolomics mass spectrometry imaging, as detailed below. It should be noted that "acquisition" in this invention refers to acquiring isolated brain tissue, and does not include the process of performing surgery on a living organism.
[0035] Step 1: Obtaining, embedding, and preparing serial frozen sections of brain tissue
[0036] Five-week-old obese rats and normal-weight rats were anesthetized intraperitoneally with 2% sodium pentobarbital. Brain tissue was harvested after confirming deep anesthesia. The brain tissue was gently separated from the meninges and cranial nerves while minimizing damage to the brain parenchyma, and the entire brain tissue was extracted intact. Residual fluid on the surface of the brain tissue was gently blotted with absorbent paper, and the tissue was placed in Leica cryo-embedding medium. The orientation of the brain tissue was adjusted, and the vertical and horizontal positions were marked on the outside of the embedding cassette, and photographs were taken. The embedded brain tissue samples were then rapidly frozen horizontally in a dry ice box and subsequently transferred to a -80°C freezer for storage.
[0037] Brain tissue embedded in embedding medium was removed from a -80°C freezer and equilibrated in a cryostat set to -20°C for approximately 30 minutes. The brain tissue was then fixed to a sample base using embedding medium, and its orientation was adjusted according to the target section. After the embedding medium was completely frozen, rough trimming was performed to remove excess embedding medium. As the tissue approached the target anatomical region, anti-rolling adjustments were made, and sections were continuously sliced. Adjacent sections were flattened and attached to positively charged desorption plates and ordinary glass slides, respectively, to obtain mass spectrometry imaging sections and histological sections. The mass spectrometry imaging sections were stored at -80°C for later use. The histological sections were dried in a 37°C incubator for subsequent H&E staining and histological localization.
[0038] Step 2: Histological section staining and digital imaging
[0039] The above histological sections were stained with hematoxylin and eosin (H&E): the sections were immersed in hematoxylin staining solution for about 7 minutes, rinsed with pure water, and then blued; rapid differentiation was performed with ethanol differentiation solution, followed by re-bluing in pure water; then eosin staining was performed. After staining, the sections were sequentially dehydrated with graded ethanol and cleared with xylene, and finally mounted with neutral resin mounting medium and coverslips. After the mounting medium had fully cured, the histological images were observed and acquired under a digital section scanner.
[0040] Step 3: Registration of histological images and spectral maps
[0041] Based on the sagittal plane orientation of the slices and the main visible anatomical structures, a section approximately 0.40 mm lateral to the midsagittal plane was selected from the rat brain stereotaxic atlas as a reference atlas. This selection was based on the fact that, compared to adjacent sections in the atlas series, this section showed the largest exposed area and clearest boundaries for key hypothalamic nuclei (such as the arcuate nucleus and paraventricular nucleus), making it the most suitable standard atlas for this experiment's slices and providing accurate anatomical reference for subsequent nucleus localization.
[0042] Registration was performed using the BigWarp plugin in ImageJ software. The aforementioned histological image was set as the target image, and the reference atlas image (i.e., the atlas section approximately 0.40 mm lateral to the midsagittal plane) was set as the moving image. Corresponding landmarks for the third ventricle, optic chiasm, fornix, hippocampus, outer margins of brain tissue, and other identifiable anatomical structures were marked in both images. By adding and adjusting the landmarks, the external shape and internal anatomical structures of the brain tissue in the reference atlas image were matched with the histological sections. During registration, only the reference atlas image underwent spatial transformation, maintaining the original histological section's morphology unchanged, thereby mapping the nucleus boundaries in the reference atlas onto the histological image.
[0043] After registration, the transformed reference spectral image is exported using nearest-neighbor interpolation. The color channels of this image are then separated, retaining the red channel (the red channel, after separation, provides the largest grayscale difference between the target region and the background, helping to improve the accuracy of subsequent thresholding and grayscale quantification). The red channel image is inverted and converted to an 8-bit grayscale image, serving as the spectral boundary layer.
[0044] The processed grayscale image of the spectral boundaries is merged with the original histological image through channel merging to generate a superimposed image containing histological structures and nucleus boundaries, such as... Figure 1 As shown. On the overlaid image, based on the registered nucleus boundaries, the polygon selection tool can be used to draw point by point along the edge of the target hypothalamic nucleus to identify one or more target hypothalamic nuclei, forming a closed region of interest (ROI), which is then saved to the ROI Manager for subsequent nucleus region delineation in mass spectrometry imaging data.
[0045] In this embodiment, the paraventricular nucleus, posterior nucleus, dorsomedial nucleus, ventromedial nucleus, anterolateral nucleus, and arcuate nucleus in the hypothalamus are circled, as shown in the following figure. Figure 2 As shown.
[0046] Step 4: Mass spectrometry imaging scan and generation of reference ion map
[0047] Take out the mass spectrometry imaging slices stored in the -80℃ freezer in step one, dry them at about -20℃ for 1 hour, and then continue to dry them at room temperature for 2 hours. After the slice surface is fully dry, perform the instrumental analysis.
[0048] The slides were analyzed using an AFADESI-MSI platform (Aerodynamically Assisted Desorption / Electrospray Ionization Mass Spectrometry Imaging Platform) coupled with a Q-Orbitrap high-resolution mass spectrometer. Depending on the detection mode, corresponding spray solvent and ion source parameters were set: in negative ion mode, the spray solvent was an acetonitrile / water mixture (8:2, v / v); in positive ion mode, the spray solvent was an acetonitrile / water mixture containing 0.1% formic acid (8:2, v / v). In both modes, the spray solvent flow rate was set to 5 μL / min, the transport gas flow rate to 5 L / min, the spray voltage to 7 kV, and the distance from the sample surface to the sprayer to the ion transport tube to be 3 mm.
[0049] The mass spectrometry acquisition resolution was set to 70,000, the mass scan range was 70–1000 m / z, and the AGC target value was 2 × 10⁻⁶. 6 The maximum ion implantation time was 200 ms, the S-lens RF voltage was 55 V, and the capillary temperature was 350℃. The X-axis scan speed was set to 0.2 mm / s, and the Y-axis scan step size was set to 100 μm.
[0050] Under the above parameter conditions, the tissue region containing the corresponding region of interest in the mass spectrometry imaging slice is scanned line by line, and high-resolution mass spectrometry data at each scanning position is acquired to obtain mass spectrometry imaging data.
[0051] After mass spectrometry acquisition, the raw data is converted to imzML format (a standardized data exchange format for mass spectrometry imaging), and the converted imzML format data is imported into MSiReader software (available from the official website https: / / msireader.com / support) for data reading and analysis. MSiReader software can read the XY spatial coordinates and corresponding mass spectrometry or ion intensity information for each scan position.
[0052] Based on the preliminary qualitative analysis results of the compounds, a characteristic ion that can fully display the tissue profile was selected from the identified metabolites. The mass-to-charge ratio (m / z value) of this characteristic ion was entered into the "MS Navigation" interface of the MSiReader software. The software then extracted the signal intensity distribution of this ion across the entire tissue section and generated a corresponding tissue profile reference ion map.
[0053] Step 5: Spatial registration of the overlay image containing histological structures and nucleus boundaries with the reference ion map.
[0054] Launch the image overlay tool in MSiReader software and import the overlay image with nucleus boundaries processed in step three. Adjust the transparency of the histological image so that the histological results and the reference ionogram can be displayed simultaneously. Use the rotation, translation, and scaling functions in the image overlay tool to adjust the image. Registration and matching are performed using the outer contour of the tissue slice as the primary reference, combined with identifiable anatomical findings such as the optic chiasm for correction. Repeatedly fine-tune each parameter until the tissue contour in the histological image and the tissue contour in the reference ionogram achieve the best spatial match. This establishes a spatial coordinate correspondence between the histological image and the mass spectrometry imaging data, ensuring that the identified target hypothalamic nucleus corresponds to its corresponding spatial location in the mass spectrometry imaging slice.
[0055] After registration, save the transformation parameters (including rotation angle, translation amount, scaling ratio, etc.) of the current image overlay as a .mat file. Name the file according to the format "sample number + ion acquisition mode" for subsequent repeated loading and result verification.
[0056] Step 6: Acquisition of mass spectrometry imaging data corresponding to the target hypothalamic nuclei and extraction of metabolite signals
[0057] Based on the spatial coordinate correspondence established in the above steps between histological images and mass spectrometry imaging data, the target hypothalamic nuclei are mapped to the spatial coordinate system of the mass spectrometry imaging data. The boundaries of each target hypothalamic nucleus are verified, and the local contours and spatial positions are adjusted by dragging nodes or moving the entire selection area. For multiple nuclei, different colored contours can be used to assist in differentiation, but the contour color is not used as the basis for determining brain region classification.
[0058] After confirming the ROI location is correct, export the ROI spatial location information (i.e., the XY coordinates of each node forming the polygon in the image pixel coordinate system) using the "Export ROI Location Info" function. Name the file according to the format "Sample Number-Cluster-Ion Mode" and save the .mat and .txt files exported by MSiReader for subsequent ion information extraction, statistical analysis, and result verification.
[0059] Based on the XY spatial coordinates corresponding to the ROI of each target hypothalamic nucleus, the mass spectrometry or ion intensity information of each scanned pixel falling within the ROI range is obtained from the mass spectrometry imaging data, thereby obtaining the mass spectrometry imaging data corresponding to each target hypothalamic nucleus, and extracting the signals of each target metabolite from it.
[0060] For samples acquired simultaneously in both positive and negative ion modes, the saved ROI location file was loaded into the mass spectrometry imaging data of the other ion mode, and the corresponding histological images were re-overlaid. The correspondence between the ROI and the target hypothalamic nucleus was checked and corrected. ROIs for different samples and different target hypothalamic nuclei underwent independent histological verification to ensure the accuracy of target hypothalamic nucleus localization and extraction of corresponding metabolite signals.
[0061] Step 7: Metabolomics analysis of hypothalamic nuclei in obese and normal rat groups.
[0062] Following steps one through seven, brain tissue samples from normal and obese rats were processed to obtain mass spectrometry imaging data of key hypothalamic nuclei and corresponding ROI spatial location files for both groups of animals.
[0063] (1) Construction of a spatial distribution map of the nuclear region of differential metabolites
[0064] During the experiment, differential metabolite expression maps of key hypothalamic nuclei (arctic nucleus, paraventricular nucleus, anterolateral nucleus, ventromedial nucleus, dorsomedial nucleus, and posterior nucleus) were obtained in both the normal and obese groups. Based on literature review and the research objective of this study (the relationship between hypothalamic appetite control and the development of obesity), the focus was placed on nuclei closely related to energy metabolism and appetite regulation. Ultimately, the arctic nucleus, ventromedial nucleus, and paraventricular nucleus were selected as the primary focus areas, and differential metabolite maps of these three nuclei were constructed for further in-depth analysis.
[0065] (2) Principal component analysis (PCA) was used to assess the trend of metabolomic profile separation between groups.
[0066] To assess the impact of obesity on the overall metabolic profile of three different hypothalamic nuclei—the arcuate nucleus, ventromedial nucleus, and paraventricular nucleus—principal component analysis was performed on the metabolite signals detected in these regions. The results are as follows: Figure 3 As shown.
[0067] Principal component analysis showed that the normal and obese groups exhibited significant intergroup separation in all three hypothalamic nuclei, suggesting a significant trend of remodeling of the overall metabolic profile of different hypothalamic nuclei associated with obesity. Figure 3 As shown in (a), in the arcuate nucleus, the first principal component (PC1) and the second principal component (PC2) explained 34.7% and 25.2% of the total variance, respectively, with a cumulative explanation rate of 59.9%. The PCA score plot showed that the normal group samples were mainly distributed in the positive PC2 region, while the obese group samples were mainly located in the negative PC2 region. The two groups were clearly separated overall, with only a small overlap in distribution, indicating that the composition and relative abundance of metabolites in the arcuate nucleus under obese conditions underwent a significant overall change.
[0068] like Figure 3 As shown in (b), in the ventromedial nucleus, PC1 and PC2 explained 69.7% and 6.55% of the total variance, respectively, with a cumulative explained rate of 76.25%. The two groups of samples mainly separated along the PC1 direction: the normal group samples were concentrated in the positive PC1 region, while the obese group samples were mainly distributed in the negative PC1 region, with relatively clear inter-group cluster boundaries. Compared to the arcuate nucleus, the metabolic characteristics of the normal and obese groups differed more significantly in the ventromedial nucleus, suggesting that obesity may be accompanied by more extensive metabolic remodeling in the ventromedial nucleus.
[0069] like Figure 3 As shown in (c), in the paraventricular nucleus, PC1 and PC2 explained 62.5% and 6.06% of the total variance, respectively, with a cumulative explained rate of 68.56%. The normal group samples were generally biased towards the negative PC1 region, while the obese group samples were mainly distributed in the positive PC1 region, showing a clear spatial separation trend between the two groups. However, some overlap in the distribution could still be observed in the region where PC1 was close to zero. This indicates that obesity is also accompanied by changes in the overall metabolic pattern of the paraventricular nucleus, but there may be some metabolic heterogeneity among different individuals.
[0070] In summary, the PCA results showed that the obese and normal groups exhibited varying degrees of metabolic profile separation in the arcuate nucleus, ventromedial nucleus, and paraventricular nucleus, with the ventromedial nucleus showing particularly prominent intergroup separation. These results validate the application value of the method of this invention in accurately distinguishing the metabolic characteristics of different hypothalamic nuclei and revealing nucleus-specific metabolic remodeling.
[0071] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention. Therefore, all technical solutions obtained through equivalent substitution or transformation fall within the protection scope of the present invention.
Claims
1. A method for locating and classifying hypothalamic nuclei using spatial metabolomics mass spectrometry imaging, characterized in that, The specific steps are as follows: S1: Obtain isolated brain tissue containing the hypothalamus and its surrounding complete anatomical structures, and perform serial slicing on the brain tissue sample to obtain adjacent slices as histological slices and mass spectrometry imaging slices, respectively. S2: Perform histological staining and digital imaging on the histological sections to obtain histological images; S3: Based on the slice direction, stereoscopic coordinates, and anatomical landmarks of the histological image, select a reference atlas slice image that matches the histological image from the corresponding animal model brain stereoscopic atlas; S4: Spatial registration is performed between the reference atlas section image and the histological image. The boundary of the hypothalamic nucleus in the reference atlas section image is mapped onto the histological image. The boundary of the nucleus is corrected according to the histological structure shown in the histological image to obtain a superimposed image containing the histological structure and the boundary of the nucleus. Based on the boundaries of the nuclei in the overlaid images, one or more target hypothalamic nuclei are identified to form a region of interest. S5: Perform mass spectrometry imaging scans on the corresponding regions of interest in the mass spectrometry imaging slices to obtain mass spectrometry imaging data; convert the mass spectrometry imaging data into a standardized format into a reference ion map reflecting the spatial contour of the tissue in the mass spectrometry imaging slices using mass spectrometry imaging software. S6: Overlay and spatially register the superimposed image obtained in step S4 and the reference ion map obtained in step S5 to match the tissue contours in the two images and establish the spatial coordinate correspondence between the histological image and the mass spectrometry imaging data. S7: Based on the spatial coordinate correspondence, obtain the mass spectrometry imaging data corresponding to the target hypothalamic nucleus in the superimposed image, and finally obtain the signal of the target metabolite in the different target hypothalamic nuclei.
2. The method for locating and dividing hypothalamic nuclei using spatial metabolomics mass spectrometry imaging according to claim 1, characterized in that, The brain tissue is rat brain tissue or mouse brain tissue, and a stereotactic brain atlas matching the animal model is selected.
3. The method for locating and dividing hypothalamic nuclei using spatial metabolomics mass spectrometry imaging according to claim 1, characterized in that, In step S1, the thickness of the continuous slices is controlled at 8~12 μm.
4. The method for locating and dividing hypothalamic nuclei using spatial metabolomics mass spectrometry imaging according to claim 1, characterized in that, In step S1, the histological sections are flattened and attached to a glass slide, and placed in a 37°C constant temperature oven for later use; the mass spectrometry imaging sections are flattened and attached to a positive charge desorption plate, and stored in a -80°C freezer for later use.
5. The method for locating and dividing hypothalamic nuclei using spatial metabolomics mass spectrometry imaging according to claim 1, characterized in that, The histological sections in step S2 are stained using hematoxylin-eosin staining; the digital imaging is performed using a digital slide scanner.
6. The method for locating and dividing hypothalamic nuclei using spatial metabolomics mass spectrometry imaging according to claim 1, characterized in that, The anatomical landmarks mentioned in step S3 are selected from at least three of the following: optic chiasm, fornix, mammillary body, pituitary stalk, hippocampus, and outer edge of brain tissue.
7. The method for locating and dividing hypothalamic nuclei using spatial metabolomics mass spectrometry imaging according to claim 1, characterized in that, In step S4, spatial registration is performed using the image processing software ImageJ, specifically as follows: the histological image is set as the target image, and the reference atlas section image is set as the moving image; several corresponding anatomical landmark structures are marked on the histological image and the reference atlas section image respectively; the reference atlas section image is spatially transformed according to the marked anatomical landmark structures, so that the external shape and internal anatomical structure of the brain tissue in the reference atlas section image match the corresponding tissue structure in the histological image, and the boundary of the hypothalamic nucleus in the reference atlas section image is mapped onto the histological image.
8. The method for locating and classifying hypothalamic nuclei using spatial metabolomics mass spectrometry imaging according to claim 1, characterized in that, The standardized format mentioned in step S5 is the imzML format; the mass spectrometry imaging software uses MSiReader software, which displays the two-dimensional spatial coordinates of each scanned pixel and its corresponding ion intensity information.
9. The method for locating and dividing hypothalamic nuclei using spatial metabolomics mass spectrometry imaging according to claim 1, characterized in that, The target hypothalamic nuclei include one or more of the arcuate nucleus, paraventricular nucleus, ventromedial nucleus, dorsomedial nucleus, and posterior nucleus.
10. An application of the hypothalamic nucleus localization and segmentation method using spatial metabolomics mass spectrometry imaging as described in any one of claims 1 to 9, characterized in that, The signals of metabolites in the target hypothalamic nucleus in brain tissue samples under different experimental conditions were obtained, and the metabolic characteristics of the target hypothalamic nucleus under different experimental conditions were analyzed based on the metabolite signals.
