A device and method for acquiring a three-dimensional model of abdominal organs

Abstract

This invention provides a device and method for acquiring a three-dimensional model of abdominal organs, comprising: a coil support assembly, an MRI scanner, and an image processing device; the coil support assembly is clamped and fixed to the side of the examination bed to support the abdominal coil and reduce pressure on the abdomen of the user; the MRI scanner is positioned above the abdomen of the user, with the axis of the examination positioning light aligned with the midpoint of the axis of the coil support assembly, and a magnetic resonance imaging (MRI) scan sequence is acquired by scanning with the MRI scanner; the image processing device is communicatively connected to the MRI scanner, acquires the MRI scan sequence, and obtains a three-dimensional model of the abdominal organs based on the MRI scan sequence. This invention achieves precise extraction of abdominal organs at all levels through multi-region seed point strategies and adaptive threshold segmentation, overcoming the inefficiency and subjective limitations of traditional manual segmentation, and ensuring the reliability of the results through automated algorithms, thus providing support for the image analysis of abdominal organs.

Description

Technical Field

[0001] This invention relates to the field of medical imaging technology, and in particular to a device and method for acquiring a three-dimensional model of abdominal organs. Background Technology

[0002] In clinical practice, obtaining three-dimensional models of abdominal organs mainly relies on two techniques: CT scanning and MR scanning. Both require data acquisition before three-dimensional image reconstruction, but each has its limitations in application.

[0003] CT scans, due to their ionizing radiation, pose potential risks to radiation-sensitive groups such as pregnant women and adolescents. MR scans, which do not involve ionizing radiation and are therefore safer, face several key challenges in practice: 1. Compression risk: The abdominal radiofrequency coil used in MR scans has considerable weight, directly compressing the patient's abdominal cavity during the scan. For pregnant women and other vulnerable groups, this compression could cause fetal distress, posing a safety hazard; 2. Motion artifact interference: MR sequences typically have long acquisition times and are sensitive to motion artifacts. Abdominal organs are significantly affected by physiological movements such as respiration, gastrointestinal peristalsis, and vascular pulsation. Although respiratory gating technology is currently used, it can only mitigate artifacts caused by respiratory motion to a certain extent and cannot resolve interference from other physiological movements, affecting image quality; 3. Insufficient tissue contrast: MR scan sequence parameters are complex, and different scanning methods can affect the strength of echo signals and image contrast between different tissues. Summary of the Invention

[0004] Therefore, it is necessary to provide a device and method for acquiring a three-dimensional model of abdominal organs to address the aforementioned technical problems.

[0005] A device for acquiring a three-dimensional model of abdominal organs includes: a coil support assembly, a magnetic resonance imaging (MRI) scanner, and an image processing device;

[0006] The coil support assembly is clamped and fixed to the side of the examination bed to support the abdominal coil and reduce pressure on the abdomen of the user being tested. The MRI scanner is positioned above the abdomen of the user being tested, with the axis of the examination positioning light aligned with the midpoint of the axis of the coil support assembly. The MRI scanner is used to obtain a magnetic resonance scanning sequence. The image processing device is communicatively connected to the MRI scanner to obtain the magnetic resonance scanning sequence and to obtain a three-dimensional model of the abdominal organs based on the magnetic resonance scanning sequence.

[0007] The step of obtaining a three-dimensional model of the abdominal organs based on the magnetic resonance scanning sequence includes:

[0008] The intra-abdominal organ tissue segmentation results were obtained based on the magnetic resonance scanning sequence and the automated intra-abdominal organ image segmentation technology based on grayscale features.

[0009] An initial abdominal visceral tissue model is generated based on the segmentation results of the abdominal visceral tissue. The triangular facets are then adjusted using surface simplification coefficients to optimize the initial abdominal visceral tissue model, resulting in an optimized three-dimensional mesh model.

[0010] The optimized 3D mesh model was validated in multiple dimensions and standardized for export to obtain a 3D model of the abdominal organs.

[0011] In one embodiment, the coil support assembly includes a side fixing bracket group, the lower part of which is provided with a fixing clamping mechanism for clamping and fixing the side fixing bracket group to the side of the examination bed; the upper part of the side fixing bracket group is provided with a height adjustment mechanism, the top output end of which is provided with a rotating shaft structure, the rotating shaft structure is connected to an arc-shaped support plate, the arc-shaped support plate is rotatably mounted on the top output end of the height adjustment mechanism via the rotating shaft structure; at least two abdominal coil binding holes are evenly provided on the arc-shaped support plate, the abdominal coil binding holes are used for passing through the binding strap of the abdominal coil.

[0012] In one embodiment, the MRI scanner is positioned above the abdomen of the user being tested. The alignment of the positioning light's axis with the midpoint of the coil support assembly's axis is checked. Obtaining a magnetic resonance imaging (MRI) sequence using the MRI scanner includes:

[0013] Step S1: Use a fast localization scanning sequence to acquire sagittal, coronal, and transverse localization images, and determine the scanning range of the main sequence based on the localization images;

[0014] Step S2: Define the direction of the main magnetic field as the Z-axis, which is parallel to the long axis of the human body and points to the foot side; the X-axis points to the right side of the human body; and the Y-axis points to the front side of the human body. Set the gradient field spatial encoding with the Z-axis as the layer direction. Set the direction with the smaller value among the anterior-posterior diameter thickness and lateral diameter width of the abdomen as the phase encoding direction, and the direction with the larger value as the frequency encoding direction.

[0015] Step S3: Based on the gradient field spatial encoding, the scanning range is continuously excited by small-angle radio frequency pulses with a flip angle of 30°-65°, and the readout gradient field is used to quickly switch and acquire echo signals to obtain the raw MR signal data.

[0016] Step S4: Process the raw MR signal data, and use half-echo acquisition based on Fourier space conjugate symmetry to reconstruct the magnetic resonance scanning sequence.

[0017] In one embodiment, step S3 further includes:

[0018] Add a combination of preparatory pulses at 90°, 180°, and -90° before the main sequence.

[0019] In one embodiment, step S3 further includes:

[0020] A preset current is applied to the gradient coils in the X, Y, and Z axes.

[0021] In one embodiment, the intra-abdominal visceral tissue segmentation result obtained based on the magnetic resonance scanning sequence and the automated intra-abdominal visceral image segmentation technology based on grayscale features includes:

[0022] Seed points are set in the high-signal area, low-signal area and medium-signal area of ​​the target abdominal viscera respectively, and an initial mask covering the entire range of the abdominal viscera is generated by the seed region growth algorithm.

[0023] The initial mask is subjected to boundary contraction to remove the blood vessel wall, leaving the pure parenchyma area, thus obtaining the organ parenchyma;

[0024] For abdominal organs containing fluid, the fluid region is extracted using an adaptive threshold segmentation algorithm;

[0025] For small low-signal abdominal organs, a signal threshold is obtained, and small low-signal structures are extracted based on volume screening conditions and the signal threshold.

[0026] The intra-abdominal organ tissue segmentation results include the organ parenchyma, the fluid region, and the small low-signal structures.

[0027] In one embodiment, an initial abdominal visceral tissue model is generated based on the segmentation results of the abdominal visceral tissue, and the initial abdominal visceral tissue model is optimized by adjusting the triangular facets using surface simplification factors to obtain an optimized three-dimensional mesh model, including:

[0028] Adjust the triangular facets according to the following formula:

[0029] ;

[0030] in, Indicates the number of original mesh patches. This indicates the number of target facets after simplification. This represents the simplification factor.

[0031] In one embodiment, the optimized 3D mesh model is validated in multiple dimensions and standardized for export to obtain a 3D model of the abdominal organs, including:

[0032] The optimized 3D mesh model is validated in two dimensions. The first validation model is obtained when the fit error between the segmented contour and the original image is ≤1 pixel.

[0033] The original image is superimposed on the first verification model by volume rendering, and a three-dimensional model of the abdominal organs is obtained in response to the absence of perforations or missing parts.

[0034] The export formats for the three-dimensional models of abdominal organs include STL, VTK, and DICOM-SEG formats.

[0035] A method for acquiring a three-dimensional model of abdominal organs, used in a device for acquiring a three-dimensional model of abdominal organs as described above, comprising:

[0036] The coil support assembly is clamped and fixed to the side of the examination bed;

[0037] The user to be tested lies supine on the examination bed, and the abdominal coil is fixed above the abdomen by the coil support assembly, with the abdomen located at the center of the abdominal coil.

[0038] The MRI scanner is positioned above the abdomen of the user to be tested. The axis line of the positioning light is aligned with the midpoint of the axis of the coil support assembly. The MRI scanner is used to obtain the magnetic resonance scan sequence.

[0039] The image processing device is communicatively connected to the magnetic resonance imaging (MRI) scanner to acquire the MRI scan sequence and obtain a three-dimensional model of the abdominal organs based on the MRI scan sequence.

[0040] The step of obtaining a three-dimensional model of the abdominal organs based on the magnetic resonance scanning sequence includes:

[0041] The intra-abdominal organ tissue segmentation results were obtained based on the magnetic resonance scanning sequence and the automated intra-abdominal organ image segmentation technology based on grayscale features.

[0042] An initial abdominal visceral tissue model is generated based on the segmentation results of the abdominal visceral tissue. The triangular facets are then adjusted using surface simplification coefficients to optimize the initial abdominal visceral tissue model, resulting in an optimized three-dimensional mesh model.

[0043] The optimized 3D mesh model was validated in multiple dimensions and standardized for export to obtain a 3D model of the abdominal organs.

[0044] Compared to existing technologies, the advantages and beneficial effects of this invention are as follows: The coil support assembly designed in this invention, through an adjustable side fixing bracket, height adjustment mechanism, and arc-shaped support plate, can stably support the abdominal coil to avoid abdominal compression on the patient (especially pregnant women and other special groups), and ensure coil fit through binding perforations, thus improving signal acquisition stability. Employing small-angle radio frequency pulses, extremely short repetition times, and half-echo acquisition techniques significantly shortens data acquisition time. Combined with preparatory pulses to enhance the contrast between soft tissue and fluid, it effectively suppresses motion artifacts and magnetic susceptibility artifacts caused by respiration and gastrointestinal peristalsis, balancing imaging efficiency and quality. The automated segmentation technology based on grayscale features, through multi-region seed point growth and adaptive threshold screening strategies, achieves precise extraction of organs such as the liver, gallbladder, and gallstones, as well as fine structures. Combined with surface simplification coefficients to optimize 3D modeling, the final generated model supports multi-format export and is suitable for clinical scenarios such as 3D printing and surgical planning. This invention forms a synergistic advantage in safety, imaging quality, automation, and clinical applicability, providing an efficient solution for the accurate acquisition of 3D models of intra-abdominal organs. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of a device for obtaining a three-dimensional model of abdominal organs according to one embodiment of the method;

[0046] Figure 2 This is a schematic diagram of the coil support assembly structure in one embodiment;

[0047] Figure 3 This is a schematic diagram illustrating the reduction of scan time based on phase encoding direction optimization in one embodiment;

[0048] Figure 4 This is a schematic diagram of the echo signal sequence acquired in one embodiment;

[0049] Figure 5 This is a schematic diagram of the flip angle of a radio frequency pulse in one embodiment;

[0050] Figure 6 This is a schematic diagram of the longitudinal magnetization vector in one embodiment;

[0051] Figure 7 This is a schematic diagram of a pulse combination of 90°, 180°, and -90° in one embodiment;

[0052] Figure 8 This is a schematic diagram illustrating the import of a magnetic resonance scanning sequence in one embodiment;

[0053] Figure 9 This is a schematic diagram of trimming three-dimensional data in one embodiment;

[0054] Figure 10This is a schematic diagram of the initial three-dimensional mask data in one embodiment;

[0055] Figure 11 This is a schematic diagram of three-dimensional liver data in one embodiment;

[0056] Figure 12 This is a schematic diagram of gallstones in one embodiment;

[0057] Figure 13 This is a schematic diagram of a triangular mesh for adjusting the surface simplification factor in one embodiment;

[0058] Figure 14 This is a schematic diagram of smoothing processing in one embodiment;

[0059] Figure 15 This is a schematic diagram of a three-dimensional model of abdominal organs in one embodiment;

[0060] Figure 16 This is a schematic diagram of the model export type in one embodiment.

[0061] In the figure, 1-coil support assembly, 2-MRI machine, 3-image processing equipment, 10-side fixing bracket group, 11-fixing clamping mechanism, 12-height adjustment mechanism, 13-rotation shaft structure, 14-arc support plate, 15-abdominal coil binding perforation. Detailed Implementation

[0062] Before describing the specific embodiments of the present invention, the overall concept of the present invention will be explained as follows:

[0063] This invention is mainly developed to address the problems of insufficient safety, interference with image quality, and limited accuracy of 3D modeling in the process of acquiring 3D models of abdominal organs by MR scanning. Currently, CT scans, which are relied upon in clinical practice, pose a risk of ionizing radiation, while traditional MR scans face problems such as the risk of abdominal coil compression (especially for pregnant women), significant motion artifact interference, and insufficient tissue contrast. These issues result in unstable quality of the acquired image data, making it difficult to meet the precise clinical needs for subsequent 3D model segmentation and reconstruction, and may even affect examination efficiency due to scan interruption.

[0064] Through analysis, the inventors discovered that the main reasons for these problems are: the existing coils lack adaptable support structures, leading to direct compression of the abdominal cavity; traditional scanning sequences have excessively long acquisition times and are sensitive to physiological motion, while lacking targeted artifact suppression mechanisms; sequence parameter design does not adequately highlight the signal differences between soft tissue and fluid, and subsequent image segmentation relies heavily on manual operation, resulting in low standardization. These problems can be avoided by innovating the design from multiple dimensions, including scanning device structure, sequence parameter optimization, motion artifact suppression, and automated segmentation modeling. Therefore, this invention proposes an MR scanning device and method for acquiring three-dimensional models of abdominal organs. It eliminates the risk of compression through adjustable coil support components, improves imaging quality through rapid acquisition sequences and multi-dimensional artifact suppression technology, and combines automated segmentation and precise modeling processes based on grayscale features to achieve safe, efficient, and high-precision acquisition of three-dimensional models of abdominal organs, effectively meeting the needs of clinical diagnosis and treatment planning.

[0065] After introducing the overall concept of the present invention, in order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below through specific embodiments and in conjunction with the accompanying drawings.

[0066] It should be noted that, unless otherwise defined, the technical or scientific terms used in one or more embodiments of this specification should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in one or more embodiments of this specification do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word covers the element or object listed following the word and its equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0067] For ease of understanding, the terms used in the embodiments of this invention are explained below:

[0068] MR stands for Magnetic Resonance. In the medical field, "MR" is usually combined with "imaging," namely Magnetic Resonance Imaging (MRI).

[0069] CT stands for Computed Tomography. It is an imaging technique that uses X-rays to perform tomographic scans of specific parts of the human body, and then processes the data using a computer to reconstruct the tomographic images.

[0070] TR time: Repetition Time. It is a key parameter in MRI scan sequences, referring to the time interval between two adjacent radiofrequency pulses.

[0071] T1 value: Longitudinal Relaxation Time. It is a time constant that describes the recovery of an atomic nucleus (mainly hydrogen protons) from an excited state to an equilibrium state (longitudinal magnetization vector recovery) after the radio frequency pulse stops, reflecting the speed of longitudinal magnetization recovery.

[0072] K-space: Fourier Space. It is a core concept in MRI image reconstruction. Essentially, it is a frequency domain data space used to store the frequency and phase information of magnetic resonance signals, rather than a direct anatomical spatial image.

[0073] TE: Echo Time. It is a key parameter in magnetic resonance imaging (MRI), referring to the time interval between the excitation of the radiofrequency pulse and the acquisition of the echo signal.

[0074] T2 value: Transverse Relaxation Time. It is a time constant describing the decay of the transverse magnetization vector of hydrogen protons in tissue after the radio frequency pulse stops, reflecting the rate at which transverse magnetization disappears.

[0075] SAR value: Specific Absorption Rate. It is a physical quantity that measures the rate at which the human body absorbs electromagnetic energy in an electromagnetic field.

[0076] DICOM: Digital Imaging and Communications in Medicine. It is an international standard data format in the field of medical imaging, used for storing, transmitting, and sharing medical images and related information. It is the core data carrier for hospital PACS (Picture Archiving and Communication System) and RIS (Radiology Information System).

[0077] STL: Standard Tessellation Language. It is a file format used to represent the surface geometry of three-dimensional models.

[0078] VTK: Visualization Toolkit. It is an open-source, cross-platform 3D visualization and graphics processing library.

[0079] DICOM-SEG: This stands for DICOM Segmentation Image IOD (Information Object Definition). It is a file format in the DICOM standard specifically used for storing and transmitting medical image segmentation results.

[0080] RAI coordinate system: In the field of medical imaging, it is a form of anatomical coordinate system. R (Right) represents the negative left direction, that is, from right to left; A (Anterior) represents the negative rear direction, that is, from front to back; and I (Inferior) represents the negative up direction.

[0081] ROI: Region of Interest. It refers to the specific area in an image or 3D data that researchers or clinicians focus their attention on.

[0082] SSF: Surface Simplification Factor.

[0083] Delta B-zero describes the inhomogeneity of the main magnetic field, representing the difference between the actual main magnetic field and the ideal uniform main magnetic field, i.e., the inhomogeneity value of the main magnetic field.

[0084] In one embodiment, such as Figure 1 As shown, a device for acquiring a three-dimensional model of abdominal organs is provided, including: a coil support assembly 1, a magnetic resonance imaging (MRI) instrument 2, and an image processing device 3;

[0085] The coil support assembly 1 is clamped and fixed to the side of the examination bed to support the abdominal coil and reduce pressure on the abdomen of the user being tested. The MRI scanner 2 is set above the abdomen of the user being tested, and the axis line of the examination positioning light is aligned with the midpoint of the axis of the coil support assembly 1. The MRI scanner 2 scans and obtains the magnetic resonance scan sequence. The image processing device 3 is connected to the MRI scanner 2 to obtain the magnetic resonance scan sequence and obtain a three-dimensional model of the abdominal organs based on the magnetic resonance scan sequence.

[0086] like Figure 2As shown, the coil support assembly 1 includes a side fixing bracket group 10. A fixing clamping mechanism 11 is provided at the lower part of the side fixing bracket group 10. The fixing clamping mechanism 11 is used to clamp and fix the side fixing bracket group 10 to the side of the examination bed. A height adjustment mechanism 12 is provided at the upper part of the side fixing bracket group 10. A rotating shaft structure 13 is provided at the top output end of the height adjustment mechanism 12. An arc-shaped support plate 14 is connected to the rotating shaft structure 13. The arc-shaped support plate 14 is rotatably mounted on the top output end of the height adjustment mechanism 12 through the rotating shaft structure 13. At least two abdominal coil binding holes 15 are evenly opened on the arc-shaped support plate 14. The abdominal coil binding holes 15 are used for the binding strap of the abdominal coil to pass through.

[0087] The abdominal coil is placed on the examination bed, and the side fixing bracket group 10 is fixed to the left and right sides of the examination bed by the fixing clamping mechanism 11. The arc-shaped support plate 14 is flipped open, so that the abdomen of the user to be tested can easily lie on the examination bed and the abdominal coil. Then, the arc-shaped support plate 14 is flipped so that it is fastened to the limiting support platform on the side fixing bracket group 10. The anterior abdominal coil is then placed on the upper surface of the arc-shaped support plate 14. The extension height of the side fixing bracket 10 is adjusted by the height adjustment mechanism 12 so that the lower surface of the arc-shaped support plate 14 basically fits the upper abdomen of the user to be tested without causing pressure. Then, the binding strap is passed through the abdominal coil binding hole 15 to tighten and bind the two abdominal coils, so that the anterior and posterior abdominal coils fit the human body better and the examination effect is better.

[0088] In this embodiment, the coil support group 1 adopts a combination design of side fixing bracket group 10 and fixing clamping mechanism 11, which can stably clamp the side of the examination bed, is easy to install and adaptable to different sizes of examination beds; the height adjustment mechanism 12 precisely and synchronously adjusts the height of the two side supports to ensure that the arc support plate 14 fits smoothly against the abdomen of the user to be tested, effectively distributing the weight of the abdominal coil and avoiding the risk of compression on the abdominal cavity (especially pregnant women and other special groups); the arc support plate 14 can be flexibly flipped through the rotating shaft structure 13, which not only facilitates the user to get on and off the bed, but also forms a stable support through the arch bridge design, which, together with the abdominal coil, ensures its fit with the human body and improves the stability of signal acquisition; the overall structure takes into account both safety and comfort, and adapts to different body types of examinees through multi-dimensional adjustment, providing a reliable coil support solution for high-quality MR scanning.

[0089] In one embodiment, the MRI scanner 2 is positioned above the abdomen of the user being tested. The alignment of the positioning light's axis with the midpoint of the coil support assembly 1's axis is checked. The MRI scanner 2 is used to obtain a magnetic resonance imaging (MRI) scan sequence, including:

[0090] Step S1: Use a fast localization scanning sequence to acquire sagittal, coronal, and transverse localization images, and determine the scanning range of the main sequence based on the localization images.

[0091] Step S2: Define the direction of the main magnetic field as the Z-axis, which is parallel to the long axis of the human body and points to the foot side; the X-axis points to the right side of the human body; and the Y-axis points to the front side of the human body. Set the gradient field spatial encoding with the Z-axis as the layer direction. Set the direction with the smaller value among the anterior-posterior diameter thickness and lateral diameter width of the abdomen as the phase encoding direction, and the direction with the larger value as the frequency encoding direction.

[0092] Step S3: Based on the gradient field spatial encoding, the scanning range is continuously excited by small-angle radio frequency pulses with a flip angle of 30°-65°, and the readout gradient field is used to quickly switch and acquire echo signals to obtain the raw MR signal data.

[0093] Step S4: Process the raw MR signal data, and use half-echo acquisition based on Fourier space conjugate symmetry to reconstruct the magnetic resonance scanning sequence.

[0094] Specifically, a rapid localization scanning sequence is used to acquire sagittal, coronal, and transverse localization images. The final scanning range of the main sequence is determined based on the acquired localization images, which must include the entire abdominal viscera. The main magnetic field direction is defined as the Z-axis of the MRI scanner 2. That is, when the user is supine with feet first, the Z-axis is parallel to the long axis of the body and points towards the feet; the X and Y axes are perpendicular to the Z-axis, with the Y-axis pointing anteriorly to the anatomical position and the X-axis pointing to the right. Gradient fields are applied to the gradient coils in the X, Y, and Z directions for spatial encoding, with the Z-axis direction set as the slice direction. Based on individual differences in the anteroposterior diameter thickness and lateral diameter width of the user's abdomen, directions with smaller values ​​are set as phase encoding directions, and those with larger values ​​are set as frequency encoding directions.

[0095] like Figure 3 The number of scanning layers can be increased or decreased based on the size of the area to be scanned, and the scanning time will also increase or decrease accordingly, typically fluctuating by several seconds. The scanning time is directly related to the number of steps in phase encoding. Setting the direction with the shorter anatomical diameter as the phase encoding direction can reduce the scanning time by setting a rectangular field of view. In the figure, Kx represents the number of steps in the frequency encoding direction, and Ky represents the number of steps in the phase encoding direction.

[0096] Precisely measure the dimensions of the area to be scanned along its long and short axes. If the scanning area is large, such as a full scan of the entire abdomen to observe multiple organs, the anatomical diameter of this region is long, and the number of scanning layers should be appropriately increased to fully cover this area. Conversely, if only a small, specific intra-abdominal organ (such as a localized gallbladder) is being scanned, the number of scanning layers can be reduced accordingly due to its short anatomical diameter. Acquiring sagittal, coronal, and transverse localization images through rapid initial localization scanning sequences allows for precise measurement of the specific diameter values ​​of the area to be scanned, providing a reliable basis for the preliminary determination of the number of scanning layers. Assuming that the long diameter of a patient's liver in the coronal plane is measured to be 15cm based on localization images, and according to the equipment parameters, if a slice thickness of 3mm is used, the initial estimated number of scanning layers to fully cover the liver is approximately 150mm ÷ 3mm = 50 layers, but further adjustments are needed considering other factors.

[0097] Small-angle radio frequency pulses with a flip angle of 30°-65° are used to continuously excite the selected layer, and the echo signal is acquired by rapidly switching the readout gradient field. sequence diagrams as follows Figure 4 As shown.

[0098] This sequence uses extremely short repetition times (TR < 5 ms) to continuously excite the tissue, generating a radiofrequency magnetic field (B1 field). Each radiofrequency pulse has the same number of flip angles, but the polarity differs between every two radiofrequency pulses (e.g., ...). Figure 5 As shown, the flip angle of each radio frequency pulse is 45°, but the first one is in the same direction as the main magnetic field, the second one is in the opposite direction, and so on. After the first 4-6 radio frequency pulses, the echo signal is not collected. Instead, the echo signal is collected only after the macroscopic longitudinal magnetization vector and macroscopic transverse magnetization vector of the tissue have reached a relatively stable state. The rapid switching of the readout gradient field is then used to collect the echo signal.

[0099] Taking the macroscopic longitudinal magnetization vector as an example, before each radio frequency pulse excitation, its magnitude is the superposition of the residual longitudinal component and the longitudinal magnetization vector recovered in the previous repetition time (TR). Because the tissue is continuously excited using an extremely short repetition time shorter than the tissue's T1 value, each tissue cannot complete sufficient longitudinal relaxation within each TR cycle. Assuming the excitation pulse flip angle is α and the initial longitudinal magnetization vector magnitude is 1, then after the first excitation, the residual longitudinal component is 1×cosα. Before the second radio frequency pulse excitation, the longitudinal magnetization vector becomes the sum of cosα and M1 (where M1 is the magnitude of the magnetization vector recovered to the longitudinal direction within the first TR); after the second excitation, the residual longitudinal component is (cosα+M1)×cosα, while M2 represents the magnitude of the magnetization vector recovered to the longitudinal direction within the second TR. Figure 6As shown, similarly, after multiple excitations, the longitudinal magnetization vector of the tissue gradually reaches a relatively stable state. Gradient fields of the same magnitude but opposite direction to the corresponding encoding gradients are applied to the slice selection direction, phase encoding direction, and frequency encoding direction, correcting the proton group phase dephase caused by these three encoding gradient fields.

[0100] Since the original data space of the acquired MR signal—the Fourier space (K-space)—is mirror-symmetric (conjugate symmetric) in the frequency coding direction, only half of the echo is acquired. The other half of the echo can be simulated and filled according to the principle of symmetry to obtain the magnetic resonance scanning sequence. Using half-echo acquisition will further shorten the scanning time of the main sequence.

[0101] The conjugate symmetry of K-space data stems from the real-valued nature of MRI signals. Assume the acquired time-domain signal... If it is a real number, then its Fourier transform satisfy:

[0102] ;

[0103] in This indicates complex conjugation. This means that in the K-space, k and... The positional data are conjugate complex numbers, with symmetric real parts and antisymmetric imaginary parts. If data with k ≥ 0 are collected... ,but Data can be obtained through generate.

[0104] In this embodiment, the use of a rapid localization scanning sequence to acquire sagittal, coronal, and transverse localization images and determine the main sequence scanning range has several significant advantages: First, the rapid localization sequence can complete multi-planar image acquisition in a very short time, reducing the patient's waiting time and breath-holding burden, making it especially suitable for special groups who cannot maintain a fixed posture for a long time (such as pregnant women and elderly patients); Second, through localization images in three orthogonal planes, the anatomical location, morphology, and adjacent relationships of abdominal organs can be displayed intuitively from all angles, providing a three-dimensional reference for accurately defining the main sequence scanning range, avoiding unnecessary time waste due to an excessively large scanning range, and preventing the omission of abdominal organs or lesions due to an excessively small range; In addition, this method can flexibly adjust the scanning boundary according to individual anatomical differences (such as variations in the size and position of abdominal organs), ensuring that the main sequence acquisition data completely covers the target area (such as completely including the liver and gallbladder, the entire bladder, etc.), laying a complete and accurate data foundation for subsequent image reconstruction and 3D modeling, and improving the efficiency and reliability of the overall examination.

[0105] Small-angle radio frequency pulses reduce excessive consumption of the longitudinal magnetization vector of the tissue, enabling continuous excitation with an extremely short repetition time (TR < 5ms). This shortens the overall data acquisition time, allows the user to complete the scan by holding their breath, effectively avoids artifact interference caused by physiological movements such as breathing and gastrointestinal peristalsis, and maintains stable signal output, ensuring the image signal-to-noise ratio. At the same time, the design of opposite polarity between every two radio frequency pulses, combined with the strategy of waiting for the magnetization vector to stabilize after the initial 4-6 excitations before acquiring the signal, reduces the cumulative error of the residual magnetization vector and improves signal consistency. The rapid switching of the readout gradient field accelerates the acquisition of echo signals, further compressing the sequence length. Furthermore, by applying reverse gradient fields in three encoding directions to correct proton group dephase, artifacts are significantly reduced. Ultimately, while ensuring high contrast between soft tissue and fluid, high-quality and highly stable raw data are provided for subsequent 3D model reconstruction, balancing scanning efficiency and imaging accuracy.

[0106] The conjugate symmetry of K-space data (i.e., the k and -k position data are conjugate complex numbers) provides a theoretical basis for half-echo acquisition. By acquiring only half of the signal after the echo peak (corresponding to the k≥0 region), the other half of the echo data can be simulated and filled through the principle of symmetry. This can significantly shorten the main sequence scanning time while ensuring image quality and reduce the burden of breath-holding on the user, especially suitable for areas such as the abdomen that are easily affected by physiological movement. At the same time, half-echo acquisition avoids the longer gradient switching time required for full echo acquisition, reduces the risk of artifacts caused by gradient field changes, and ensures that image resolution and signal integrity are not compromised by completing the data through mathematical symmetry during reconstruction. Ultimately, while improving scanning efficiency, it provides a high-quality imaging basis for the accurate segmentation and reconstruction of 3D models of abdominal organs, achieving an effective balance between efficiency and accuracy.

[0107] In one embodiment, step S3 further includes:

[0108] Add a combination of preparatory pulses at 90°, 180°, and -90° before the main sequence.

[0109] Specifically, such as Figure 7 As shown, a combination of 90°, 180°, and -90° pulses is added before the main sequence as a preparatory pulse combination. The first 90° pulse of this preparatory pulse transforms the macroscopic longitudinal magnetization vector of the tissue into a macroscopic transverse magnetization vector. After the 90° pulse is turned off, a 180° radio frequency pulse is applied at an appropriate time (1 / 2 TE of the preparatory pulse). Subsequently, the macroscopic transverse magnetization vector will refocus at the TE time of the preparatory pulse. Since different tissues have different T2 relaxation rates, there is a T2 contrast, and the refocused macroscopic transverse magnetization vectors are different. Then, a -90° pulse is used to return the macroscopic transverse magnetization vector to the longitudinal direction, becoming the macroscopic longitudinal magnetization vector.

[0110] In this embodiment, a 90° pulse first converts the longitudinal magnetization vector to the transverse direction. A 180° pulse at a specific moment causes the transverse magnetization vector to refocus, highlighting the T2 relaxation differences between different tissues. Finally, a -90° pulse converts the refocused transverse magnetization vector back to the longitudinal direction. This process enhances the T2 contrast between soft tissues and fluids (such as bile, amniotic fluid, and blood), making the boundaries of abdominal organs (such as the liver and gallbladder, and the fetus and amniotic fluid) more clearly distinguishable in the image. It also reduces the interference of magnetic susceptibility artifacts on image quality through magnetic field adjustment. At the same time, this preparation pulse combination is completed before the main sequence starts, without adding extra breath-holding time for the user. It is especially suitable for special groups such as pregnant women who need to shorten the scan time. Ultimately, it provides high-quality image data with better contrast and fewer artifacts for the accurate segmentation and reconstruction of the subsequent three-dimensional model of abdominal organs, significantly improving the recognition accuracy of fine structures (such as the common bile duct and stones) in clinical applications.

[0111] In one embodiment, since MRI acquires K-space data line by line using spatial encoding (phase encoding, frequency encoding, and slice encoding) of gradient magnetic fields, the scan time is proportional to the number of encoding steps. Reducing the number of encoding steps (i.e., "undersampling") can shorten the scan time, but undersampling leads to incomplete K-space data, resulting in aliasing artifacts if reconstructed directly. Spatial sensitivity information of the foot and ankle coils can be obtained in advance through a short calibration pre-scan, and then the undersampled data can be "de-aliased" using the following algorithm to reconstruct an artifact-free image.

[0112] ;

[0113] in, This represents the original signal of the coil. Indicates the first 1 pixel, This represents the sum of pixels within the scanned area. Represents pixels The true signal strength Indicates the first The coil unit for the first... Sensitivity per pixel Indicates noise.

[0114] Specifically, before the main sequence scan, a very short calibration pre-scan is performed, typically within a few seconds. This allows M units of the abdominal coil (e.g., an 8-channel or 16-channel array coil) to simultaneously acquire a small amount of fully sampled reference data (covering the key layers of the area to be scanned; complete coverage is not required, but signal characteristics of all pixels must be included). At this time, multiple units of the abdominal coil (e.g., different channels of the array coil) simultaneously acquire a small amount of fully sampled reference data (covering the key layers of the area to be scanned). By analyzing this reference data, the sensitivity coefficient of each coil unit to N pixels within the scan range can be calculated (i.e., the signal reception sensitivity of the coil unit to that pixel, expressed in matrix form as c, where...). This represents the sensitivity of the m-th coil unit to the n-th pixel. This sensitivity information is stored for subsequent dealiasing.

[0115] During the main sequence scan, the number of coding steps in the phase coding direction and / or inter-slice coding direction is actively reduced (i.e., undersampling), for example, only 50%-70% of the regular coding steps are acquired. Since the MRI scan time is directly related to the number of coding steps, this can significantly shorten the data acquisition time (e.g., from 10 seconds to 5-7 seconds) and reduce radiofrequency energy deposition (SAR value). However, undersampling will result in incomplete K-space data. When directly reconstructing the image through inverse Fourier transform, the signals of adjacent pixels will overlap, forming striped or overlapping artifacts.

[0116] Using the coil spatial sensitivity matrix c obtained in the first step, the undersampled K-space data is processed to eliminate aliasing artifacts. First, an inverse Fourier transform is performed on the undersampled K-space data to obtain the original image containing aliasing artifacts, and the original coil signal is recorded. The abdominal coil's M units synchronously acquire undersampled data, and the raw signal output by each unit is the one in the formula. The signal contains the actual signal of the target pixel, aliasing interference, and noise. (such as electronic noise, thermal noise).

[0117] A mathematical model is constructed based on the coil sensitivity matrix c. This is achieved by solving the inverse problem (e.g., using compressed sensing or parallel imaging algorithms). Separate the real signal of each pixel from the data. This allows the actual signal of each pixel to be extracted individually.

[0118] After dealiasing processing, the true signal intensity of each pixel can be obtained. Then, through conventional image reconstruction procedures (such as Fourier transform and grayscale correction), an aliasing-free magnetic resonance image is generated. This image retains the time advantage of undersampling (scanning time is reduced by 30%-50%), and the missing data is compensated for by coil sensitivity information. The image resolution and detail are consistent with the fully sampled result.

[0119] In this embodiment, by utilizing the spatial sensitivity differences of the abdominal coils, a complete image can be reconstructed while minimizing the amount of data acquired, thereby further shortening the scan time and reducing the SAR value. By "sacrificing a small amount of calibration time" in exchange for "significantly shortening the main sequence time," and by utilizing the inherent spatial sensitivity differences of the coils to solve the artifact problem caused by undersampling, this method is particularly suitable for abdominal scanning. It reduces the breath-holding burden on the user (shortening the scan time) and avoids the image quality degradation caused by undersampling, providing clear original images for subsequent abdominal organ segmentation and 3D modeling. This method can be used simultaneously in both phase encoding and inter-slice encoding directions, greatly reducing the sequence acquisition time and SAR value.

[0120] In one embodiment, step S3 further includes:

[0121] A preset current is applied to the gradient coils in the X, Y, and Z axes.

[0122] Specifically, for scanning ranges, especially those near the top of the diaphragm or where there is interference from gas in the abdominal cavity, appropriate currents can be applied to gradient coils in the X, Y, and Z axes, superimposed on the original main magnetic field, thereby compensating for the inhomogeneity of the original main magnetic field and further reducing the generation of magnetic susceptibility artifacts.

[0123] The specific compensation logic is as follows:

[0124] Inhomogeneities in the main magnetic field can lead to differences in proton resonant frequencies, causing magnetic susceptibility artifacts. Applying current through a gradient coil can generate a linearly varying gradient field, the direction of which is... Conversely, this cancels out the inhomogeneity. For example, if there is a vertical magnetic field drift in the diaphragm top region ( Along the Z-axis, applying a current to the Z-axis gradient coil generates a Z-axis gradient field, causing the magnetic field to change linearly in the Z-axis direction, thus canceling out the current. .

[0125] The gradient field strength is proportional to the coil current. in, For gradient field intensity, The sensitivity coefficient of the gradient coil is determined by the coil design (for example, for a certain device, the Z-axis coil K=2mT / m / A, and the X / Y-axis coil K=1.5mT / m / A). This represents the coil current.

[0126] The specific superposition method and example scenario for calculating the applied current are as follows:

[0127] The problem of uneven compensation of the Z-axis magnetic field in the top region of the diaphragm manifests as follows: respiratory movements cause unevenness of the main magnetic field in the Z-axis direction (vertical direction) of the top region of the diaphragm, resulting in signal loss or distortion.

[0128] First, the inhomogeneity value of the main magnetic field is measured. The inhomogeneity value of the main magnetic field in the diaphragm top region is obtained by field pattern scanning (such as the multi-echo phase difference method). By using the gyromagnetic ratio of hydrogen protons, This is converted into the corresponding resonant frequency shift. ,in, This indicates a shift in the hydrogen proton resonance frequency caused by inhomogeneity in the main magnetic field. The gyromagnetic ratio of hydrogen protons, This indicates the non-uniformity value of the main magnetic field.

[0129] Then calculate the required gradient field, which needs to be compensated. The spatial variation is calculated using the following formula to determine the required gradient field: ,in, Indicates the required gradient field. This indicates the non-uniformity value of the main magnetic field. The gyromagnetic ratio of hydrogen protons, This indicates the scanning range along the Z-axis.

[0130] Finally, calculate the applied current. ,in, Indicates the application of current. Indicates the sensitivity of the Z-axis coil. This represents the desired gradient field.

[0131] After applying current, the Z-axis gradient field ,offset This restores the uniformity of the magnetic field in the diaphragmatic dome region and reduces respiratory artifacts.

[0132] Superimposed current differs from passive shimming (such as applying shimming patches or adjusting patient position) and is more suitable for complex clinical scanning needs. The current is applied through the shimming coil (surrounding the scanning aperture) built into the MRI equipment, without any physical instruments contacting the patient's body.

[0133] The current parameters can be adjusted in real time according to individual patient differences (such as body shape, weight, anatomical variations) or the characteristics of the scanning site (such as the top of the diaphragm, the gas area of ​​the abdominal cavity, and the vicinity of metal implants), and the compensation speed is very fast, thereby making the magnetic field in the scanning area more uniform.

[0134] The magnitude, direction, and area of ​​effect of the current can all be quantitatively adjusted via the MRI console. The strength of the main magnetic field is maintained by a fixed current in the superconducting magnet, and the superimposed current only makes minor corrections to the local magnetic field without affecting the overall strength of the main magnetic field.

[0135] In one embodiment, the image processing device 3 and the magnetic resonance imaging (MRI) scanner 2 are communicatively connected, and acquiring the MRI scan sequence and obtaining a three-dimensional model of the abdominal organs based on the MRI scan sequence includes:

[0136] The intra-abdominal organ tissue segmentation results were obtained based on the magnetic resonance scanning sequence and the automated intra-abdominal organ image segmentation technology based on grayscale features.

[0137] An initial abdominal visceral tissue model is generated based on the segmentation results of the abdominal visceral tissue. The triangular facets are then adjusted using surface simplification coefficients to optimize the initial abdominal visceral tissue model, resulting in an optimized three-dimensional mesh model.

[0138] The optimized 3D mesh model was validated in multiple dimensions and standardized for export to obtain a 3D model of the abdominal organs.

[0139] The intra-abdominal visceral tissue segmentation results obtained based on the magnetic resonance scanning sequence and the automated intra-abdominal visceral image segmentation technology based on grayscale features include:

[0140] Seed points are set in the high-signal, low-signal, and intermediate-signal regions of the abdominal organs, and an initial mask covering the entire area of ​​the abdominal organs is generated by the seed region growth algorithm.

[0141] The initial mask is subjected to boundary contraction to remove the blood vessel wall, leaving the pure parenchyma area, thus obtaining the organ parenchyma;

[0142] For abdominal organs containing fluid, the fluid region is extracted using an adaptive threshold segmentation algorithm;

[0143] For small low-signal abdominal organs, a signal threshold is obtained, and small low-signal structures are extracted based on volume screening conditions and the signal threshold.

[0144] The segmentation result of the abdominal viscera tissue includes the organ parenchyma, the fluid region, and the small low-signal structures. An initial abdominal viscera tissue model is generated based on the segmentation result, and the initial model is optimized by adjusting the triangular facets using surface simplification factors, resulting in an optimized 3D mesh model including:

[0145] Adjust the triangular facets according to the following formula:

[0146] ;

[0147] in, Indicates the number of original mesh patches. This indicates the number of target facets after simplification. This represents the simplification factor.

[0148] The optimized 3D mesh model was validated in multiple dimensions and standardized for export, resulting in a 3D model of the abdominal organs, including:

[0149] The optimized 3D mesh model is validated in two dimensions. The first validation model is obtained when the fit error between the segmented contour and the original image is ≤1 pixel.

[0150] The original image is superimposed on the first verification model by volume rendering, and a three-dimensional model of the abdominal organs is obtained in response to the absence of perforations or missing parts.

[0151] The export formats for the three-dimensional models of abdominal organs include STL, VTK, and DICOM-SEG formats.

[0152] This embodiment uses the medical image processing software 3D Slicer as an example to explain the above embodiments. The operation of this invention includes, but is not limited to, this medical image processing software; any medical image processing software that can achieve this function is acceptable.

[0153] Specifically, firstly, such as Figure 8 As shown, the magnetic resonance imaging (MRI) scan sequence (DICOM data) is imported into 3D Slicer. This sequence possesses the characteristic of "high liquid contrast" (sensitive to fluid signals such as bile and blood), laying the foundation for subsequent grayscale feature-based segmentation. By screening for high liquid contrast sequences of three-dimensional gradient echoes of the target organ, it is ensured that the imported MRI scan sequence is the target high liquid contrast sequence. The MRI scan sequence is then loaded as volume data to obtain three-dimensional data.

[0154] like Figure 9 As shown, to trim the 3D data, the 3D data orientation is uniformly marked into the international standard RAI (Right-Front-Down) coordinate system to ensure spatial positioning consistency; the voxel spacing is strictly verified, requiring ≤1.2mm×1.2mm×1.2mm. If the original slice thickness exceeds 1.5mm, the imported 3D data is spatially trimmed, retaining the region of interest and removing irrelevant parts, and 0.8mm isocuboid voxel resampling is performed. By improving the spatial resolution, the impact of slice thickness differences on the accuracy of subsequent 3D modeling is eliminated, ensuring the preservation of details of the edges of abdominal organs and microstructures (such as the common bile duct and stones).

[0155] After data preprocessing, positive and negative seed points are set in the high signal area, intermediate signal parenchyma area and surrounding low signal background area of ​​the abdominal organs, respectively, by taking advantage of the signal differences of tissues in the high liquid contrast sequence. Then, an initial coarse mask covering the entire range of the abdominal organs is generated by the seed region growth algorithm, which combines tissue signal characteristics and spatial continuity.

[0156] The initial coarse mask is subjected to a boundary contraction operation to remove non-solid structures such as blood vessel walls that may be included at the edge of the mask, while preserving the pure solid areas of the abdominal organs.

[0157] If the abdominal organs contain fluid cavities (such as the cavity structure of fluid-containing organs), the fluid region is extracted using an adaptive threshold segmentation algorithm. The threshold range is [SI_mean-0.5SD, SI_mean+2SD], where SI_mean is the signal mean of the fluid region and SD is the signal standard deviation. The extracted tubular or irregular structural regions are then smoothed using an edge optimization algorithm to eliminate step artifacts and obtain the organ parenchyma.

[0158] For small low-signal structures that may exist within abdominal organs, a specific threshold screening method is used based on their signal contrast with surrounding tissues. The threshold is set to be less than the mean signal value of the normal tissue surrounding the small structure minus 2.5 SD. For candidate voxels that pass the threshold screening, a volume screening operation is performed to retain clinically significant small low-signal structures, and their volume, maximum diameter and other quantitative parameters are automatically calculated. Finally, a complete segmentation result including the parenchyma of abdominal organs, internal fluid and small abnormal structures is obtained.

[0159] Specifically, such as Figure 10 As shown, the above embodiments are illustrated using the liver as an example.

[0160] Create a new segmentation node to integrate the segmentation results of liver-related structures.

[0161] A liver parenchyma segmentation sub-item was added, employing a seed region growth method. Leveraging the sequence's characteristic of high signal intensity for blood flow (portal vein and hepatic vein), intermediate signal intensity for liver parenchyma, and low signal intensity for the background, a multi-region seed point strategy was implemented: 3-5 positive seed points were placed in the main portal vein (high signal area), 2-3 negative seed points were placed in the extrahepatic background (low signal area), and 4-6 core seed points were randomly placed within the liver parenchyma (intermediate signal area). The segmentation range was guided by the signal gradient differences between seed points. An initial mask containing liver parenchyma and blood vessels was generated, achieving automated preliminary localization of the overall liver contour.

[0162] Specifically, such as Figure 11 As shown, the above embodiments are illustrated using liver parenchyma edge optimization and gallbladder-common bile duct separation as examples.

[0163] To address the issue of blood vessels being included in the initial mask, a boundary adjustment operation is performed on the segmentation results (such as the contour mask of organs and lesions). By shrinking the boundary, the blood vessel walls are precisely removed, preserving the pure liver parenchyma region, thus resolving the segmentation ambiguity caused by the overlap of blood vessel and parenchyma signals.

[0164] A new gallbladder bile segmentation sub-item is created. Utilizing the characteristic that bile exhibits significantly high signal intensity and the gallbladder wall shows low signal intensity in the sequence, an adaptive threshold segmentation algorithm is employed: automatic segmentation based on image grayscale values ​​is performed, and the mean (SI_bile_mean) and standard deviation (SD) of the bile signal are obtained by manually drawing ROIs within the gallbladder cavity. The threshold range is dynamically set to [SI_bile_mean – 0.5SD, SI_bile_mean + 2SD]. This threshold range can adapt to the fluctuations in bile signals among different individuals, avoiding the inapplicability of fixed thresholds to signal variations.

[0165] Optimization operations are performed on the scattered small regions that may exist in the segmentation results (from all the segmented regions, only the main region with the largest volume and the best connectivity is retained, and other small regions are automatically deleted), and small volume artifacts (such as interference from nearby high signal blood vessels) are removed through morphological screening to ensure the complete extraction of the gallbladder cavity.

[0166] A new common bile duct segmentation sub-item is created. Taking advantage of the consistency of bile signals between the common bile duct and the gallbladder, 2-3 seed points are placed in the common bile duct region of the gallbladder neck-pancreatic head segment. High-signal bile columns are quickly extracted using a semi-automatic segmentation tool based on a region growing algorithm. The segmentation results (such as the mask of the gallbladder bile region) are smoothed using "median filtering" with a filter kernel size of 1.0 mm to eliminate step artifacts, thus achieving continuous segmentation of the gallbladder and common bile duct.

[0167] Specifically, such as Figure 12 As shown, the above embodiments are illustrated using the automatic detection and marking of gallstones as an example.

[0168] For the stone segmentation sub-item, a specific threshold screening method is used to target the characteristic that stones present a significantly low signal in a high-signal bile background.

[0169] Using a threshold segmentation tool, a low signal threshold was set to < mean bile signal - 2.5 SD (standard deviation), and candidate voxels were extracted by utilizing the significant signal difference between gallstones and bile.

[0170] The process optimizes small, scattered regions that may exist during candidate voxel processing (from all segmented regions, only the N largest and most connected main regions are retained, and other small regions are automatically deleted) (N value is set according to clinical needs, usually 5). By filtering by volume, minute noise artifacts are removed, and clinically significant stones are accurately located.

[0171] By using segmentation statistical tools, quantitative parameters such as volume and maximum diameter are calculated and exported for each selected stone region, providing objective data support for clinical diagnosis.

[0172] Based on the segmentation results of abdominal viscera tissue, an initial 3D mesh model is generated. This model contains all the geometric details of the original image, but often increases the data volume and computational burden due to redundant patches (such as small undulations caused by noise). Figure 13 As shown, the Surface Simplification Factor (SSF) is introduced to adjust the retention ratio of triangle faces in the mesh. This reduces the number of redundant faces while ensuring that key anatomical details (such as liver capsule texture, gallbladder neck and common bile duct connection) are not lost, thereby balancing the model's visualization accuracy and computational efficiency.

[0173] The formula for calculating the surface simplification factor is:

[0174] ;

[0175] in, This represents the original number of mesh patches. To simplify the number of target patches, The value ranges from 0.2 to 0.5. In practical applications, it can be dynamically adjusted according to the complexity of the organ: for detailed regions such as the liver capsule and the gallbladder neck-common bile duct junction, SSF is set to 0.4-0.5 to retain more facets; for simpler regions such as gallstones, SSF is set to 0.2-0.3 to reduce computational load. Through parameterized control, key clinical anatomical features can be accurately preserved while significantly reducing model complexity, thus meeting the efficiency requirements of subsequent visualization analysis and surgical planning.

[0176] like Figure 14As shown, for minor protrusions on the surface of organ parenchyma (caused by noise), a smoothing tool (10-20 iterations) is used for surface smoothing. For liquid regions, multiple independent discrete regions in the segmentation result are merged into a continuous whole to eliminate gaps and ensure continuity. For small low-signal target structures, a simplification tool is used to reduce the number of polygons to simplify the mesh while retaining core morphological features. The optimized model can accurately reflect the spatial relationships between organs, providing a high-fidelity digital model for clinical applications such as 3D printing and surgical navigation.

[0177] like Figure 15 As shown, the optimized three-dimensional mesh model is verified in two dimensions. The segmented contour is compared with the original image layer by layer. The fit error between the segmented contour and the original image is ≤1 pixel, and the first verification model is obtained.

[0178] By overlaying the original image with the 3D model through volumetric rendering, the model is checked for perforations (such as incorrect connections between the liver and diaphragm) or omissions (such as omissions of the liver lobe edge) to ensure the accuracy of the 3D spatial structure and obtain a 3D model of the abdominal organs.

[0179] like Figure 16 As shown, the 3D model of abdominal organs supports exporting to STL format (suitable for 3D printing), VTK format (suitable for mechanical studies such as finite element analysis), and DICOM-SEG format (suitable for clinical system archiving), enabling adaptation to multiple application scenarios.

[0180] In this embodiment, a multi-region seed point strategy is used to precisely combine high, medium, and low signal region features, enabling the rapid generation of an initial mask covering the entire abdominal organs, balancing segmentation efficiency and regional integrity. Precise removal of vessel walls through boundary contraction effectively solves the boundary blurring problem caused by signal overlap between vessels and organ parenchyma, ensuring the purity of the organ parenchyma region. Adaptive threshold segmentation is used for organs containing fluid, dynamically matching individual signal differences and avoiding the limitations of fixed thresholds for fluid region extraction. Furthermore, a strategy for extracting small low-signal structures based on signal thresholds and volume selection accurately captures lesion features while eliminating noise interference. The final integrated segmentation result fully encompasses organ parenchyma, fluid regions, and small abnormal structures, achieving full-level coverage from macroscopic organs to microscopic lesions. The automated process significantly improves segmentation efficiency and objectivity, providing a precise and comprehensive imaging foundation for clinical diagnosis and quantitative analysis.

[0181] The method of optimizing 3D mesh models by adjusting triangular facets using surface simplification factors offers the dual advantages of accuracy and practicality. On one hand, it generates an initial model based on complete segmentation results (including organ parenchyma, fluid regions, and minute low-signal structures), ensuring faithful reproduction of 3D morphology and anatomical features and providing a reliable structural benchmark for subsequent analysis. On the other hand, by selectively adjusting the number of triangular facets using surface simplification factors, it reduces redundant facets while preserving key morphological details (such as organ edges and the outlines of minute lesions), significantly reducing the model's data volume. This optimization ensures both the clarity and spatial accuracy of the model during visualization and improves its efficiency in interactive operations, data transmission, and subsequent applications (such as surgical simulation and 3D printing), achieving a balance between accuracy and lightweight design. It provides high-quality 3D structural support for clinical diagnosis, preoperative planning, and other scenarios.

[0182] Two-dimensional verification rigorously controls the fit error between the segmented contour and the original image (≤1 pixel), ensuring the accuracy of local details in the model from the perspective of tomographic images. Three-dimensional volumetric rendering and overlay verification further checks for perforations or missing parts from the overall structure, ensuring the model fully restores the spatial morphology and anatomical continuity of organs. This dual verification mechanism provides a high-quality three-dimensional structural foundation for subsequent applications. Support for exporting in multiple formats such as STL, VTK, and DICOM-SEG not only meets the needs of engineering applications such as 3D printing and surgical simulation, but also adapts to clinical and research scenarios such as medical image archiving and cross-platform analysis, achieving efficient transformation of high-precision models into practical applications across multiple fields.

[0183] This invention provides a device for acquiring three-dimensional models of abdominal organs. Through multi-region seed point strategies and adaptive threshold segmentation, it achieves precise extraction of organ parenchyma, fluid regions, and minute low-signal structures at all levels, balancing efficiency and integrity. Boundary optimization and surface simplification steps preserve key anatomical details while achieving model lightweighting, balancing accuracy and operational performance. A multi-dimensional verification mechanism (two-dimensional fit control and three-dimensional structural integrity check) ensures model authenticity at every level, while standardized export in multiple formats is adaptable to diverse scenarios such as clinical diagnosis, surgical planning, and 3D printing. The overall process overcomes the inefficiency and subjective limitations of traditional manual segmentation, and ensures the reliability of results through automated algorithms and rigorous verification, providing a highly efficient, accurate, and compatible complete solution for the image analysis and clinical application of abdominal organs.

[0184] It should be noted that the above description describes some embodiments of the present invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than that shown in the above embodiments and still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0185] Based on the same inventive concept, corresponding to the device in any of the above embodiments, the present invention also provides a method for obtaining a three-dimensional model of abdominal organs.

[0186] The method for obtaining a three-dimensional model of abdominal organs includes:

[0187] The coil support assembly is clamped and fixed to the side of the examination bed;

[0188] The user to be tested lies supine on the examination bed, and the abdominal coil is fixed above the abdomen by the coil support assembly, with the abdomen located at the center of the abdominal coil.

[0189] The MRI scanner is positioned above the abdomen of the user to be tested. The axis line of the positioning light is aligned with the midpoint of the axis of the coil support assembly. The MRI scanner is used to obtain the magnetic resonance scan sequence.

[0190] The image processing device is communicatively connected to the magnetic resonance imaging (MRI) scanner to acquire the MRI scan sequence and obtain a three-dimensional model of the abdominal organs based on the MRI scan sequence.

[0191] The step of obtaining a three-dimensional model of the abdominal organs based on the magnetic resonance scanning sequence includes:

[0192] The intra-abdominal organ tissue segmentation results were obtained based on the magnetic resonance scanning sequence and the automated intra-abdominal organ image segmentation technology based on grayscale features.

[0193] An initial abdominal visceral tissue model is generated based on the segmentation results of the abdominal visceral tissue. The triangular facets are then adjusted using surface simplification coefficients to optimize the initial abdominal visceral tissue model, resulting in an optimized three-dimensional mesh model.

[0194] The optimized 3D mesh model was validated in multiple dimensions and standardized for export to obtain a 3D model of the abdominal organs.

[0195] Specifically, the user lies supine on the examination bed with cotton balls or earplugs placed in both ears, feet first, with the long axis of the body aligned with the long axis of the bed surface. Hands are placed naturally at the sides, avoiding crossing the hands and feet to form a loop. An abdominal coil is selected and fixed above the abdomen using the coil support assembly, placing the abdomen in the center of the coil. The gaps inside the abdominal coil are filled with high-density sponge padding to minimize magnetic susceptibility artifacts.

[0196] The MRI scanner is positioned above the abdomen of the user being tested. The axis of the positioning light is aligned with the midpoint of the coil support assembly axis. The MRI scanner is then used to obtain the magnetic resonance scan sequence.

[0197] The image processing equipment communicates with the MRI scanner to acquire MRI scan sequences and obtain three-dimensional models of abdominal organs based on the MRI scan sequences.

[0198] The method described above is used in any of the foregoing embodiments for obtaining a three-dimensional model of an abdominal organ, and has the beneficial effects of the corresponding device embodiments, which will not be repeated here.

[0199] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples; within the framework of the invention, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of the invention as described above, which are not provided in the details for the sake of brevity.

[0200] While specific details have been set forth to describe exemplary embodiments of the invention, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive. Although the invention has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description.

[0201] The embodiments of this invention are intended to cover all such substitutions, modifications, and variations falling within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this invention should be included within the protection scope of this invention.

Claims

1. A device for acquiring a three-dimensional model of abdominal organs, characterized in that, include: Coil support assembly, nuclear magnetic resonance spectrometer and image processing equipment; The coil support assembly is clamped and fixed to the side of the examination bed to support the abdominal coil and reduce pressure on the abdomen of the user being tested. The MRI scanner is positioned above the abdomen of the user being tested, with the axis of the examination positioning light aligned with the midpoint of the axis of the coil support assembly. The MRI scanner is used to obtain a magnetic resonance scanning sequence. The image processing device is communicatively connected to the MRI scanner to obtain the magnetic resonance scanning sequence and to obtain a three-dimensional model of the abdominal organs based on the magnetic resonance scanning sequence. The MRI scanner is positioned above the abdomen of the user being tested. The alignment of the positioning light's axis with the midpoint of the coil support assembly's axis is checked. The MRI scanner is used to obtain a magnetic resonance imaging (MRI) scan sequence, including: Step S1: Use a fast localization scanning sequence to acquire sagittal, coronal, and transverse localization images, and determine the scanning range of the main sequence based on the localization images; Step S2: Define the direction of the main magnetic field as the Z-axis, which is parallel to the long axis of the human body and points to the foot side; the X-axis points to the right side of the human body; and the Y-axis points to the front side of the human body. Set the gradient field spatial encoding with the Z-axis as the layer direction. Set the direction with the smaller value among the anterior-posterior diameter thickness and lateral diameter width of the abdomen as the phase encoding direction, and the direction with the larger value as the frequency encoding direction. Step S3: Based on the gradient field spatial encoding, the scanning range is continuously excited by small-angle radio frequency pulses with a flip angle of 30°-65°, and the readout gradient field is used to quickly switch and acquire echo signals to obtain the raw MR signal data. Step S4: Process the raw MR signal data, and use half-echo acquisition based on Fourier space conjugate symmetry to reconstruct the magnetic resonance scanning sequence. The step of obtaining a three-dimensional model of the abdominal organs based on the magnetic resonance scanning sequence includes: The intra-abdominal organ tissue segmentation results were obtained based on the magnetic resonance scanning sequence and the automated intra-abdominal organ image segmentation technology based on grayscale features. An initial abdominal visceral tissue model is generated based on the segmentation results of the abdominal visceral tissue. The model is then optimized by adjusting the triangular facets using a surface simplification coefficient based on the following formula, resulting in an optimized 3D mesh model: in, Indicates the number of original mesh patches. This indicates the number of target facets after simplification. Represents the simplification coefficient; The optimized 3D mesh model is validated in multiple dimensions and standardized for export to obtain a 3D model of the abdominal organs. The process of obtaining the intra-abdominal visceral tissue segmentation results based on the magnetic resonance scanning sequence and the automated intra-abdominal visceral image segmentation technology based on grayscale features includes: Seed points are set in the high-signal, low-signal, and intermediate-signal regions of the abdominal organs, and an initial mask covering the entire area of ​​the abdominal organs is generated by the seed region growth algorithm. The initial mask is subjected to boundary contraction to remove the blood vessel wall, leaving the pure parenchyma area, thus obtaining the organ parenchyma; For abdominal organs containing fluid, the fluid region is extracted using an adaptive threshold segmentation algorithm; For abdominal organs with small low signal intensity, a signal threshold is obtained, and small low signal intensity structures are extracted based on volume screening conditions and the signal threshold. The intra-abdominal organ tissue segmentation results include the organ parenchyma, the fluid region, and the small low-signal structures.

2. The device for acquiring a three-dimensional model of abdominal organs according to claim 1, characterized in that, The coil support assembly includes a side fixing bracket group, the lower part of which is provided with a fixing clamping mechanism for clamping and fixing the side fixing bracket group to the side of the examination bed; the upper part of the side fixing bracket group is provided with a height adjustment mechanism, the top output end of which is provided with a rotating shaft structure, the rotating shaft structure is connected to an arc-shaped support plate, the arc-shaped support plate is rotatably mounted on the top output end of the height adjustment mechanism through the rotating shaft structure; at least two abdominal coil binding holes are evenly opened on the arc-shaped support plate for passing through the binding strap of the abdominal coil.

3. The device for acquiring a three-dimensional model of abdominal organs according to claim 1, characterized in that, Step S3 further includes: Add a combination of preparatory pulses at 90°, 180°, and -90° before the main sequence.

4. The device for acquiring a three-dimensional model of abdominal organs according to claim 1, characterized in that, Step S3 further includes: A preset current is applied to the gradient coils in the X, Y, and Z axes.

5. The device for acquiring a three-dimensional model of abdominal organs according to claim 1, characterized in that, The process of performing multi-dimensional verification and standardized export of the optimized three-dimensional mesh model to obtain a three-dimensional model of the abdominal organs includes: The optimized 3D mesh model is validated in two dimensions. The first validation model is obtained when the fit error between the segmented contour and the original image is ≤1 pixel. The original image is superimposed on the first verification model by volume rendering, and a three-dimensional model of the abdominal organs is obtained in response to the absence of perforations or missing parts. The export formats for the three-dimensional models of abdominal organs include STL, VTK, and DICOM-SEG formats.

6. A method for obtaining a three-dimensional model of abdominal organs, characterized in that, An apparatus for acquiring a three-dimensional model of an abdominal organ as described in any one of claims 1-5, comprising: The coil support assembly is clamped and fixed to the side of the examination bed; The user to be tested lies supine on the examination bed, and the abdominal coil is fixed above the abdomen by the coil support assembly, with the abdomen located at the center of the abdominal coil. The MRI scanner is positioned above the abdomen of the user to be tested. The axis line of the positioning light is aligned with the midpoint of the axis of the coil support assembly. The MRI scanner is used to obtain the magnetic resonance scan sequence. The image processing device is communicatively connected to the magnetic resonance imaging (MRI) scanner to acquire the MRI scan sequence and obtain a three-dimensional model of the abdominal organs based on the MRI scan sequence. The step of obtaining a three-dimensional model of the abdominal organs based on the magnetic resonance scanning sequence includes: The intra-abdominal organ tissue segmentation results were obtained based on the magnetic resonance scanning sequence and the automated intra-abdominal organ image segmentation technology based on grayscale features. An initial abdominal visceral tissue model is generated based on the segmentation results of the abdominal visceral tissue. The triangular facets are then adjusted using surface simplification coefficients to optimize the initial abdominal visceral tissue model, resulting in an optimized three-dimensional mesh model. The optimized 3D mesh model was validated in multiple dimensions and standardized for export to obtain a 3D model of the abdominal organs.

Images