Temporomandibular joint three-dimensional reconstruction system based on multi-modal image fusion

Abstract

The present application relates to the technical field of image analysis, in particular to a kind of temporomandibular joint three-dimensional reconstruction system based on multimodal image fusion, comprising: obtaining the three-dimensional body data of CBCT and MRI two kinds of medical images of temporomandibular joint, determine key point;According to physiological movement process determines small opening process or big opening process;When small opening process, according to the difference between opening degree and preset opening threshold obtains target interval;When big opening process, directly according to the interval of condylar apical point and attachment point under preset opening threshold, and the translation distance in translation process after exceeding preset opening threshold, determine target interval;Finally, according to target interval and model interval, determine whether model reconstruction is needed or not.Compared with prior art, the present application can realize the error analysis of overall physiological movement process by relatively clear point change, so as to obtain more accurate and reliable three-dimensional model.

Description

Technical Field

[0001] This invention relates to the field of image analysis technology, specifically to a three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion. Background Technology

[0002] In the diagnosis and treatment of temporomandibular joint disorders (TMJ) in dentistry, 3D modeling technology can transform 2D images into three-dimensional models, intuitively presenting the spatial morphology and positional relationships of structures such as the glenoid fossa and condyles. Furthermore, 3D models constructed by fusing different types of images can further leverage the advantages of 3D modeling technology in reconstructing the TMJ. Different types of images have different focuses; some excel at capturing the morphological features of subtle structures, while others present functional information more clearly. Fusing these images for 3D modeling can integrate the advantages of various image types, avoiding the limitations of a single image, and enabling a more comprehensive reconstruction of the overall state of the TMJ.

[0003] Based on the core concept of the Joint Kinectics System (JK system), while the 3D model of the temporomandibular joint constructed through image fusion can integrate the advantages of various images and improve its diagnostic and therapeutic reference value, the image alignment before fusion relies on one-to-one corresponding feature point pairs in different images. However, different images have different advantageous regions, and often appear blurry in non-advantageous regions. For example, CBCT shows soft tissue blurring and MRI shows unclear bony edges, thus affecting the alignment accuracy. The blurring between medical image models from different sources leads to mismatch in the overall model matching, resulting in insufficient accuracy in the 3D model construction. Summary of the Invention

[0004] To address the technical problem in related technologies where ambiguity between medical image models from different sources leads to mismatch in the overall model and insufficient accuracy in 3D model construction, this invention provides a 3D reconstruction system for the temporomandibular joint based on multimodal image fusion. The specific technical solution adopted is as follows:

[0005] This invention proposes a three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion, comprising:

[0006] The acquisition module is used to acquire three-dimensional volume data of the temporomandibular joint from CBCT and MRI medical images; and to identify key points in different types of medical images, including: the apex of the condyle, the deepest point of the glenoid fossa, and the attachment point of the posterior band of the articular disc.

[0007] The classification module is used to determine whether a physiological movement process is a small-opening process or a large-opening process based on the comparison between the opening degree during the physiological movement process and the preset opening threshold.

[0008] The first analysis module is used to determine the rotation adjustment coefficient during physiological movement based on the difference between the opening degree and the preset opening threshold during the small opening process; and to dynamically adjust the distance between the condyle apex and the attachment point based on the rotation adjustment coefficient while keeping the deepest point of the glenoid fossa unchanged, so as to obtain the target distance.

[0009] The second analysis module is used to determine the target spacing directly based on the distance between the condyle apex and the attachment point under the preset opening threshold, and the translation distance during the translation process after exceeding the preset opening threshold, during the large opening process.

[0010] The judgment module is used to compare the target spacing with the model spacing of the condyle vertices and attachment points during the 3D model matching process to determine whether model reconstruction is needed.

[0011] Furthermore, the distance from closing to opening during the physiological movement of the temporomandibular joint is taken as the opening degree.

[0012] Furthermore, based on the comparison between the opening degree and the preset opening threshold, it is determined whether the physiological movement process belongs to a small opening process or a large opening process, including:

[0013] When the opening is less than the preset opening threshold, the physiological movement process is determined to be a small opening process; otherwise, the physiological movement process is determined to be a large opening process.

[0014] Furthermore, the preset opening threshold is 20 mm.

[0015] Furthermore, based on the difference between the opening degree and the preset opening threshold, the rotational adjustment coefficient during physiological movement is determined, including:

[0016] The ratio of the opening degree to the preset opening threshold is used as the rotation adjustment coefficient.

[0017] Furthermore, the distance between the condylar apex and the attachment point is dynamically adjusted according to the rotation adjustment coefficient to obtain the target distance, including:

[0018] The distance between the condyle apex and the attachment point when the mouth is closed is used as the initial distance, and the distance under the preset opening threshold is used as the analysis distance.

[0019] The difference between the analysis interval and the initial interval is used as the adjustment interval;

[0020] Calculate the product of the rotation adjustment coefficient and the adjustment spacing, and use the sum of the product and the initial spacing as the target spacing.

[0021] Furthermore, the target spacing is determined directly based on the distance between the condyle apex and the attachment point below a preset opening threshold, and the translation distance during the translation process after exceeding the preset opening threshold, including:

[0022] The sum of the analysis spacing and the translation distance is used as the target spacing.

[0023] Furthermore, by comparing the target spacing with the model spacing between the condyle vertices and attachment points during the 3D model matching process, it is determined whether model reconstruction is necessary, including:

[0024] Based on the difference between the target spacing and the model spacing, the reconstruction judgment index is determined;

[0025] If the reconstruction judgment index is greater than the preset reconstruction threshold, it is determined that model reconstruction should be performed; otherwise, it is determined that model reconstruction should not be performed.

[0026] Furthermore, based on the difference between the target spacing and the model spacing, reconstruction judgment indicators are determined, including:

[0027] Calculate the absolute value of the difference between the target spacing and the model spacing, and use the ratio of the absolute value of the difference to the target spacing as the reconstruction judgment index.

[0028] Furthermore, the preset reconstruction threshold is 0.75.

[0029] The present invention has the following beneficial effects:

[0030] This invention analyzes the overall model construction error using three relatively clear points from CBCT and MRI medical images: the condylar apex, the deepest point of the glenoid fossa, and the attachment point of the posterior glenoid disc band. First, three-dimensional volumetric data of the temporomandibular joint are acquired from CBCT and MRI images, with the condylar apex, the deepest point of the glenoid fossa, and the attachment point of the posterior glenoid disc band as key points. Then, through mouth opening analysis, small and large opening processes are identified, based on each physiological movement process. During the small opening process, the rotation adjustment coefficient during physiological movement is determined based on the difference between the opening degree and a preset opening threshold. With the concaveest point of the glenoid fossa unchanged, the distance between the condylar apex and attachment point is dynamically adjusted according to the rotation adjustment coefficient to obtain the target distance. During the large opening process, the target distance is determined directly based on the distance between the condylar apex and attachment point under the preset opening threshold, and the translation distance during the translation process after exceeding the preset opening threshold. The above classification and discussion through small and large opening processes can more comprehensively restore the overall state of the temporomandibular joint and effectively characterize the morphological features of fine structures. Finally, the target distance is compared with the model distance between the condylar apex and attachment point during the 3D model matching process to determine whether model reconstruction is necessary. Through relatively clear point changes, error analysis of the overall physiological movement process is achieved, thereby obtaining a more accurate and reliable 3D model. Attached Figure Description

[0031] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 A structural diagram of a three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion provided in an embodiment of the present invention;

[0033] Figure 2 This is a schematic diagram of a CBCT medical image provided according to an embodiment of the present invention;

[0034] Figure 3 A schematic diagram of a medical MRI image provided according to an embodiment of the present invention;

[0035] Figure 4 This is a schematic diagram of three key points provided in an embodiment of the present invention: the apex of the condyle, the most concave point of the glenoid fossa, and the attachment point of the posterior band of the articular disc.

[0036] The labels in the diagram are: A: condylar apex; B: deepest point of the glenoid fossa; C: attachment point of the posterior band of the articular disc. Detailed Implementation

[0037] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0039] The following description, in conjunction with the accompanying drawings, details a specific scheme for a three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion provided by this invention.

[0040] Please see Figure 1 The diagram illustrates a three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion according to an embodiment of the present invention. The system includes: an acquisition module 101, a motion classification module 102, a first analysis module 103, a second analysis module 104, and a judgment module 105.

[0041] The acquisition module 101 is used to acquire three-dimensional volume data of the temporomandibular joint from CBCT and MRI medical images; and to determine key points in different medical images, including: the apex of the condyle, the deepest point of the glenoid fossa, and the attachment point of the posterior band of the articular disc.

[0042] In temporomandibular joint (TMJ) diagnosis and treatment in dentistry, 3D modeling technology can transform 2D images into three-dimensional models, intuitively presenting the spatial morphology and positional relationships of structures such as the glenoid fossa and condyles. Furthermore, 3D models constructed by fusing different types of images can further leverage the advantages of 3D modeling technology in reconstructing the TMJ. Different types of images have different focuses; some excel at capturing the morphological features of subtle structures, while others present functional information more clearly. Fusing these images for 3D modeling integrates the advantages of various image types, avoiding the limitations of single images, and enabling a more comprehensive reconstruction of the overall state of the TMJ. This fused 3D model provides a more complete reference for diagnosis, effectively improving the accuracy of TMJ diagnosis and treatment in dentistry.

[0043] While image fusion-based 3D models of the temporomandibular joint (TMJ) can integrate the advantages of various images and enhance their diagnostic and therapeutic value, image alignment before fusion relies on one-to-one feature point pairs from different images. However, different images have different advantageous regions, and feature point pairs in non-advantageous regions are often inaccurately located, affecting alignment accuracy and consequently the accuracy of 3D modeling. Therefore, a method for accuracy verification based on actual movement is urgently needed. This invention addresses this by analyzing the physiological characteristics of the TMJ and the movement process of the joint structure during physiological activities to achieve accuracy detection in 3D modeling, determining whether model reconstruction is necessary to avoid the influence of inaccurate models and improve the reliability of the 3D model.

[0044] This invention proposes a three-dimensional reconstruction system for the temporomandibular joint (TMJ) that integrates different types of images. This system addresses the differences in strengths and weaknesses between CBCT (cone-beam computed tomography) and MRI (magnetic resonance imaging) in specific areas (bone and soft tissue), achieving complementary advantages through the fusion of the two image types. To ensure the accuracy of the three-dimensional modeling, the system incorporates the physiological activity characteristics of the TTMJ to calibrate the feature points used for image alignment, and optimizes the final three-dimensional reconstruction process based on the calibration analysis results.

[0045] The CBCT imaging process involves the subject sitting with their head fixed, wearing an occlusal splint to maintain mandibular stability, and scanning static images in both closed and wide-open positions to ensure clear identification of bony structures such as the condyles and glenoid fossae, and consistent placement of corresponding bony landmarks in different opening positions. High-resolution static data is then acquired to obtain the CBCT medical images. (See also...) Figure 2 , Figure 2 This is a schematic diagram of a CBCT medical image provided in one embodiment of the present invention.

[0046] MRI imaging: The subject lies supine and is fixed inside a coil. They are pre-trained to open and close their mouth at a uniform speed. Dynamic sequences covering the entire opening and closing process are acquired. Images with smooth movement and no artifacts are selected to clearly show the dynamic changes of soft tissues such as the joint disc and synovium, thus obtaining medical MRI images. See also Figure 3 , Figure 3 This is a schematic diagram of a medical MRI image provided according to an embodiment of the present invention.

[0047] Since the original CBCT and MRI images exist in the form of two-dimensional slices, they can only show planar images of local structures and cannot reflect the three-dimensional morphology of the temporomandibular joint in three-dimensional space. Therefore, it is necessary to reconstruct the two-dimensional images into three-dimensional volume data.

[0048] The original CBCT and MRI image sequences were imported into image processing software. First, the data quality was optimized through denoising and boundary enhancement. Then, based on the spatial positioning information of the slices, the two-dimensional slices were stacked into three-dimensional voxel data to construct three-dimensional volumetric models of the two types of images. Preliminary rigid registration was performed using the external auditory canal or nasal root point as a reference, and the coordinate system of the MRI three-dimensional volume data was roughly aligned to the CBCT coordinate system (or vice versa) to ensure that the positions of large structures in the two images are consistent (e.g., the condyle is roughly located in the glenoid fossa, rather than misaligned to other areas).

[0049] CBCT and MRI differ in their visualization of bone and soft tissue regions (CBCT shows blurred soft tissue and MRI shows indistinct bony margins). It is necessary to select key points that are clearly visible in both types of imaging and have clear structural significance as key points. These key points are used to locate the two images in the same space. The following three points are selected as alignment key points (clearly distinguishable in both CBCT and MRI and related to joint movement), see [link to relevant documentation]. Figure 4 , Figure 4 This is a schematic diagram of three key points provided in one embodiment of the present invention: the apex of the condyle, the most concave point of the glenoid fossa, and the attachment point of the posterior band of the articular disc.

[0050] Point A: Condylar apex (the highest bony point of the condyle in CBCT, and the center point of the contact surface between the articular disc and the condyle in MRI).

[0051] Point B: The deepest point of the glenoid fossa (center of the bony depression in the glenoid fossa in CBCT, lowest point of the synovial reflection in MRI).

[0052] Point C: The attachment point of the posterior band of the articular disc (the point of contact between the posterior band of the articular disc and the posterior slope of the condyle in MRI, and the midpoint of the posterior margin of the condyle in CBCT).

[0053] Points A, B, and C are the condyle apex, the deepest point of the glenoid fossa, and the attachment point of the posterior band of the articular disc, respectively. They correspond to the bony structure motion center, the spatial reference of the glenoid fossa, and the area linked with soft tissues, and can all be identified in CBCT and MRI. Subsequently, a three-dimensional spatial registration reference for cross-modal image fusion can be constructed using these three points.

[0054] The motion classification module 102 is used to determine whether a physiological motion process is a small-opening process or a large-opening process based on the comparison between the opening degree during physiological motion and the preset opening threshold.

[0055] In 3D reconstruction of the temporomandibular joint (TMJ) based on multimodal image fusion, the fusion accuracy between different images directly determines the accuracy of the final 3D model. Although the feature points used for alignment are clearly identifiable in static images, unavoidable artifacts during scanning can cause slight deviations in the spatial positioning of these feature points. This potential interference can affect the matching accuracy of bone and soft tissue structures. Therefore, it is necessary to combine the physiological movement characteristics of the TMJ and calibrate key points to improve the stability of multimodal image fusion.

[0056] It should be noted that the opening and closing movements of the temporomandibular joint (TMJ) are a smooth transition from rotation-dominant to translation-dominant. Throughout this process, condylar rotation is dominant during small openings, while translation is dominant during large openings. The physiological movements of the TMJ are continuous rather than discrete phases. The most concave point of the glenoid fossa remains fixed as a reference. During small openings, the apex of the condyle primarily rotates around the most concave point of the glenoid fossa, with the attachment point of the posterior glenoid disc rotating synchronously with the apex. During large openings, the apex of the condyle primarily translates along the slope of the glenoid fossa, and the distance between the attachment point of the posterior glenoid disc and the apex of the condyle stretches with the translation. The entire process smoothly transitions from rotation to translation. For ease of analysis and statistics, it is necessary to specifically distinguish between the small-opening and large-opening processes.

[0057] First, in this embodiment of the invention, the distance from closed to open during the physiological movement of the temporomandibular joint is used as the opening degree. The distance from closed to open is the vertical distance between the incisal edges of the upper and lower central incisors (front teeth).

[0058] Furthermore, in some embodiments of the present invention, determining whether a physiological movement process belongs to a small opening process or a large opening process based on a comparison between the opening degree and a preset opening threshold includes: determining that the physiological movement process belongs to a small opening process when the opening degree is less than the preset opening threshold, and otherwise determining that the physiological movement process belongs to a large opening process.

[0059] The preset opening threshold can be, for example, 20mm. When the opening is less than 20mm, the physiological movement of the temporomandibular joint is mainly the rotation of the condyle apex around the most concave point of the glenoid fossa during the small opening process. The attachment point of the posterior band of the articular disc rotates synchronously with the condyle apex. When the opening is greater than or equal to 20mm, it is a large opening process. The condyle apex mainly translates along the slope of the glenoid fossa, and the distance between the attachment point of the posterior band of the articular disc and the condyle apex is stretched with the translation.

[0060] The small-opening process and the large-opening process will be analyzed in detail in the following sections.

[0061] The first analysis module 103 is used to determine the rotation adjustment coefficient during the physiological movement process based on the difference between the opening degree and the preset opening threshold during the small opening process; and to dynamically adjust the distance between the condyle apex and the attachment point based on the rotation adjustment coefficient while keeping the deepest point of the glenoid fossa unchanged, so as to obtain the target distance.

[0062] Among them, the rotation adjustment coefficient is the adjustment index information of the condyle apex rotating around the concave point of the glenoid fossa. The larger the opening, the larger the angle of rotation of the condyle apex around the concave point of the glenoid fossa. At this time, the distance between the attachment point of the posterior band of the articular disc and the condyle apex increases linearly with the rotation process. Therefore, in this embodiment of the invention, the rotation adjustment coefficient is calculated by combining the characteristics of the dynamic change of temporomandibular joint rotation with the opening: the relationship between the two is close to linear when the opening is small.

[0063] In this embodiment of the invention, the ratio of the opening degree to the preset opening threshold is used as the rotation adjustment coefficient.

[0064] Furthermore, in some embodiments of the present invention, the distance between the condylar apex and the posterior girdle attachment point of the articular disc is dynamically adjusted according to the rotation adjustment coefficient to obtain the target distance, including: taking the distance between the condylar apex and the posterior girdle attachment point of the articular disc when closed as the initial distance, and the distance at a preset opening threshold as the analysis distance; taking the difference between the analysis distance and the initial distance as the adjustment distance; calculating the product of the rotation adjustment coefficient and the adjustment distance, and taking the sum of the product and the initial distance as the target distance.

[0065] Since the distance between the attachment point of the posterior band of the articular disc and the apex of the condyle increases linearly with the rotation process, the adjustment distance can be determined by determining the initial distance and the analysis distance. The initial distance and the analysis distance are fitted by existing data in the actual detection process. Of course, in the embodiments of the present invention, the test subject can also be directly simulated to determine the initial distance and the analysis distance, and there is no limitation on this.

[0066] Then, during the linear fitting process, the product of the rotation adjustment coefficient and the adjustment interval is calculated to obtain the interval value that should be increased under linear fitting. The sum of this interval value and the initial interval is taken as the target interval.

[0067] The second analysis module 104 is used to dynamically adjust the distance between the condyle apex and the attachment point directly based on the rotation adjustment coefficient under the preset opening threshold and the translation distance during the translation process after exceeding the preset opening threshold during the large opening process, so as to obtain the target distance.

[0068] Compared to the small-scale opening process, which is primarily rotational, the large-scale opening process is mainly characterized by translation. During this translation, the dynamic distance between the condylar apex and the attachment point of the posterior band of the articular disc, combined with the physiological movement of the temporomandibular joint, increases due to the influence of the articular disc structure during the large-scale opening phase. However, in reality, because the condylar apex is pulled forward and translated, while the displacement of the attachment point of the posterior band of the articular disc is extremely small and negligible, in this embodiment of the invention, the increased distance between the attachment point of the posterior band of the articular disc and the condylar apex is directly calculated by analyzing the translation distance of the same point during the translation process after exceeding the preset opening threshold. In other words, the sum of the analyzed distance and the translation distance is used as the target distance.

[0069] The judgment module 105 is used to compare the target spacing with the model spacing of the condyle apex and attachment point during the 3D model matching process to determine whether model reconstruction is needed.

[0070] The temporomandibular joint is a continuous whole. By calibrating and aligning three points—the apex of the condyle, the deepest point of the glenoid fossa, and the attachment point of the posterior band of the articular disc—overall deviations between different images can be detected.

[0071] In this embodiment of the invention, the model spacing between the target spacing and the attachment point of the condyle apex and the articular disc posterior band during the matching process of the three-dimensional model is compared to determine whether model reconstruction is required. This includes: determining a reconstruction judgment index based on the difference between the target spacing and the model spacing; determining that model reconstruction is required if the reconstruction judgment index is greater than a preset reconstruction threshold; otherwise, determining that model reconstruction is not required.

[0072] The target spacing represents the relative distance between the attachment point of the posterior band of the articular disc and the apex of the condyle under normal standard conditions. However, in actual 3D model building, it may cause misalignment between CBCT and MRI medical images during 3D model building. For example, in order to align a certain point, the surrounding tissues and bones may be misaligned, resulting in a large overall misalignment of the model, which in turn affects the overall model accuracy.

[0073] Therefore, it is necessary to analyze the identified misalignment through the target spacing. Furthermore, in some embodiments of the present invention, the model spacing between the target spacing and the condyle vertex and attachment point during the 3D model matching process is compared to determine whether model reconstruction is required. This includes: determining a reconstruction judgment index based on the difference between the target spacing and the model spacing; determining that model reconstruction is required if the reconstruction judgment index is greater than a preset reconstruction threshold, otherwise determining that model reconstruction is not required.

[0074] Among them, the reconstruction judgment index represents the judgment value of whether the model needs to be reconstructed, that is, the degree of misalignment of the model itself. Since the target spacing represents the relative distance between the attachment point of the posterior band of the articular disc and the apex of the condyle under normal standard conditions, the greater the difference between the target spacing and the model spacing, the greater the misalignment of the model spacing in point identification, and the insufficient accuracy of the 3D model construction.

[0075] Furthermore, in some embodiments of the present invention, the reconstruction judgment index is determined based on the difference between the target spacing and the model spacing, including: calculating the absolute value of the difference between the target spacing and the model spacing, and using the ratio of the absolute value of the difference to the target spacing as the reconstruction judgment index.

[0076] The difference characteristics are represented by the absolute value of the difference, that is, the absolute value of the difference between the target distance and the model distance is calculated, and the ratio of the absolute value of the difference to the target distance is used as the reconstruction judgment index.

[0077] In this embodiment of the invention, the larger the value of the reconstruction judgment index, the greater the impact of the overall misalignment difference. By standardizing the value of the index with the target distance under the same dimensions, a reconstruction judgment index that can accurately characterize the impact of misalignment difference can be obtained.

[0078] The preset reconstruction threshold is the threshold value of the reconstruction judgment index. In this embodiment of the invention, the preset reconstruction threshold can be, for example, 0.75, or it can be adjusted according to actual needs without limitation.

[0079] This invention analyzes the overall model construction error using three relatively clear points from CBCT and MRI medical images: the condylar apex, the deepest point of the glenoid fossa, and the attachment point of the posterior glenoid disc band. First, three-dimensional volumetric data of the temporomandibular joint are acquired from CBCT and MRI images, with the condylar apex, the deepest point of the glenoid fossa, and the attachment point of the posterior glenoid disc band as key points. Then, through mouth opening analysis, small and large opening processes are identified, based on each physiological movement process. During the small opening process, the rotation adjustment coefficient during physiological movement is determined based on the difference between the opening degree and a preset opening threshold. With the concaveest point of the glenoid fossa unchanged, the distance between the condylar apex and attachment point is dynamically adjusted according to the rotation adjustment coefficient to obtain the target distance. During the large opening process, the target distance is determined directly based on the distance between the condylar apex and attachment point under the preset opening threshold, and the translation distance during the translation process after exceeding the preset opening threshold. The above classification and discussion through small and large opening processes can more comprehensively restore the overall state of the temporomandibular joint and effectively characterize the morphological features of fine structures. Finally, the target distance is compared with the model distance between the condylar apex and attachment point during the 3D model matching process to determine whether model reconstruction is necessary. Through relatively clear point changes, error analysis of the overall physiological movement process is achieved, thereby obtaining a more accurate and reliable 3D model.

[0080] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0081] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion, characterized in that, The system includes: The acquisition module is used to acquire three-dimensional volume data of the temporomandibular joint from CBCT and MRI medical images; and to identify key points in different types of medical images, including: the apex of the condyle, the deepest point of the glenoid fossa, and the attachment point of the posterior band of the articular disc. The classification module is used to determine whether a physiological movement process is a small-opening process or a large-opening process based on the comparison between the opening degree during the physiological movement process and the preset opening threshold. The first analysis module is used to determine the rotation adjustment coefficient during physiological movement based on the difference between the opening degree and the preset opening threshold during the small opening process; and to dynamically adjust the distance between the condyle apex and the attachment point based on the rotation adjustment coefficient while keeping the deepest point of the glenoid fossa unchanged, so as to obtain the target distance. The second analysis module is used to determine the target spacing directly based on the distance between the condyle apex and the attachment point under the preset opening threshold, and the translation distance during the translation process after exceeding the preset opening threshold, during the large opening process. The judgment module is used to compare the target spacing with the model spacing of the condyle vertices and attachment points during the 3D model matching process to determine whether model reconstruction is needed.

2. The temporomandibular joint three-dimensional reconstruction system based on multimodal image fusion as described in claim 1, characterized in that, The distance from closing to opening during the physiological movement of the temporomandibular joint is defined as the opening degree.

3. The temporomandibular joint three-dimensional reconstruction system based on multimodal image fusion as described in claim 1, characterized in that, Based on the comparison between the opening degree and the preset opening threshold, it is determined whether the physiological movement process belongs to a small opening process or a large opening process, including: When the opening is less than the preset opening threshold, the physiological movement process is determined to be a small opening process; otherwise, the physiological movement process is determined to be a large opening process.

4. The three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion as described in claim 3, characterized in that, The preset opening threshold is 20 mm.

5. The temporomandibular joint three-dimensional reconstruction system based on multimodal image fusion as described in claim 1, characterized in that, Based on the difference between the opening degree and the preset opening threshold, the rotational adjustment coefficient during physiological movement is determined, including: The ratio of the opening degree to the preset opening threshold is used as the rotation adjustment coefficient.

6. The temporomandibular joint three-dimensional reconstruction system based on multimodal image fusion as described in claim 1, characterized in that, The distance between the condyle apex and the attachment point is dynamically adjusted based on the rotation adjustment coefficient to obtain the target distance, including: The distance between the condyle apex and the attachment point when the mouth is closed is used as the initial distance, and the distance under the preset opening threshold is used as the analysis distance. The difference between the analysis interval and the initial interval is used as the adjustment interval; Calculate the product of the rotation adjustment coefficient and the adjustment spacing, and use the sum of the product and the initial spacing as the target spacing.

7. The temporomandibular joint three-dimensional reconstruction system based on multimodal image fusion as described in claim 6, characterized in that, The target spacing is determined directly based on the distance between the condyle apex and the attachment point below a preset opening threshold, and the translation distance during the translation process after exceeding the preset opening threshold. This includes: The sum of the analysis spacing and the translation distance is used as the target spacing.

8. The temporomandibular joint three-dimensional reconstruction system based on multimodal image fusion as described in claim 1, characterized in that, By comparing the target spacing with the model spacing of the condyle vertices and attachment points during the 3D model matching process, it is determined whether model reconstruction is necessary, including: Based on the difference between the target spacing and the model spacing, the reconstruction judgment index is determined; If the reconstruction judgment index is greater than the preset reconstruction threshold, it is determined that model reconstruction should be performed; otherwise, it is determined that model reconstruction should not be performed.

9. The three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion as described in claim 8, characterized in that, Based on the difference between the target spacing and the model spacing, reconstruction judgment indicators are determined, including: Calculate the absolute value of the difference between the target spacing and the model spacing, and use the ratio of the absolute value of the difference to the target spacing as the reconstruction judgment index.

10. A three-dimensional reconstruction system for the temporomandibular joint based on multimodal image fusion as described in claim 8, characterized in that, The preset reconstruction threshold is 0.75.

Images