Bone density calibration method, apparatus, medical device, and computer program product

Abstract

The application is suitable for the medical technical field, and provides a bone density calibration method and device, medical equipment and a computer program product. The method comprises the following steps: performing a scan on a plurality of calibration phantoms with known concentrations to obtain an attenuation reconstruction image; based on an adaptive multi-threshold segmentation algorithm, automatically identifying and segmenting an interested region corresponding to each calibration phantom in the attenuation reconstruction image; calculating the average value of the gray value of each interested region to obtain a measured attenuation coefficient corresponding to each calibration phantom; based on the known concentration of each calibration phantom and the corresponding measured attenuation coefficient, establishing a calibration model through linear regression; and based on the calibration model and the obtained image gray mean value corresponding to the examinee, determining the bone density of the examinee. The application performs a scan on a plurality of different calibration phantoms, and then determines the bone density in combination with the adaptive multi-threshold segmentation algorithm, thereby not only solving the problem of multiple scans required by single-threshold segmentation, but also reducing the difference in manual calibration.

Description

Technical Field

[0001] This application belongs to the field of data processing technology, and in particular relates to a bone density calibration method, apparatus, medical device and computer program product. Background Technology

[0002] In practical applications, micro-computed tomography (MicroCT) is a high-resolution three-dimensional imaging technology based on X-ray imaging, capable of accurately quantifying parameters such as bone mineral density and bone microstructure. MicroCT can provide non-destructive, high-precision data support for research and clinical evaluation of bone diseases through high-precision scanning and three-dimensional reconstruction. Currently, researchers can use MicroCT to measure bone mineral density.

[0003] However, existing bone mineral density measurement methods typically involve scanning a specific phantom and then manually selecting the target region and performing numerical calculations based on the phantom's location in the corresponding attenuated reconstructed image; or scanning a specific phantom sequentially and then segmenting the target region using a single threshold. Clearly, existing technologies are not comprehensive enough and fail to meet practical needs. Summary of the Invention

[0004] This application provides a bone mineral density calibration method, apparatus, medical device, and computer program product to address the problem that existing technologies are not comprehensive enough and cannot meet practical needs.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, embodiments of this application provide a method for bone mineral density calibration, including: S1: Perform a single scan on multiple calibration phantoms with known concentrations to obtain attenuated reconstructed images; S2: Based on the adaptive multi-threshold segmentation algorithm, the region of interest corresponding to each calibrated phantom is automatically identified and segmented in the attenuated reconstructed image; S3: Calculate the average gray value of each region of interest to obtain the measurement attenuation coefficient corresponding to each calibration phantom; S4: Based on the known concentrations and corresponding measurement attenuation coefficients of each calibration phantom, a calibration model between concentration and measurement attenuation coefficient is established through linear regression: BMD = k μ+b; where k is the regression coefficient and b is the constant term; S5: Determine the bone density of the examinee based on the calibration model and the obtained mean grayscale value of the corresponding image of the examinee.

[0007] This application provides a bone mineral density (BMD) calibration method. It involves scanning multiple calibration phantoms with known concentrations to obtain attenuation reconstruction images. Based on an adaptive multi-threshold segmentation algorithm, the method automatically identifies and segments the regions of interest (ROIs) corresponding to each calibration phantom in the attenuation reconstruction images. The method calculates the average grayscale value of each ROI to obtain the measurement attenuation coefficient for each calibration phantom. Based on the known concentrations and corresponding measurement attenuation coefficients of each calibration phantom, a calibration model between concentration and measurement attenuation coefficient is established through linear regression: BMD = k μ+b; where k is the regression coefficient and b is a constant term; based on the calibration model and the obtained mean grayscale value of the corresponding image of the examinee, the bone density of the examinee is determined. This application can obtain the attenuated reconstruction image of all phantoms with only one scan for multiple calibration phantoms, which not only solves the problem of multiple scans required for single threshold segmentation, but also avoids systematic errors such as equipment drift and positional differences caused by multiple scans, ensuring that all calibration data are on the same scanning reference, providing more stable basic data for subsequent model fitting. Subsequently, the adaptive multi-threshold segmentation algorithm can automatically adapt to the grayscale distribution differences of different calibration phantoms, accurately segment the region of interest (ROI) corresponding to each phantom, effectively avoiding the undersegmentation and oversegmentation problems that are prone to occur in mixed images of multiple phantoms by traditional fixed threshold segmentation, and reducing soft tissue or noise interference. Subsequently, a calibration model was constructed based on the measured attenuation coefficients of multiple calibration phantoms and a set bone mineral density. The mapping relationship between grayscale and bone mineral density was quantified through regression coefficients and constant terms. Compared to the conversion method using a single calibration reference, the fitting results are more statistically reliable and reduce human calibration differences, thus significantly reducing the impact of individual calibration errors on the final results. This improves the overall accuracy of subsequent actual bone mineral density calculations.

[0008] In one possible implementation of the first aspect, based on an adaptive multi-threshold segmentation algorithm, the region of interest corresponding to each calibrated phantom is automatically identified and segmented in the attenuated reconstructed image, including: S21: Divide each pixel in the attenuated reconstructed image into different gray-level intervals based on the gray-level value, and determine the number of pixels in each gray-level interval after division. S22: Determine the initial interval corresponding to each calibration phantom based on the number of pixels in each grayscale interval and the number of phantoms in each calibration phantom. S23: Based on the boundaries and interval step size of each initial interval, determine the target interval corresponding to each calibration phantom; S24: Based on each target interval, determine the region of interest corresponding to each calibration phantom.

[0009] In the above implementation, the image pixels are divided into attenuation reconstruction points by gray-level intervals, and the number of pixels in each interval is counted. This discretizes continuous gray-level values ​​into quantifiable interval units, clearly distinguishing the gray-level distribution characteristics caused by density differences among different calibrated phantoms. This avoids the ambiguity of directly thresholding continuous gray-level values, providing data support for subsequent target interval localization and improving the objectivity of the segmentation results. Subsequently, an initial interval is determined based on the number of pixels in each gray-level interval and the number of calibrated phantoms. This automatically locks the core range of the gray-level interval corresponding to each phantom, eliminating the need for manual pre-setting of phantom gray-level positions. This solves the problem of difficulty in quickly identifying target regions in multi-phantom mixed images, significantly improving the automation level of the segmentation algorithm. Then, the target interval is determined using the initial interval boundaries and interval step size. This allows for fine-grained expansion or contraction of the initial gray-level interval, accurately matching the actual gray-level distribution range of each calibrated phantom. This effectively avoids under-segmentation or over-segmentation problems caused by gradual gray-level changes at the phantom edges, ensuring that the region of interest determined based on the target interval highly matches the actual contour of the phantom, providing accurate regional data for subsequent attenuation coefficient calculation.

[0010] In one embodiment of the first aspect, the number of grayscale intervals is determined as follows: S01: Determine the maximum gray value, minimum gray value, and total number of pixels from the attenuated reconstructed image; S02: Calculate the difference between the maximum and minimum grayscale values; S03: Determine the number of intervals based on the difference and the total number of pixels.

[0011] In the above implementation, the number of intervals is determined based on the difference between the maximum and minimum grayscale values ​​of the attenuated reconstructed image and the total number of pixels. This allows for dynamic adjustment of the binning granularity according to the actual grayscale distribution range and pixel scale of the image, avoiding the drawbacks of overly coarse binning when the grayscale range is large and overly fine binning when the grayscale range is small, under a fixed number of bins. This ensures that each grayscale interval can accurately represent the grayscale feature differences of different calibrated phantoms, laying a reliable foundation for the subsequent localization of initial and target intervals. Simultaneously, the number of bins is strongly correlated with the image grayscale difference and the total number of pixels. This reduces meaningless over-binning while maintaining grayscale feature distinguishability, and also reduces the computational load of subsequent pixel statistics and interval selection, achieving a balance between segmentation accuracy and algorithm efficiency.

[0012] In one embodiment of the first aspect, determining the initial interval corresponding to each calibration phantom based on the number of pixels in each grayscale interval and the number of calibrated phantoms includes: S221: Sort the grayscale intervals in ascending order of the number of pixels to obtain the interval order; S222: Based on the interval order, the grayscale interval that is the last in the sort and equal to the number of phantoms is determined as the initial interval.

[0013] In the above embodiments, since the calibrated phantoms typically appear as continuous regions of a certain pixel size in the attenuated reconstruction image, the number of pixels in their corresponding grayscale intervals is much higher than that of invalid regions such as noise and background. Therefore, by sorting them by the number of pixels from smallest to largest, valid phantom intervals and invalid interference intervals can be quickly distinguished. Then, by selecting the interval with the lowest sorting value based on the number of calibrated phantoms as the initial interval, the core grayscale interval corresponding to each calibrated phantom can be directly locked, avoiding interference from invalid intervals in subsequent segmentation processes and laying a precise foundation for determining the target interval. Simultaneously, the entire initial interval determination process does not require manual pre-setting of phantom grayscale ranges or interval filtering rules; it relies entirely on pixel count statistics and sorting matching, eliminating errors caused by subjective human judgment.

[0014] In one embodiment of the first aspect, the boundary includes a lower boundary limit and an upper boundary limit; based on the boundary and interval step size of each initial interval, the target interval corresponding to each calibration phantom is determined, including: S231: For any calibration phantom, perform a subtraction operation on the lower boundary value of the initial interval of the calibration phantom and the interval step size to obtain the target lower boundary value; S232: Perform an addition operation on the upper boundary value and the interval step size of the initial interval of the calibration phantom to obtain the target upper boundary value; S233: Based on the target lower limit and target upper limit, determine the target interval corresponding to the calibration phantom.

[0015] In the above embodiments, since the initial interval only locks the core grayscale interval corresponding to the calibrated phantom, while the edges of the phantom often have transitional areas with gradual grayscale changes, this embodiment can accurately cover the grayscale transitional parts outside the core area of ​​the phantom by performing a symmetrical expansion operation of subtracting the lower boundary value and adding the upper boundary value of the initial interval. This avoids the under-segmentation problem caused by missing edge grayscale, ensuring that the extracted target interval highly matches the actual grayscale distribution of the calibrated phantom, and providing complete regional data support for subsequent measurement attenuation coefficient calculation. Furthermore, this determination method can complete the boundary expansion simply through addition and subtraction operations, without the need for complex iteration or optimization processes, significantly reducing the computational load of the algorithm while ensuring segmentation accuracy.

[0016] In one embodiment of the first aspect, the region of interest corresponding to each calibration phantom is determined based on each target interval, including: S241: For any calibration phantom, generate a binary image based on each target pixel in the target interval corresponding to the calibration phantom; the binary image includes multiple connected components. S242: Determine the connected region with the largest area as the initial region corresponding to the calibration phantom, and determine the center coordinates of the initial region; S243: Determine the region diameter of the initial region based on the center coordinates; S244: The circular region constructed in the initial region with the center coordinate as the center point and the target length as the target diameter is determined as the region of interest of the calibration phantom; the target length is the product of the preset ratio and the region diameter.

[0017] In the above implementation, generating a binary image based on the target region's pixels and filtering the connected components with the largest area automatically eliminates invalid regions formed by image noise, artifacts, and scattered interfering pixels, ensuring that the initial region closely matches the actual contour of the calibrated phantom. Subsequently, locating and cropping a circular region using the center coordinates further avoids the influence of irregular edges on the calculation of the measurement attenuation coefficient, ensuring that the region of interest is the core, uniform part of the phantom, providing high-purity regional data for grayscale mean calculation. Simultaneously, constructing a circular region using the center coordinates of the initial region as the center point and a preset proportion of the region's diameter as the target diameter replaces the manual selection of the region of interest, eliminating errors from subjective human judgment. Furthermore, the entire region determination process is achieved through only basic image processing steps such as connected component analysis, coordinate calculation, and circular cropping, requiring no complex iterative optimization or model training, resulting in low computational load and fast operation.

[0018] In one embodiment of the first aspect, determining the bone density of the examinee based on a calibration model and the obtained mean grayscale value of the image corresponding to the examinee includes: S51: Scan the subject to obtain scan images and determine the skeletal regions in the scan images; S52: Determine the average gray level of the bone region as the average gray level of the image; S53: Substitute the image grayscale mean into the calibration model for calculation to obtain the subject's bone density.

[0019] In the above implementation, the skeletal region is first identified from the scanned image of the subject, and then the average grayscale value of this region is extracted as the basis for calculation. This effectively avoids interference from non-skeletal regions such as soft tissue, fat, and air, ensuring that the grayscale data input into the calibration model only reflects the attenuation characteristics of bone. Compared to directly using the average grayscale value of the entire image, this significantly reduces the interference of irrelevant regions on the calculation results, improving the accuracy and reliability of bone density values. Furthermore, the regression coefficients and constants of the calibration model are obtained based on multi-calibration simulations, making the calculation logic traceable and verifiable, facilitating the establishment of a unified bone density detection standard. Therefore, after completing the construction of the calibration model, it can be directly reused to calculate bone density for different subjects without the need for repeated calibration.

[0020] Secondly, embodiments of this application provide a bone mineral density calibration device, comprising: The first scanning unit is used to perform a single scan on multiple calibration phantoms of known concentrations to obtain attenuated reconstructed images; The image segmentation unit is used to automatically identify and segment the region of interest corresponding to each calibrated phantom in the attenuated reconstructed image based on an adaptive multi-threshold segmentation algorithm. The mean calculation unit is used to calculate the average gray value of each region of interest to obtain the measurement attenuation coefficient corresponding to each calibration phantom. The model building unit is used to establish a calibration model between concentration and measurement attenuation coefficient based on the known concentration and corresponding measurement attenuation coefficient of each calibration phantom using linear regression: BMD = k μ+b; where k is the regression coefficient and b is the constant term; The bone density determination unit is used to determine the bone density of the examinee based on the calibration model and the obtained average grayscale value of the corresponding image of the examinee.

[0021] In one embodiment of the second aspect, the image segmentation unit specifically includes: A segmentation unit is used to divide each pixel in the attenuated reconstructed image into different gray-level intervals based on gray-level values, and to determine the number of pixels in each gray-level interval after segmentation. The initial interval determination unit is used to determine the initial interval corresponding to each calibration phantom based on the number of pixels in each grayscale interval and the number of phantoms in each calibration phantom. The target interval determination unit is used to determine the target interval corresponding to each calibration phantom based on the boundary and interval step size of each initial interval. The region of interest determination unit is used to determine the region of interest corresponding to each calibration phantom based on each target interval.

[0022] In one embodiment of the second aspect, the bone mineral density calibration device further includes: The grayscale value determination unit is used to determine the maximum grayscale value, the minimum grayscale value, and the total number of pixels from the attenuated reconstructed image. The difference calculation unit is used to calculate the difference between the maximum and minimum grayscale values; The interval number determination unit is used to determine the number of intervals based on the difference and the total number of pixels.

[0023] In one embodiment of the second aspect, the initial interval determination unit specifically includes: The sorting unit is used to sort the grayscale intervals in ascending order of the number of pixels to obtain the interval order; The first interval determines the sub-unit, which is used to determine the grayscale interval that is the last in the sorted order and equal to the number of phantoms as the initial interval.

[0024] In one embodiment of the second aspect, the boundary includes a lower boundary limit and an upper boundary limit; the target interval determination unit specifically includes: The subtraction unit is used to perform a subtraction operation on the lower boundary value of the initial interval and the interval step size of any calibration phantom to obtain the target lower boundary value. The addition unit is used to perform addition operations on the upper boundary value and the interval step size of the initial interval of the calibration phantom to obtain the target upper boundary value; The second interval determination sub-unit is used to determine the target interval corresponding to the calibration phantom based on the target lower limit and target upper limit.

[0025] In one embodiment of the second aspect, the region of interest determination unit specifically includes: The image generation unit is used to generate a binary image based on each target pixel in the target interval corresponding to any calibration phantom; the binary image includes multiple connected components. The coordinate determination unit is used to determine the connected region with the largest area as the initial region corresponding to the calibration phantom, and to determine the center coordinates of the initial region; The diameter determination unit is used to determine the region diameter of the initial region based on the center coordinates. The region construction unit is used to determine the region of interest of the calibration phantom as the circular region constructed in the initial region with the center coordinate as the circle and the target length as the target diameter; the target length is the product of a preset ratio and the region diameter.

[0026] In one embodiment of the second aspect, the bone density determination unit specifically includes: The second scanning unit is used to scan the subject, obtain scanned images, and determine the skeletal regions of the scanned images. The mean value determination unit is used to determine the mean gray value corresponding to the skeletal region as the mean gray value of the image; The bone density calculation unit is used to input the average grayscale value of the image into the calibration model to calculate the bone density of the examinee.

[0027] Thirdly, embodiments of this application provide a medical device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the bone mineral density calibration method as described in any one of the first aspects above.

[0028] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the bone mineral density calibration method as described in any one of the first aspects above.

[0029] Fifthly, embodiments of this application provide a computer program product that, when run on a medical device, enables the medical device to perform the bone mineral density calibration method described in any one of the first aspects.

[0030] Compared with existing technologies, the bone mineral density calibration method provided in this application has the following advantages: 1. Improved calibration efficiency and automation: Images of all calibration phantoms with different concentrations can be acquired in a single scan, and an adaptive multi-threshold algorithm is used to achieve fully automatic and high-precision segmentation of the region of interest (ROI) of each phantom, completely replacing the cumbersome process of traditional multiple scans or manual selection, and significantly improving the efficiency and automation of calibration operations.

[0031] 2. Ensures the accuracy and consistency of calibration results: The adaptive multi-threshold algorithm can dynamically determine the segmentation threshold based on the actual grayscale distribution of each phantom in the image, overcoming the shortcomings of the existing fixed single-threshold method in inaccurate segmentation of multi-phantom images. Simultaneously, the fully automated processing eliminates the subjectivity and random errors of manual operation, ensuring high repeatability and consistency in the calibration model establishment.

[0032] 3. Enhanced system practicality and reliability: This application simplifies the complex multi-phantom calibration into a single scan plus automated processing, reducing reliance on operator experience and facilitating integration into existing MicroCT system workflows. Simultaneously, the established linear calibration model is concise and computationally fast, providing a reliable technical foundation for accurate and efficient bone density measurement of large batches of samples. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This is a flowchart illustrating the implementation of a bone mineral density calibration method provided in an embodiment of this application; Figure 2 This is a flowchart illustrating the implementation of a bone mineral density calibration method provided in another embodiment of this application; Figure 3This is a flowchart illustrating the implementation of a bone mineral density calibration method provided in another embodiment of this application; Figure 4 This is a flowchart illustrating the implementation of a bone mineral density calibration method provided in another embodiment of this application; Figure 5 This is a schematic diagram of each region of interest in an attenuation-reconstructed image provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of a bone mineral density calibration device provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of a medical device provided in one embodiment of this application. Detailed Implementation

[0035] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0036] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0037] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0038] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0039] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0040] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0041] In practical applications, micro-computed tomography (MicroCT) is a high-resolution three-dimensional imaging technology based on X-ray imaging, capable of accurately quantifying parameters such as bone mineral density and bone microstructure. MicroCT can provide non-destructive, high-precision data support for research and clinical evaluation of bone diseases through high-precision scanning and three-dimensional reconstruction. Currently, researchers can use MicroCT to measure bone mineral density.

[0042] However, existing bone density measurement methods typically involve scanning a specific phantom and then manually selecting the target region and performing numerical calculations based on the phantom's location in its corresponding attenuated reconstruction image; or scanning a specific phantom sequentially and then segmenting the target region using a single threshold.

[0043] Therefore, this application provides a bone density calibration method that can perform a single scan on multiple different calibration phantoms and then combine it with an adaptive multi-threshold segmentation algorithm to determine bone density. This not only solves the problem of requiring multiple scans for single-threshold segmentation but also reduces the differences in manual calibration.

[0044] Please see Figure 1 , Figure 1 This is a flowchart illustrating the implementation of a bone mineral density (BMD) calibration method according to an embodiment of this application. In this embodiment, the BMD calibration method is performed by a medical device. The medical device can be a MicroCT scanner or other devices capable of quantifying bone mineral density.

[0045] like Figure 1 As shown, a bone mineral density calibration method provided in one embodiment of this application may include S1 to S5, which are detailed below: In S1, a single scan is performed on multiple calibration phantoms with known concentrations to obtain attenuated reconstructed images.

[0046] It should be noted that the calibration phantom can be a hydroxyapatite rod.

[0047] In practical applications, hydroxyapatite rods refer to cylindrical rod-shaped materials made of hydroxyapatite (CaHA), typically used as tissue equivalent materials for medical physics calibration and dosimetry verification.

[0048] The bone mineral density (BMD) of each calibration phantom is a known concentration, and each known concentration can be determined according to actual needs. No specific limitation is made on the known concentration here.

[0049] It should be noted that the number of calibration phantoms can be determined according to actual needs, and there is no limit here. For example, the number of calibration phantoms can be three or five.

[0050] In this embodiment, after determining multiple different calibration phantoms, the user can arrange all the calibration phantoms on the scanning stage at a fixed interval to ensure that there is no overlap or obstruction between the phantoms. The fixed interval can be determined according to actual needs and is not limited here.

[0051] Then, the medical device can perform a simultaneous scan on the multiple calibrated phantoms arranged above, generating an attenuated reconstructed image containing phantom information of multiple calibrated phantoms, that is, all calibrated phantoms.

[0052] It should be noted that the gray value of each pixel in the attenuated reconstructed image is positively correlated with the degree of X-ray attenuation at the corresponding location.

[0053] The format of the attenuated reconstructed image can be the native image format of the medical device, which contains phantom information such as pixel grayscale matrix and scanning parameters.

[0054] In S2, based on the adaptive multi-threshold segmentation algorithm, the region of interest corresponding to each calibrated phantom is automatically identified and segmented in the attenuated reconstructed image.

[0055] It should be noted that the core of the adaptive multi-threshold segmentation algorithm is to automatically determine the segmentation threshold based on the peak value of the image grayscale distribution, without the need for manual intervention.

[0056] In this embodiment of the application, the medical device can statistically analyze the grayscale values ​​of the attenuated reconstructed image and generate a grayscale histogram.

[0057] In the histogram, the horizontal axis represents the grayscale value (0~255 or a grayscale range defined by the medical device), and the vertical axis represents the number of pixels corresponding to the grayscale value.

[0058] Then, medical devices can use peak-valley detection to identify the number of peaks in the histogram.

[0059] It should be noted that the number of peaks mentioned above is equal to the number of phantoms in the calibration phantom, that is, each peak corresponds to a gray-scale concentration area of ​​a calibration phantom.

[0060] Then, the medical device can use the gray value corresponding to the valley point between two adjacent peaks as the segmentation threshold.

[0061] It should be noted that if there are N calibration phantoms, then N-1 adaptive segmentation thresholds can be generated.

[0062] In this embodiment, after determining multiple segmentation thresholds, the medical device can perform multi-threshold segmentation on the attenuated reconstructed image based on these thresholds. For example, pixels with gray values ​​in the range [T0, T1) are divided into the region of the first calibration phantom, pixels with gray values ​​in the range [T1, T2) are divided into the region of the second calibration phantom, and so on, until pixels with gray values ​​in the range [TN-1, TN) are divided into the region of the Nth calibration phantom. Wherein, (T0 is the minimum gray value of the attenuated reconstructed image, and TN is the maximum gray value in the attenuated reconstructed image).

[0063] In some possible embodiments, the medical device may also perform morphological processing on each of the segmented regions described above: such as removing noisy pixels at the edges of the regions through erosion operations, and then restoring the complete shape of the regions through dilation operations to eliminate holes generated by segmentation.

[0064] In this embodiment, after performing morphological processing on each of the segmented regions, the medical device can also extract the minimum bounding rectangle of each processed region and determine the minimum bounding rectangle of each region as the region of interest (ROI) corresponding to each calibrated phantom, so as to ensure that each ROI contains only the effective pixels of a single calibrated phantom without interference from the background or other phantoms.

[0065] In one embodiment of this application, the medical device can specifically be implemented via, as shown in the example below. Figure 2 Steps S21 to S24, which show the determination of the region of interest corresponding to each calibration phantom, are detailed below: In S21, each pixel in the attenuated reconstructed image is divided into different gray-level intervals based on the gray-level value, and the number of pixels in each gray-level interval after division is determined.

[0066] It should be noted that after obtaining the attenuated reconstructed image, the medical device can determine the number of gray-level intervals (Bin) based on the gray values ​​and the total number of pixels in the attenuated reconstructed image. Different gray-level intervals are used to classify the gray values ​​of different pixels in the attenuated reconstructed image.

[0067] Specifically, in one embodiment of this application, the medical device may determine the number of grayscale intervals according to the following steps S01~S03, as detailed below: In S01, the maximum gray value, the minimum gray value, and the total number of pixels are determined from the attenuated reconstructed image. In S02, the difference between the maximum and minimum grayscale values ​​is calculated. In S03, the number of intervals is determined based on the difference and the total number of pixels.

[0068] In this embodiment, the medical device can compare the gray values ​​of each pixel in the attenuation map one by one to determine the maximum and minimum gray values ​​in the attenuated reconstructed image.

[0069] Medical devices can also count the number of pixels in the attenuated reconstructed image in real time to obtain the total number of pixels.

[0070] The medical device can then calculate the difference between the maximum and minimum gray values ​​to determine the gray-scale distribution span of the attenuated reconstructed image.

[0071] In this embodiment, after calculating the aforementioned difference, the medical device can determine the first baseline quantity corresponding to the current difference based on the pre-stored correspondence between different differences and different baseline interval quantities. Simultaneously, the medical device can also determine the second baseline quantity corresponding to the current total number of pixels based on the pre-stored correspondence between different total pixel quantities and different baseline interval quantities.

[0072] It should be noted that the number of benchmark intervals corresponding to different differences can be determined according to actual needs, and there is no limit here.

[0073] The number of reference intervals corresponding to different total number of pixels can also be determined according to actual needs, and there is no restriction here.

[0074] In this embodiment, the medical device can perform a weighted summation of the first and second benchmark quantities to obtain the final interval quantity.

[0075] This embodiment determines the number of intervals based on the difference between the maximum and minimum grayscale values ​​of the attenuated reconstructed image and the total number of pixels. This allows for dynamic adjustment of the binning granularity according to the actual grayscale distribution range and pixel size of the image. This avoids the drawbacks of overly coarse binning when the grayscale range is large, or overly fine binning when the grayscale range is small, under a fixed number of bins. It ensures that each grayscale interval accurately represents the grayscale feature differences of different calibrated phantoms, laying a reliable foundation for the subsequent localization of initial and target intervals. Simultaneously, the number of bins is strongly correlated with the image's grayscale difference and total number of pixels. This reduces meaningless over-binning while maintaining grayscale feature distinguishability, and also reduces the computational load of subsequent pixel statistics and interval selection, achieving a balance between segmentation accuracy and algorithm efficiency.

[0076] In this embodiment, after obtaining the number of grayscale intervals, the medical device can divide the difference between the maximum and minimum grayscale values ​​by the number of intervals to obtain the grayscale value variation range of each grayscale interval. Then, based on this grayscale value variation range, the minimum grayscale value, the maximum grayscale value, and the number of intervals, the medical device can obtain the grayscale value range corresponding to each grayscale interval.

[0077] Then, the medical device can divide each pixel into different grayscale intervals based on the grayscale values ​​of each pixel in the attenuated reconstructed image and the grayscale value ranges corresponding to each grayscale interval determined above.

[0078] In this embodiment, after dividing each pixel into different grayscale intervals, the medical device can count the number of pixels in each grayscale interval after the division.

[0079] In S22, the initial interval corresponding to each calibration model is determined based on the number of pixels in each grayscale interval and the number of models in the calibration model.

[0080] In this embodiment, after the medical device counts the number of pixels in each grayscale interval, it can determine the initial interval corresponding to each calibrated phantom based on the number of pixels in each grayscale interval and the number of phantoms in the calibrated phantom.

[0081] In one embodiment of this application, the medical device may determine the initial interval corresponding to each calibration phantom according to the following steps S221~S222, as detailed below: In S221, the grayscale intervals are sorted in ascending order of the number of pixels to obtain the interval order; In S222, based on the interval order, the grayscale interval that is the last in the sort and equal to the number of phantoms is determined as the initial interval.

[0082] In this embodiment, the medical device can sort the grayscale intervals according to the number of pixels in ascending order, thereby obtaining the interval order. The interval order includes the sorting of each grayscale interval.

[0083] Since the calibration phantom usually appears as a continuous region with a certain pixel size in the attenuation reconstruction image, the number of pixels in its corresponding grayscale interval is much higher than that of invalid regions such as noise and background. Therefore, in this embodiment, the medical device can determine the grayscale interval that is the last in the sequence and has the same number of pixels as the phantom as the initial interval, that is, the grayscale interval with the most pixels and the same number of pixels as the phantom.

[0084] For example, assuming there are three phantoms, the medical device can determine the three grayscale intervals that are last in the interval sequence as the initial intervals.

[0085] Since the calibrated phantom typically appears as a continuous region of a certain pixel size in the attenuated reconstruction image, the number of pixels in its corresponding grayscale interval is much higher than that of invalid regions such as noise and background. Therefore, this embodiment can quickly distinguish between valid phantom intervals and invalid interference intervals by sorting them by the number of pixels from smallest to largest. Then, by selecting the last few grayscale intervals in the sorted sequence based on the number of calibrated phantoms, multiple core grayscale intervals equal to the number of phantoms can be directly locked, avoiding interference from invalid intervals in subsequent segmentation processes and laying a precise foundation for determining the target interval. Furthermore, the entire initial interval determination process does not require manual pre-setting of phantom grayscale ranges or interval filtering rules; it relies entirely on pixel count statistics and sorting matching, eliminating errors caused by subjective human judgment.

[0086] In S23, the target interval corresponding to each calibration phantom is determined based on the boundary and interval step size of each initial interval.

[0087] In this embodiment, since the initial interval may only cover the core area of ​​the phantom pixels, the boundary needs to be expanded to include all effective pixels. Therefore, the medical device can expand the gray value range corresponding to each initial interval based on the boundary and interval step size of each initial interval to obtain the target interval corresponding to each calibrated phantom.

[0088] It should be noted that the above boundaries are used to describe the grayscale value range of the initial interval. Therefore, the boundaries can include a lower boundary value and an upper boundary value.

[0089] The interval step size is used to describe the size of the numerical range covered by each interval.

[0090] In some possible embodiments, the medical device may specifically determine the interval step size as the difference between the aforementioned upper boundary value and lower boundary value. For example, assuming the upper boundary value is 1 and the lower boundary value is 0, the interval step size is: upper boundary value - lower boundary value = 1 - 0 = 1.

[0091] In one embodiment of this application, the medical device can specifically be implemented via, as shown in the example below. Figure 3 Steps S231 to S233, which are shown, determine the target intervals corresponding to each calibration phantom, as detailed below: In S231, for any calibration phantom, the lower boundary value of the initial interval of the calibration phantom and the interval step size are subtracted to obtain the target lower boundary value.

[0092] In S232, the upper limit value of the initial interval of the calibration phantom and the interval step size are added together to obtain the target upper limit value.

[0093] In S233, the target interval corresponding to the calibration phantom is determined based on the target lower limit and the target upper limit.

[0094] In this embodiment, for any calibration phantom, the medical device can perform a subtraction operation on the lower boundary value of the initial interval and the interval step size of the calibration phantom to obtain the target lower boundary value. The target lower boundary value describes the lower boundary value of the target interval corresponding to the calibration phantom.

[0095] For example, assuming the interval step size is 1 and the lower boundary limit is 2, then the target lower limit is 2-1=1.

[0096] The medical device can then perform an addition operation on the upper boundary value and the interval step size of the initial interval of the calibration phantom to obtain the target upper boundary value. The target upper boundary value describes the upper boundary value of the target interval corresponding to the calibration phantom.

[0097] For example, assuming the interval step size is 1 and the upper limit of the boundary is 3, then the target upper limit is 3+1=4.

[0098] In this embodiment, the medical device can obtain the final target range by obtaining the target lower limit and target upper limit values ​​of the target range corresponding to the above-mentioned calibrated phantom.

[0099] For example, assuming the target lower limit is 1 and the target upper limit is 4, then the gray value range of the target interval corresponding to the above-mentioned calibration phantom is [1, 4].

[0100] Combining steps S231-S233, since the initial interval only locks the core grayscale interval corresponding to the calibrated phantom, while the edges of the phantom often have transitional areas with gradual grayscale changes, this embodiment can accurately cover the grayscale transitional parts outside the core area of ​​the phantom by performing a symmetrical expansion operation of subtracting the lower boundary value and adding the upper boundary value of the initial interval. This avoids under-segmentation problems caused by missing edge grayscale values, ensuring that the extracted target interval highly matches the actual grayscale distribution of the calibrated phantom, providing complete regional data support for subsequent measurement attenuation coefficient calculation. Furthermore, this determination method can complete the boundary expansion through simple addition and subtraction operations, without the need for complex iteration or optimization processes, significantly reducing the computational load of the algorithm while ensuring segmentation accuracy.

[0101] In S24, the region of interest corresponding to each calibration phantom is determined based on each target interval.

[0102] In this embodiment, after determining the target interval corresponding to each calibrated phantom, the medical device can extract the pixel points corresponding to each target interval from the attenuated reconstruction image to obtain the pixel point set corresponding to each target interval.

[0103] Subsequently, the medical device can perform morphological processing on each set of pixels: including removing isolated noise pixels at the edges of each set of pixels through erosion; then dilating each set of pixels after erosion through dilation to restore the complete shape of its corresponding calibration phantom and eliminate holes generated by segmentation; finally, extracting the largest connected component in each set of pixels after dilation, and determining each largest connected component as the initial ROI of its corresponding calibration phantom.

[0104] In this embodiment, after obtaining each initial ROI, the medical device can also extract the minimum bounding rectangle of each initial ROI and determine the minimum bounding rectangle of each initial ROI as the region of interest corresponding to each calibrated phantom, so as to ensure that each ROI contains only the effective pixels of a single calibrated phantom without interference from the background or other phantoms.

[0105] In one embodiment of this application, the medical device can specifically be implemented via, as shown in the example below. Figure 4 Steps S241 to S244, which show how to determine the region of interest for each calibration phantom, are described in detail below: In S241, for any calibration phantom, a binary image is generated based on each target pixel in the target interval corresponding to the calibration phantom; the binary image includes multiple connected components.

[0106] In this embodiment, for any calibrated phantom, the medical device can set the pixel value of each target pixel in the target interval corresponding to the calibrated phantom to 1 in the attenuation reconstruction image, and set all other pixels except the target pixels to 0, thereby generating an initial binary image corresponding to the calibrated phantom.

[0107] Subsequently, the medical device can connect adjacent target pixels with a pixel value of 1 in the initial binary image to obtain a binary image containing multiple connected components.

[0108] In S242, the connected region with the largest area is determined as the initial region corresponding to the calibration phantom, and the center coordinates of the initial region are determined.

[0109] In this embodiment, after generating the binary image corresponding to the calibration phantom, the medical device can calculate the area of ​​each connected region in the binary image one by one, and compare each area to determine the connected region with the largest area.

[0110] Subsequently, the medical device can determine the initial region corresponding to the calibration phantom, i.e. the initial ROI, by the connected domain with the largest area, and determine the center coordinates of the initial region.

[0111] In S243, the region diameter of the initial region is determined based on the center coordinates.

[0112] In this embodiment, after determining the center coordinates of the initial region corresponding to the above-mentioned calibration phantom, the medical device can determine the region diameter of the initial region as the longest line segment that passes through the center coordinates and whose endpoints all fall on the boundary of the initial region.

[0113] In S244, the circular region constructed in the initial region with the center coordinate as the center point and the target length as the target diameter is determined as the region of interest of the calibration phantom; the target length is the product of a preset ratio and the region diameter.

[0114] In this embodiment, after obtaining the diameter of the initial region corresponding to the aforementioned calibrated phantom, the medical device can multiply this diameter by a preset ratio and determine the product as the target length. The preset ratio can be determined according to actual needs and is not limited here.

[0115] It should be noted that the preset ratio is less than 1. For example, the preset ratio could be 0.8.

[0116] In this embodiment, after determining the target length, the medical device can construct a circular region in the initial region, with the center coordinates of the initial region as the center point and the target length as the target diameter, and determine the circular region as the final region of interest of the above-mentioned calibration phantom.

[0117] Please see Figure 5 , Figure 5 This is a schematic diagram of the regions of interest in a decay-reconstructed image provided in an embodiment of this application. For example... Figure 5 As shown, the circular areas containing labels ①, ②, and ③ are the regions of interest for each calibration phantom.

[0118] Combining steps S241-S244, this embodiment generates a binary image based on the target interval pixels and filters the connected components with the largest area. This automatically eliminates invalid regions formed by image noise, artifacts, and scattered interfering pixels, ensuring that the initial region closely matches the actual contour of the calibrated phantom. Subsequently, by locating and cropping a circular region using the center coordinates, the influence of irregular parts of the phantom's edges on the calculation of the measurement attenuation coefficient can be further avoided, ensuring that the region of interest is the core uniform part of the phantom, providing high-purity regional data for grayscale mean calculation. Simultaneously, constructing a circular region with the center coordinates of the initial region as the center point and the region diameter of a preset ratio as the target diameter replaces the manual selection of the region of interest, eliminating errors from subjective human judgment. Furthermore, the entire region determination process is achieved only through basic image processing steps such as connected component analysis, coordinate calculation, and circular cropping, without the need for complex iterative optimization or model training, resulting in low computational load and fast operation.

[0119] In S3, the average gray value of each region of interest is calculated to obtain the measurement attenuation coefficient corresponding to each calibration phantom.

[0120] In this embodiment, after determining the region of interest (ROI) for each calibration phantom, the medical device can extract all grayscale values ​​within each ROI and calculate the average value of all grayscale values ​​within each ROI. Then, the medical device can determine the average grayscale value of each ROI as the measurement attenuation coefficient of its corresponding calibration phantom.

[0121] In S4, based on the known concentrations and corresponding measured attenuation coefficients of each calibration phantom, a calibration model between concentration and measured attenuation coefficient is established through linear regression: BMD = k μ+b; where k is the regression coefficient and b is the constant term.

[0122] In this embodiment of the application, since bone mineral density and the measured attenuation coefficient are linearly positively correlated, the medical device can construct a calibration model of bone mineral density and the measured attenuation coefficient through linear regression.

[0123] Specifically, medical devices can construct calibration models based on various measurement attenuation coefficients and the known concentrations of their corresponding calibration phantoms: BMD = kμ + b. Here, BMD represents bone mineral density, and μ represents the measurement attenuation coefficient.

[0124] It is understandable that, since the known concentration of each calibration phantom is used to describe the bone mineral density of each calibration phantom as a known concentration, the above BMD can also be expressed as the known concentration of a certain calibration phantom.

[0125] Then, the medical device can substitute the measured attenuation coefficients and the known concentrations of their respective calibration phantoms into the above calibration model to calculate the regression coefficients and constant terms of the calibration model, namely k and b.

[0126] In S5, the bone density of the examinee is determined based on the calibration model and the mean gray value of the image corresponding to the examinee.

[0127] In this embodiment of the application, after the medical device calculates the above-mentioned calibration model, it can substitute the average gray value of the image corresponding to the examinee into the above-mentioned calibration model to determine the bone density of the examinee.

[0128] In one embodiment of this application, the medical device may determine the bone density of the examinee according to the following steps S51-S53, as detailed below: In S51, the subject is scanned to obtain scanned images, and the skeletal regions of the scanned images are determined. In S52, the average gray value corresponding to the skeletal region is determined as the average gray value of the image; In S53, the average grayscale value of the image is substituted into the calibration model for calculation to obtain the bone density of the examinee.

[0129] In this embodiment, the medical device can scan the target skeletal part of the subject (such as the lumbar spine or femoral neck) to obtain an attenuated reconstruction image of the subject, i.e., a scanned image.

[0130] Subsequently, the medical device can determine the skeletal region of the scanned image, i.e. the region of interest corresponding to the target bone site, based on the existing ROI selection method for clinical bone density testing.

[0131] In this embodiment, after determining the aforementioned bone region, the medical device can calculate the average gray value of all gray values ​​in the bone region and determine the average gray value as the image gray value, i.e., the measurement attenuation coefficient.

[0132] The medical device can then substitute the average grayscale value of the image into the calibration model for calculation, thereby obtaining the bone density of the examinee.

[0133] This embodiment first identifies the skeletal region from the scanned image of the subject, then extracts the average grayscale value of this region as the basis for calculation. This effectively avoids interference from non-skeletal regions such as soft tissue, fat, and air, ensuring that the grayscale data input into the calibration model only reflects the attenuation characteristics of bone. Compared to directly using the average grayscale value of the entire image, this significantly reduces the interference of irrelevant regions on the calculation results, improving the accuracy and reliability of bone density values. Furthermore, the regression coefficients and constants of the calibration model are obtained based on multi-calibration simulations, making the calculation logic traceable and verifiable, facilitating the establishment of a unified bone density detection standard. Therefore, after completing the construction of the calibration model, it can be directly reused to calculate bone density for different subjects without the need for repeated calibration.

[0134] As can be seen from the above, the bone mineral density calibration method provided in this application involves scanning multiple calibration phantoms with known concentrations to obtain attenuation reconstruction images; automatically identifying and segmenting the regions of interest corresponding to each calibration phantom in the attenuation reconstruction images based on an adaptive multi-threshold segmentation algorithm; calculating the average gray value of each region of interest to obtain the measurement attenuation coefficient corresponding to each calibration phantom; and establishing a calibration model between concentration and measurement attenuation coefficient through linear regression based on the known concentration and corresponding measurement attenuation coefficient of each calibration phantom: BMD = k μ+b; where k is the regression coefficient and b is a constant term; based on the calibration model and the obtained mean grayscale value of the corresponding image of the examinee, the bone density of the examinee is determined. This application can obtain the attenuated reconstruction image of all phantoms with only one scan for multiple calibration phantoms, which not only solves the problem of multiple scans required for single threshold segmentation, but also avoids systematic errors such as equipment drift and positional differences caused by multiple scans, ensuring that all calibration data are on the same scanning reference, providing more stable basic data for subsequent model fitting. Subsequently, the adaptive multi-threshold segmentation algorithm can automatically adapt to the grayscale distribution differences of different calibration phantoms, accurately segment the region of interest (ROI) corresponding to each phantom, effectively avoiding the undersegmentation and oversegmentation problems that are prone to occur in mixed images of multiple phantoms by traditional fixed threshold segmentation, and reducing soft tissue or noise interference. Subsequently, a calibration model was constructed based on the measured attenuation coefficients of multiple calibration phantoms and a set bone mineral density. The mapping relationship between grayscale and bone mineral density was quantified through regression coefficients and constant terms. Compared to the conversion method using a single calibration reference, the fitting results are more statistically reliable and reduce human calibration differences, thus significantly reducing the impact of individual calibration errors on the final results. This improves the overall accuracy of subsequent actual bone mineral density calculations.

[0135] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0136] Corresponding to the bone mineral density calibration method described in the above embodiments, Figure 6 This is a schematic diagram of a bone mineral density calibration device provided in one embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown. (Refer to...) Figure 6 The bone density calibration device 600 includes: a first scanning unit 61, an image segmentation unit 62, a mean calculation unit 63, a model construction unit 64, and a bone density determination unit 65. Wherein: The first scanning unit 61 is used to perform a single scan on multiple calibration phantoms of known concentrations to obtain attenuated reconstructed images.

[0137] The image segmentation unit 62 is used to automatically identify and segment the region of interest corresponding to each calibrated phantom in the attenuated reconstructed image based on an adaptive multi-threshold segmentation algorithm.

[0138] The mean calculation unit 63 is used to calculate the average gray value of each region of interest to obtain the measurement attenuation coefficient corresponding to each calibration phantom.

[0139] Model building unit 64 is used to establish a calibration model between concentration and measurement attenuation coefficient based on the known concentration and corresponding measurement attenuation coefficient of each calibration phantom through linear regression: BMD = k μ+b; where k is the regression coefficient and b is the constant term.

[0140] The bone density determination unit 65 is used to determine the bone density of the examinee based on the calibration model and the obtained average grayscale value of the corresponding image of the examinee.

[0141] In one embodiment of this application, the image segmentation unit 62 specifically includes: a segmentation unit, an initial interval determination unit, a target interval determination unit, and a region of interest determination unit. Wherein: The segmentation unit is used to divide each pixel in the attenuated reconstructed image into different gray-level intervals based on the gray-level value, and to determine the number of pixels in each gray-level interval after segmentation.

[0142] The initial interval determination unit is used to determine the initial interval corresponding to each calibration model based on the number of pixels in each grayscale interval and the number of calibration models.

[0143] The target interval determination unit is used to determine the target interval corresponding to each calibration phantom based on the boundary and interval step size of each initial interval.

[0144] The region of interest determination unit is used to determine the region of interest corresponding to each calibration phantom based on each target interval.

[0145] In one embodiment of this application, the bone density calibration device 600 further includes: a grayscale value determination unit, a difference calculation unit, and an interval number determination unit. Wherein: The grayscale value determination unit is used to determine the maximum grayscale value, the minimum grayscale value, and the total number of pixels from the attenuated reconstructed image.

[0146] The difference calculation unit is used to calculate the difference between the maximum gray value and the minimum gray value.

[0147] The interval number determination unit is used to determine the number of intervals based on the difference and the total number of pixels.

[0148] In one embodiment of this application, the initial interval determination unit specifically includes: a sorting unit and a first interval determination subunit. Wherein: The sorting unit is used to sort the grayscale intervals in ascending order of the number of pixels to obtain the interval order.

[0149] The first interval determination sub-unit is used to determine the grayscale interval that is the last in the sorted interval and equal to the number of phantoms as the initial interval, based on the interval order.

[0150] In one embodiment of this application, the boundary includes a lower boundary limit and an upper boundary limit; the target interval determination unit specifically includes: a subtraction operation unit, an addition operation unit, and a second interval determination subunit. Wherein: The subtraction unit is used to perform a subtraction operation on the lower boundary value of the initial interval and the interval step size of any calibration phantom to obtain the target lower boundary value.

[0151] The addition unit is used to perform addition operations on the upper boundary value and the interval step size of the initial interval of the calibration phantom to obtain the target upper boundary value.

[0152] The second interval determination sub-unit is used to determine the target interval corresponding to the calibration phantom based on the target lower limit and target upper limit.

[0153] In one embodiment of this application, the region of interest determination unit specifically includes: an image generation unit, a coordinate determination unit, a diameter determination unit, and a region construction unit. Wherein: The image generation unit is used to generate a binary image based on each target pixel in the target interval corresponding to any calibration phantom; the binary image includes multiple connected components.

[0154] The coordinate determination unit is used to determine the connected region with the largest area as the initial region corresponding to the calibration phantom, and to determine the center coordinates of the initial region.

[0155] The diameter determination unit is used to determine the region diameter of the initial region based on the center coordinates.

[0156] The region construction unit is used to determine the region of interest of the calibration phantom as the circular region constructed in the initial region with the center coordinate as the circle and the target length as the target diameter; the target length is the product of the preset ratio and the region diameter.

[0157] In one embodiment of this application, the bone density determination unit 64 specifically includes: a second scanning unit, a mean value determination unit, and a bone density calculation unit. Wherein: The second scanning unit is used to scan the subject, obtain scanned images, and determine the skeletal regions of the scanned images.

[0158] The mean value determination unit is used to determine the mean gray value of the image by the mean gray value corresponding to the bone region.

[0159] The bone density calculation unit is used to input the average grayscale value of the image into the calibration model to calculate the bone density of the examinee.

[0160] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0161] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0162] Figure 7 This is a schematic diagram of the structure of a medical device provided in one embodiment of this application. Figure 7 As shown, the medical device 7 of this embodiment includes: at least one processor 70 ( Figure 7 (Only one is shown in the diagram), memory 71, and computer program 72 stored in said memory 71 and executable on said at least one processor 70, wherein the processor 70 executes said computer program 72 to implement the steps in any of the above-described bone density calibration method embodiments.

[0163] The medical device 7 may include, but is not limited to, a processor 70 and a memory 71. Those skilled in the art will understand that... Figure 7 This is merely an example of medical device 7 and does not constitute a limitation on medical device 7. It may include more or fewer components than shown in the figure, or combine certain components, or different components, such as input / output devices, network access devices, etc.

[0164] The processor 70 may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0165] In some embodiments, the memory 71 may be an internal storage unit of the medical device 7, such as the RAM of the medical device 7. In other embodiments, the memory 71 may be an external storage device of the medical device 7, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the medical device 7. Furthermore, the memory 71 may include both internal and external storage units of the medical device 7. The memory 71 is used to store the operating system, applications, boot loader, data, and other programs, such as the program code of the computer program. The memory 71 can also be used to temporarily store data that has been output or will be output.

[0166] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0167] This application provides a computer program product that, when run on a medical device, enables the medical device to perform the steps described in the above-described method embodiments.

[0168] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a medical device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, such as a USB flash drive, a portable hard drive, a magnetic disk, or an optical disk.

[0169] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0170] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for bone mineral density calibration, characterized in that, include: S1: Perform a single scan on multiple calibration phantoms with known concentrations to obtain attenuated reconstructed images; S2: Based on the adaptive multi-threshold segmentation algorithm, the region of interest corresponding to each of the calibration phantoms is automatically identified and segmented in the attenuated reconstructed image; S3: Calculate the average gray value of each region of interest to obtain the measurement attenuation coefficient corresponding to each calibration phantom; S4: Based on the known concentrations and corresponding measurement attenuation coefficients of each calibration phantom, a calibration model between concentration and measurement attenuation coefficient is established through linear regression: BMD = k μ+b; where k is the regression coefficient and b is the constant term; S5: Determine the bone density of the examinee based on the calibration model and the obtained average grayscale value of the image corresponding to the examinee.

2. The bone mineral density calibration method as described in claim 1, characterized in that, The adaptive multi-threshold segmentation algorithm automatically identifies and segments the region of interest corresponding to each calibrated phantom in the attenuated reconstructed image, including: S21: Divide each pixel in the attenuated reconstructed image into different grayscale intervals based on the grayscale value, and determine the number of pixels in each of the grayscale intervals after division. S22: Based on the number of pixels in each grayscale interval and the number of calibrated phantoms, determine the initial interval corresponding to each calibrated phantom; S23: Based on the boundaries and interval step size of each initial interval, determine the target interval corresponding to each calibration phantom; S24: Based on each of the target intervals, determine the region of interest corresponding to each of the calibration phantoms.

3. The bone mineral density calibration method as described in claim 2, characterized in that, The number of grayscale intervals is determined as follows: S01: Determine the maximum gray value, the minimum gray value, and the total number of pixels from the attenuated reconstructed image; S02: Calculate the difference between the maximum gray value and the minimum gray value; S03: Determine the number of intervals based on the difference and the total number of pixels.

4. The bone mineral density calibration method as described in claim 2, characterized in that, The step of determining the initial interval corresponding to each calibration phantom based on the number of pixels in each grayscale interval and the number of phantoms in each calibration phantom includes: S221: Sort the grayscale intervals in ascending order of the number of pixels to obtain the interval order; S222: Based on the interval order, the grayscale interval that is the last in the sort and equal to the number of the phantoms is determined as the initial interval.

5. The bone mineral density calibration method as described in claim 2, characterized in that, The boundary includes a lower boundary limit and an upper boundary limit; determining the target interval corresponding to each calibration phantom based on the boundary and interval step size of each initial interval includes: S231: For any of the calibration phantoms, perform a subtraction operation on the lower boundary value of the initial interval of the calibration phantom and the interval step size to obtain the target lower boundary value; S232: Perform an addition operation on the upper boundary value of the initial interval of the calibration phantom and the interval step size to obtain the target upper boundary value; S233: Based on the target lower limit and the target upper limit, determine the target interval corresponding to the calibration phantom.

6. The bone mineral density calibration method as described in claim 2, characterized in that, The step of determining the region of interest corresponding to each of the calibrated phantoms based on each of the target intervals includes: S241: For any one of the calibration phantoms, generate a binary image based on each target pixel in the target interval corresponding to the calibration phantom; the binary image includes multiple connected components; S242: Determine the connected region with the largest area as the initial region corresponding to the calibration phantom, and determine the center coordinates of the initial region; S243: Determine the region diameter of the initial region based on the center coordinates; S244: The circular region constructed in the initial region with the center coordinates as the center point and the target length as the target diameter is determined as the region of interest of the calibration phantom; the target length is the product of a preset ratio and the diameter of the region.

7. The bone mineral density calibration method according to any one of claims 1-6, characterized in that, The step of determining the bone density of the examinee based on the calibration model and the obtained average grayscale value of the image corresponding to the examinee includes: S51: Scan the subject to obtain a scanned image and determine the skeletal region of the scanned image; S52: Determine the average gray value corresponding to the bone region as the average gray value of the image; S53: Substitute the mean grayscale value of the image into the calibration model for calculation to obtain the bone density of the examinee.

8. A bone mineral density calibration device, characterized in that, include: The first scanning unit is used to perform a single scan on multiple calibration phantoms of known concentrations to obtain attenuated reconstructed images; An image segmentation unit is used to automatically identify and segment the region of interest corresponding to each of the calibration phantoms in the attenuated reconstructed image based on an adaptive multi-threshold segmentation algorithm. The mean calculation unit is used to calculate the average gray value of each region of interest to obtain the measurement attenuation coefficient corresponding to each calibration phantom. The model building unit is used to establish a calibration model between concentration and measurement attenuation coefficient based on the known concentration and corresponding measurement attenuation coefficient of each calibration phantom through linear regression: BMD = k μ+b; where k is the regression coefficient and b is a constant; The bone density determination unit is used to determine the bone density of the examinee based on the calibration model and the obtained average grayscale value of the image corresponding to the examinee.

9. A medical device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the bone mineral density calibration method as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, Includes a computer program that, when run, implements the bone mineral density calibration method as described in any one of claims 1 to 7.

Images