Human body brown adipose tissue characterization method and device and metabolic function evaluation system
By combining multi-wavelength photoacoustic imaging technology with ultrasound image segmentation and light flux correction, the problems of radiation damage and imaging accuracy in the imaging assessment of human adipose tissue in existing technologies have been solved. Real-time, non-invasive, and quantitative assessment of human brown adipose tissue has been achieved, and an automated image processing and result presentation system has been established to meet the needs of lipid metabolism assessment of metabolic syndrome and insulin resistance efficacy detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- PEKING UNION MEDICAL COLLEGE HOSPITAL
- Filing Date
- 2026-01-21
- Publication Date
- 2026-05-15
AI Technical Summary
Current PET-CT technology suffers from high image composition, ionizing radiation damage, and inability to meet the short-term follow-up assessment needs of drug treatment research in human adipose tissue imaging evaluation. Photoacoustic imaging needs improvement in spatial resolution, depth, and positioning accuracy, and lacks standardized methods and automated evaluation systems.
By employing multi-wavelength photoacoustic imaging technology, combined with ultrasound image segmentation and light flux correction, light flux compensation and linear spectral mixing are performed to establish a real-time, non-invasive, and quantitative characterization method for human adipose tissue. An automated image processing and result presentation system is also developed, including data acquisition, preprocessing, core processing, and post-processing modules.
It enables real-time, non-invasive, and quantitative assessment of human brown adipose tissue, improving imaging accuracy and clinical applicability. It also establishes an automated image processing and result presentation system, meeting the needs of lipid metabolism assessment for metabolic syndrome and insulin resistance efficacy detection.
Smart Images
Figure CN122030879A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical image processing technology, and in particular to a method for characterizing human brown adipose tissue, a device for characterizing human brown adipose tissue, and a metabolic function assessment system, which are mainly used for lipid metabolism assessment of metabolic syndrome and detection of insulin resistance efficacy. Background Technology
[0002] PET-CT (Positron Emission Tomography-Computed Tomography) is currently the most commonly used imaging method for evaluating human adipose tissue. It can locate, characterize, and quantify activated BAT (Brown Adipose Tissue) in the human body. However, this technology has limitations such as high imaging composition, ionizing radiation damage, and reliance on radionuclide imaging, which cannot meet the clinical application needs of short-term follow-up evaluation in drug treatment research.
[0003] Photoacoustic imaging (PAI) is a hybrid imaging technology that combines high optical absorption contrast with deep ultrasound penetration depth. Through multi-wavelength photoacoustic imaging, it can perform contrast-free, specific photoexcitation imaging based on the differences in light absorption coefficients of different components within the tissue. By extracting multi-scale imaging parameters that reflect the characteristics of tissue structure, function, and metabolism, it can achieve the identification and imaging of diseased tissues, making it one of the technologies with the greatest scientific research value and clinical application prospects in the field of medical imaging.
[0004] Currently, the spatial resolution, depth, and positioning accuracy of photoacoustic imaging for adipose tissue characterization need improvement. Human application research is still in the exploratory stage, lacking standardized methods, multi-parameter image databases, and characterization parameters. Furthermore, there is a lack of automated evaluation systems in image processing methods and result presentation modes. Summary of the Invention
[0005] To overcome the shortcomings of the prior art, the technical problem to be solved by the present invention is to provide a method for characterizing human brown adipose tissue, which can establish a real-time, non-invasive, and quantitative photoacoustic imaging characterization technique for human adipose tissue.
[0006] The technical solution of this invention is: a method for characterizing human brown adipose tissue, which includes the following steps: (1) Collect multi-wavelength photoacoustic and ultrasound images of brown fat-rich areas in the target subjects before and after cold stimulation; (2) Perform optical flux correction on multi-wavelength photoacoustic images at different wavelengths, segment different component tissue structures through ultrasound images, and then perform optical flux compensation; (3) Linear spectral mixing was performed on the multi-wavelength photoacoustic image after light flux compensation to obtain the distribution images of different tissue components, including: oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, white fat, and brown fat. (4) Analyze the trends, standard deviations, and ratios of photoacoustic signal intensity in brown adipose tissue enrichment areas before and after cold stimulation in healthy subjects and patients, and draw statistical evaluation images; analyze the concentration changes of other tissue components besides brown adipose tissue before and after cold stimulation in healthy subjects and patients, and draw statistical evaluation images.
[0007] This invention establishes a real-time, non-invasive, and quantitative photoacoustic imaging characterization technique for human adipose tissue, and develops an automated image processing method and a result presentation evaluation system to achieve automated processing and visual comparison of multi-wavelength photoacoustic / ultrasound images before and after activation of brown adipose tissue, thereby improving clinical applicability and image processing efficiency.
[0008] It also provides a device for characterizing human brown adipose tissue, which includes: The data acquisition module is configured to acquire multi-wavelength photoacoustic and ultrasound images of the target subject in the region of interest before and after cold stimulation. The preprocessing module is configured to perform luminous flux correction on multi-wavelength photoacoustic images at different wavelengths, segment different tissue structures through ultrasound images, and then perform luminous flux compensation. The core processing module is configured to perform spectral mixing on the multi-wavelength photoacoustic image after light flux compensation to obtain the distribution images of different tissue components, including: oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, white fat, and brown fat. The post-processing module is configured to calculate the photoacoustic signal characteristic parameters of brown fat and other tissue component concentration changes based on the processed image, and output the characterization evaluation results.
[0009] It also provides a metabolic function assessment system, which includes a human brown adipose tissue characterization device and a result visualization interface (GUI). The GUI constructs an interactive operating platform that integrates automatic data reading, multi-process sequential processing, visualization analysis, and report generation. The GUI adopts a three-column logical structure of input, display, and information: The first column is the processing control module, which integrates file input, threshold adjustment, and depth setting input functions, and has file processing progress prompts and file processing completion prompts; the second column is the analysis result visualization module, which integrates the analysis results display of oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, fat content, total blood volume, and fat type; the third column is the unmixing result statistics module, which quantitatively analyzes the mean and maximum values of oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, fat content, total blood volume, fat type, and multiple parameter indicators. Attached Figure Description
[0010] Figure 1 This is a schematic diagram of the working principle of the human brown adipose tissue characterization device according to the present invention.
[0011] Figure 2 This is the result visualization interface according to the present invention. Detailed Implementation
[0012] like Figure 1 As shown, this method for characterizing human brown adipose tissue includes the following steps: (1) Collect multi-wavelength photoacoustic and ultrasound images of brown fat-rich areas in the target subjects before and after cold stimulation; (2) Perform optical flux correction on multi-wavelength photoacoustic images at different wavelengths, segment different component tissue structures through ultrasound images, and then perform optical flux compensation; (3) Linear spectral mixing was performed on the multi-wavelength photoacoustic image after light flux compensation to obtain the distribution images of different tissue components, including: oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, white fat, and brown fat. (4) Analyze the trends, standard deviations, and ratios of photoacoustic signal intensity in brown adipose tissue enrichment areas before and after cold stimulation in healthy subjects and patients, and draw statistical evaluation images; analyze the concentration changes of other tissue components besides brown adipose tissue before and after cold stimulation in healthy subjects and patients, and draw statistical evaluation images.
[0013] This invention establishes a real-time, non-invasive, and quantitative photoacoustic imaging characterization technique for human adipose tissue, and develops an automated image processing method and a result presentation evaluation system to achieve automated processing and visual comparison of multi-wavelength photoacoustic / ultrasound images before and after activation of brown adipose tissue, thereby improving clinical applicability and image processing efficiency.
[0014] Preferably, in step (1), the brown fat-rich area is selected from at least one of the supraclavicular fossa, neck, axilla, mediastinum, paravertebral region, or perirenal region.
[0015] Preferably, in step (1), the cold stimulation time is 10 to 120 minutes; the excitation wavelength of the multi-wavelength photoacoustic image is located in the near-infrared band and is multiple discrete wavelengths in the range of 600 nm to 1100 nm.
[0016] Preferably, step (2) includes the following sub-steps: (2.1) Based on the energy values corresponding to the multi-wavelength photoacoustic images recorded by the energy meter, for each frame of the original photoacoustic image, divide all its pixel values by the energy value of the pulse to normalize the photoacoustic images acquired at each wavelength. (2.2) The method of segmenting different tissue structures in the region of interest using the ultrasound image includes: using a deep learning-based image segmentation model or image processing algorithm to segment the imaging region into skin, subcutaneous fat, muscle and other tissues. The deep learning-based image segmentation model is the Attention U-Net model. The absorption, scattering coefficients and anisotropy factors of each tissue at each wavelength are input, and the effective attenuation coefficient of each tissue is calculated. (2.3) Create a lookup table based on literature data and assign an effective attenuation coefficient value according to each tissue type and wavelength.
[0017] Preferably, step (3) includes the following sub-steps: (3.1) Establish a geometric model containing multiple layers of tissue based on the segmented tissue mask, and set the optical parameters for each layer; (3.2) The distribution of light energy within the tissue is obtained by calculation using a light transmission model, which includes a Monte Carlo simulation or a diffusion approximation model. For example, by running a Monte Carlo simulation, the distribution of light energy in the three-dimensional space within the tissue under specified wavelength and light source illumination conditions is obtained. For thicker tissues, a diffusion approximation model is used as a fast solver for light transmission. By solving the diffusion equation, a three-dimensional distribution map of light flux similar to that of the Monte Carlo simulation is calculated. The photoacoustic image after light flux compensation is obtained by dividing the original photoacoustic image by the light flux distribution image. (3.3) Preset the known absorption spectrum vector of key chromophores in a specified wavelength range. Key chromophores include oxyhemoglobin HbO2, deoxyhemoglobin Hb, lipids, and water. For each pixel in the image, a series of photoacoustic signal values measured at different wavelengths constitute a measurement spectrum vector. (3.4) Solve the following linear equations: , in It is a known absorption coefficient. Given the concentration to be determined, algorithms such as the least squares method are used to optimally calculate the relative concentration value of each component. , Blood oxygen saturation is obtained using the following formula: , This yields distribution images of different tissue components.
[0018] Preferably, in step (3), the distribution images of different tissue components are superimposed on the ultrasound grayscale image for display.
[0019] Preferably, in step (4), statistical analysis is performed on the photoacoustic signal data of subjects with different physiological characteristics before and after cold stimulation to obtain the characteristics of photoacoustic signal intensity changes, dispersion, and ratio information; based on the statistical analysis results, a brown adipose tissue function assessment model or reference database is constructed. Specifically, the trends, standard deviations, and ratios of photoacoustic signal intensity changes in brown adipose tissue regions at different wavelengths before and after cold stimulation are analyzed in healthy young subjects aged 20-24, healthy middle-aged subjects aged 40-65, and patients with metabolic syndrome, and a line graph with standard deviation is plotted; the concentration changes of other tissue components before and after cold stimulation in healthy subjects and patients are analyzed, and a box plot assessment image is plotted.
[0020] Preferably, the method further includes step (5), visualizing the changes in various parameters of the participants before and after cold stimulation through the graphical user interface (GUI) of Matlab software, and completing the above automated processing from data input to output through GUI operation; outputting a quantitative analysis report.
[0021] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium. When executed, the program includes the steps of the methods of the above embodiments. The storage medium can be ROM / RAM, magnetic disk, optical disk, memory card, etc. Therefore, corresponding to the method of the present invention, the present invention also includes a human brown adipose tissue characterization device, which is typically represented in the form of functional modules corresponding to the steps of the method. Figure 1 As shown, the device includes: The data acquisition module is configured to acquire multi-wavelength photoacoustic and ultrasound images of the target subject in the region of interest before and after cold stimulation. The preprocessing module is configured to perform luminous flux correction on multi-wavelength photoacoustic images at different wavelengths, segment different tissue structures through ultrasound images, and then perform luminous flux compensation. The core processing module is configured to perform spectral mixing on the multi-wavelength photoacoustic image after light flux compensation to obtain the distribution images of different tissue components, including: oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, white fat, and brown fat. The post-processing module is configured to calculate the photoacoustic signal characteristic parameters of brown fat and other tissue component concentration changes based on the processed image, and output the characterization evaluation results.
[0022] like Figure 2 As shown, a metabolic function assessment system is also provided, which includes a human brown adipose tissue characterization device and a result visualization interface (GUI). The GUI constructs an interactive operating platform that integrates automatic data reading, multi-process sequential processing, visualization analysis, and report generation. The GUI adopts a three-column logical structure of input, display, and information: The first column is the processing control module, which integrates file input, threshold adjustment, and depth setting input functions, and has file processing progress prompts and file processing completion prompts; the second column is the analysis result visualization module, which integrates the analysis results display of oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, fat content, total blood volume, and fat type; the third column is the unmixing result statistics module, which quantitatively analyzes the mean and maximum values of oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, fat content, total blood volume, fat type, and multiple parameter indicators.
[0023] The following is a specific embodiment.
[0024] (1) Healthy subjects and patients with metabolic syndrome were recruited at Peking Union Medical College Hospital. Multi-wavelength photoacoustic / ultrasound images were collected from the participants 30-60 mins before and after cold stimulation. (2) The automatic image analysis and evaluation system of the present invention automatically performs light flux correction and compensation on the multi-wavelength photoacoustic images and performs spectral mixing to obtain the content and distribution information of various tissue components. (3) The multi-wavelength photoacoustic images and content information of various tissue components of all healthy subjects and patients are automatically analyzed to draw statistical images and evaluate the brown adipose tissue function level of healthy subjects and patients.
[0025] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.
Claims
1. A method for characterizing human brown adipose tissue, characterized in that: It includes the following steps: (1) Collect multi-wavelength photoacoustic and ultrasound images of brown fat-rich areas in the target subjects before and after cold stimulation; (2) Perform optical flux correction on multi-wavelength photoacoustic images at different wavelengths, segment different component tissue structures through ultrasound images, and then perform optical flux compensation; (3) Linear spectral mixing was performed on the multi-wavelength photoacoustic image after light flux compensation to obtain the distribution images of different tissue components, including: oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, white fat, and brown fat. (4) Analyze the trends, standard deviations, and ratios of photoacoustic signal intensity in brown adipose tissue enrichment areas before and after cold stimulation in healthy subjects and patients, and draw statistical evaluation images; analyze the concentration changes of other tissue components besides brown adipose tissue before and after cold stimulation in healthy subjects and patients, and draw statistical evaluation images.
2. The method for characterizing human brown adipose tissue according to claim 1, characterized in that: In step (1), the brown fat-rich area is selected from at least one of the supraclavicular fossa, neck, axilla, mediastinum, paravertebral region, or perirenal region.
3. The method for characterizing human brown adipose tissue according to claim 2, characterized in that: In step (1), the cold stimulation time is 10 to 120 minutes; the excitation wavelength of the multi-wavelength photoacoustic image is located in the near-infrared band and consists of multiple discrete wavelengths in the range of 600 nm to 1100 nm.
4. The method for characterizing human brown adipose tissue according to claim 3, characterized in that: Step (2) includes the following sub-steps: (2.1) Based on the energy values corresponding to the multi-wavelength photoacoustic images recorded by the energy meter, for each frame of the original photoacoustic image, divide all its pixel values by the energy value of the pulse to normalize the photoacoustic images acquired at each wavelength. (2.2) The method of segmenting different tissue structures in the region of interest using the ultrasound image includes: using a deep learning-based image segmentation model or image processing algorithm to segment the imaging region into skin, subcutaneous fat, muscle and other tissues. The deep learning-based image segmentation model is the Attention U-Net model. The absorption, scattering coefficients and anisotropy factors of each tissue at each wavelength are input, and the effective attenuation coefficient of each tissue is calculated. (2.3) Create a lookup table based on literature data and assign an effective attenuation coefficient value according to each tissue type and wavelength.
5. The method for characterizing human brown adipose tissue according to claim 4, characterized in that: Step (3) includes the following sub-steps: (3.1) Establish a geometric model containing multiple layers of tissue based on the segmented tissue mask, and set the optical parameters for each layer; (3.2) The distribution of light energy within the tissue is obtained by calculation using a light transport model, which includes a Monte Carlo simulation or a diffusion approximation model; (3.3) Preset the known absorption spectrum vector of key chromophores in a specified wavelength range. Key chromophores include oxyhemoglobin HbO2, deoxyhemoglobin Hb, lipids, and water. For each pixel in the image, a series of photoacoustic signal values measured at different wavelengths constitute a measurement spectrum vector. (3.4) Solve the following linear equations: ， in It is a known absorption coefficient. Given the concentration to be determined, algorithms such as the least squares method are used to optimally calculate the relative concentration value of each component. , Blood oxygen saturation is obtained using the following formula: ， This yields distribution images of different tissue components.
6. The method for characterizing human brown adipose tissue according to claim 5, characterized in that: In step (3), the distribution images of different tissue components are superimposed on the ultrasound grayscale image for display.
7. The method for characterizing human brown adipose tissue according to claim 6, characterized in that: In step (4), the photoacoustic signal data of subjects with different physiological characteristics before and after cold stimulation are statistically analyzed to obtain the characteristics of photoacoustic signal intensity change, dispersion and ratio information; based on the statistical analysis results, a brown fat function assessment model or reference database is constructed.
8. The method for characterizing human brown adipose tissue according to claim 7, characterized in that: The method also includes step (5), which uses the graphical user interface (GUI) of Matlab software to visualize the changes of various parameters of the participants before and after cold stimulation, and completes the above automated processing from data input to output through GUI operation; Output a quantitative analysis report.
9. A device for characterizing human brown adipose tissue, characterized in that: It includes: The data acquisition module is configured to acquire multi-wavelength photoacoustic and ultrasound images of the target subject in the region of interest before and after cold stimulation. The preprocessing module is configured to perform luminous flux correction on multi-wavelength photoacoustic images at different wavelengths, segment different tissue structures through ultrasound images, and then perform luminous flux compensation. The core processing module is configured to perform spectral mixing on the multi-wavelength photoacoustic image after light flux compensation to obtain the distribution images of different tissue components, including: oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, white fat, and brown fat. The post-processing module is configured to calculate the photoacoustic signal characteristic parameters of brown fat and other tissue component concentration changes based on the processed image, and output the characterization evaluation results.
10. A metabolic function assessment system, characterized in that: It includes a human brown adipose tissue characterization device and a result visualization interface (GUI); The GUI is an interactive operation platform that integrates automatic data reading, multi-process sequential processing, visual analysis and report generation. The GUI adopts a three-column logical structure of input, display and information: the first column is the processing control module, which integrates the input functions of file input, threshold adjustment and depth setting, and has file processing progress prompts and file processing completion prompts. The second section is the analysis results visualization module, which integrates the analysis results display of oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, fat content, total blood volume, and fat type. The third column is the unmixing results statistics module, which quantitatively analyzes the mean and maximum values of oxyhemoglobin, deoxyhemoglobin, blood oxygen saturation, fat content, total blood volume, fat type, and multiple parameter indicators.