CN111563953A - Jaw bone defect reconstruction method, device, terminal and medium based on machine learning - Google Patents
Jaw bone defect reconstruction method, device, terminal and medium based on machine learning Download PDFInfo
- Publication number
- CN111563953A CN111563953A CN202010288381.3A CN202010288381A CN111563953A CN 111563953 A CN111563953 A CN 111563953A CN 202010288381 A CN202010288381 A CN 202010288381A CN 111563953 A CN111563953 A CN 111563953A
- Authority
- CN
- China
- Prior art keywords
- jaw bone
- machine learning
- jaw
- data
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
The application provides a jaw bone defect reconstruction method, a jaw bone defect reconstruction device, a jaw bone defect reconstruction terminal and a jaw bone defect reconstruction medium based on machine learning, wherein the jaw bone defect reconstruction method comprises the following steps: collecting jaw CT data of a plurality of sampled persons in a preset crowd range; establishing a plurality of jaw bone feature points characteristic of the upper jaw bone surface and the lower jaw bone surface of each sampled person based on the jaw bone CT data; and obtaining the correlation between the jaw bone characteristic points through a machine learning algorithm so as to reconstruct the jaw bone defect based on the correlation between the jaw bone characteristic points. The jaw bone feature point restoration method based on the machine learning can provide an accurate and personalized scheme in the complex jaw bone reconstruction process by utilizing the machine learning, and solves the clinical problem that the reconstruction of the jaw bone defect cases in a large range across the midrange only depends on experience and no reference. In addition, the jaw bone reconstruction strategy is provided for a specific group in a targeted manner based on jaw bone CT data in a preset crowd range, so that the blood supply of a bone segment is ensured, and the planting site is kept, and the jaw bone reconstruction strategy is closer to the facial appearance of the specific group.
Description
Technical Field
The present application relates to the field of jaw bone defect reconstruction, and in particular, to a jaw bone defect reconstruction method, apparatus, terminal and medium based on machine learning.
Background
The jaw bone is the major bony structure in the upper and lower third of the face, which is essential to the facial appearance and physiological function. How to better reconstruct jaw defects caused by trauma, deformity or tumor surgery is a clinical problem that afflicts oral and maxillofacial surgery and plastic surgeons.
Although the techniques of new biocompatible materials, bone tissue engineering, prosthesis implantation and the like are developed in recent years, the current mainstream repair method is to repair the defective jawbone by autologous bone transplantation using tissue flaps such as fibula, ilium or scapula for vascularization because tissues in the oral cavity are influenced by various factors such as microbial environment, mechanical distribution and the like. When the vascularization free composite tissue flap is transplanted, the shaping and positioning of the bone block are always difficult in the operation, but the specific research of the bone flap shaping method is rarely mentioned in domestic and foreign literature reports. It is common practice to design based on symmetry, however for cases with extensive defects across the midline, it is not possible to design directly by mirror image means due to lack of a healthy side as a reference.
The machine learning algorithm is a research hotspot in the field of computers under the background of big data at present. The method simulates the link mode of a multilayer topological network among human brain neurons by algorithm logic, realizes the layer-by-layer analysis capability of a program on original information, even if a machine has certain learning capability. In the medical field, expert diagnostic systems, disease prognosis models, and digital pathological diagnosis [9, 10] are the latest exploration of the combination of artificial intelligence technology and clinical application background. CampanellaG et al [11] collected pathological imaging data of 1.5 ten thousand patients, trained a data set based on a machine learning algorithm model, and finally constructed a set of pathological artificial intelligence decision-making systems related to prostate cancer, skin cancer and the like. The research result is helpful for relieving the huge reading pressure of clinical pathologists, and simultaneously discloses the application potential of the machine learning algorithm in medical research.
Therefore, how to provide a preoperative design solution suitable for a transmidline and large-scale jaw bone defect by using computer technology is a technical problem to be solved in the field.
Content of application
In view of the above-mentioned drawbacks of the prior art, the present application aims to provide a solution to the technical problem of the prior art that it does not provide a preoperative design solution suitable for a transmidline, large range of jaw defects.
To achieve the above and other related objects, a first aspect of the present application provides a jaw bone defect reconstruction method based on machine learning, including: collecting jaw CT data of a plurality of sampled persons in a preset crowd range; establishing a plurality of jaw bone feature points characteristic of the upper jaw bone surface and the lower jaw bone surface of each sampled person based on the jaw bone CT data; and obtaining the correlation between the jaw bone characteristic points through a machine learning algorithm so as to reconstruct the jaw bone defect based on the correlation between the jaw bone characteristic points.
In some embodiments of the first aspect of the present application, the method of reconstructing a jaw defect further comprises: and carrying out statistical analysis on the jaw bone characteristic points to obtain the difference of the jaw bone characteristic points of different genders on the linear variable and the angle variable respectively.
In some embodiments of the first aspect of the present application, the predetermined population comprises a population meeting a predetermined acquisition requirement; the preset acquisition requirement comprises: any one or more of the same group, the same age group, the same height range and the history of cranialess injury.
In some embodiments of the first aspect of the present application, the three-dimensional coordinate information of all jaw bone feature points of the facial features of the upper and lower jaw bones corresponding to each of the sampled persons is categorized into a group of data; the method comprises the following steps: using a portion of the group data as a training set for training the machine learning algorithm; and taking the residual grouped data as a test set for testing the trained machine learning algorithm, and counting the difference between the output value and the actual value of the machine learning algorithm.
In some embodiments of the first aspect of the present application, the method further comprises: simulating an extreme jaw bone missing condition by using the test set, outputting a coordinate predicted value of a characteristic point corresponding to a jaw bone missing part, and counting the difference between the coordinate predicted value and a coordinate actual value; wherein the extreme mandibular loss condition comprises bilateral complete loss of maxilla or complete loss of mandible.
In some embodiments of the first aspect of the present application, the jaw bone feature points comprise: upper alveolar edge point, bilateral orbital medial point, bilateral orbital lateral point, bilateral zygomatic arch lateral inferior point, bilateral mesiodial alveolar edge point, anterior mental point, bilateral submental point, bilateral mandibular angular point, bilateral condylar lateral point.
To achieve the above and other related objects, a second aspect of the present application provides a machine learning-based jaw bone defect reconstruction device, comprising: the data acquisition module is used for acquiring jaw CT data of a plurality of sampled persons in a preset crowd range; a feature point establishing module for establishing a plurality of jaw bone feature points of the upper jaw bone surface and the lower jaw bone surface of each sampled person based on the jaw bone CT data; and the defect reconstruction module is used for acquiring the correlation among the jaw bone characteristic points through a machine learning algorithm so as to reconstruct the jaw bone defect based on the correlation among the jaw bone characteristic points.
In some embodiments of the second aspect of the present application, the apparatus further comprises: and the statistical analysis module is used for performing statistical analysis on the jaw bone characteristic points so as to acquire the difference of the jaw bone characteristic points of different genders on the linear variable and the angle variable respectively.
To achieve the above and other related objects, a third aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the machine-learning based jaw bone defect reconstruction method.
To achieve the above and other related objects, a fourth aspect of the present application provides an electronic terminal comprising: a processor and a memory; the memory is used for storing computer programs, and the processor is used for executing the computer programs stored by the memory so as to enable the terminal to execute the jaw bone defect reconstruction method based on machine learning.
As described above, the machine learning-based jaw bone defect reconstruction method, apparatus, terminal and medium of the present application have the following beneficial effects: the jaw bone feature point restoration method based on the machine learning can provide an accurate and personalized scheme in the complex jaw bone reconstruction process by utilizing the machine learning, and solves the clinical problem that the reconstruction of the jaw bone defect cases in a large range across the midrange only depends on experience and no reference. In addition, the jaw bone reconstruction strategy is provided for a specific group in a targeted manner based on jaw bone CT data in a preset crowd range, so that the blood supply of a bone segment is ensured, and the planting site is kept, and the jaw bone reconstruction strategy is closer to the facial appearance of the specific group.
Drawings
Fig. 1 is a flowchart illustrating a jaw defect reconstruction method based on machine learning according to an embodiment of the present application.
Fig. 2 is a schematic layout diagram of jaw bone feature points according to an embodiment of the present application.
Fig. 3 is a schematic structural diagram of a jaw defect reconstruction device based on machine learning according to an embodiment of the present application.
Fig. 4 is a schematic structural diagram of an electronic terminal according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application is provided by way of specific examples, and other advantages and effects of the present application will be readily apparent to those skilled in the art from the disclosure herein. The present application is capable of other and different embodiments and its several details are capable of modifications and/or changes in various respects, all without departing from the spirit of the present application. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
It is noted that in the following description, reference is made to the accompanying drawings which illustrate several embodiments of the present application. It is to be understood that other embodiments may be utilized and that mechanical, structural, electrical, and operational changes may be made without departing from the spirit and scope of the present application. The following detailed description is not to be taken in a limiting sense, and the scope of embodiments of the present application is defined only by the claims of the issued patent. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. Spatially relative terms, such as "upper," "lower," "left," "right," "lower," "below," "lower," "above," "upper," and the like, may be used herein to facilitate describing one element or feature's relationship to another element or feature as illustrated in the figures.
In this application, unless expressly stated or limited otherwise, the terms "mounted," "connected," "secured," "retained," and the like are to be construed broadly and can, for example, be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meaning of the above terms in the present application can be understood by those of ordinary skill in the art as appropriate.
Also, as used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context indicates otherwise. It will be further understood that the terms "comprises," "comprising," and/or "comprising," when used in this specification, specify the presence of stated features, operations, elements, components, items, species, and/or groups, but do not preclude the presence, or addition of one or more other features, operations, elements, components, items, species, and/or groups thereof. The terms "or" and/or "as used herein are to be construed as inclusive or meaning any one or any combination. Thus, "A, B or C" or "A, B and/or C" means "any of the following: a; b; c; a and B; a and C; b and C; A. b and C ". An exception to this definition will occur only when a combination of elements, functions or operations are inherently mutually exclusive in some way.
Currently, there are several general methods for preoperative planning of digital jaw reconstruction: (1) the method is used for small-range bone defects, can be used for fine modification of local defects, but still has high subjectivity ratio and limited applicability; (2) model surgery, namely, a surgeon can intuitively simulate various surgical operations on a 3D printing model and change a design scheme in time, but a mature software system is not available at present to well link the operation with CT, MR and other examination image data, and additional material cost also exists; (3) mirror image technology, namely, mirror image technology is used for carrying out the case of jaw reconstruction after the removal of huge mandibular amelogenesis cytoma, thereby achieving good repairing effect. However, it is undeniable that defects in the jaw bone near the midline region have a more harsh appearance, but the mirror image technique is also difficult to benefit from due to the lack of a healthy side as a reference; (4) the database matching method is that a CT image database of the skull of a female in northern China is established, and a data set which is most similar to the skull structure of a patient is searched and adopted to provide a design basis for a maxillofacial bone plastic surgery scheme. Under the support of a relatively large-scale sample, the method is probably the design method most close to the personalized requirement, but the method still needs to undergo long and tedious early accumulation to construct a database, and the regression equations for calculating all parameters are more, so the clinical practicability is poor.
In view of the above, the invention is based on the "four-segment type" jaw bone reconstruction principle, and combines the machine learning algorithm to measure and study the correlation of the upper and lower jaw bone surface feature points, and provides a preoperative technical scheme suitable for the transmidline and large-range jaw bone defects. By measuring jaw key feature point data of a certain group, the internal relation of each feature point between the upper jaw and the lower jaw is calculated and analyzed, reference is provided for the personalized reconstruction of the cross-midline jaw defects, blood supply of bone segments is guaranteed, and the implant site is kept and simultaneously the bone segments are closer to the facial appearance of the group.
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions in the embodiments of the present invention are further described in detail by the following embodiments in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The first embodiment is as follows:
as shown in fig. 1, a flow chart of a jaw defect reconstruction method based on machine learning in an embodiment of the present invention is shown, and the jaw defect reconstruction method of the present embodiment includes the following steps.
It should be understood that the jaw bone defect reconstruction method based on machine learning provided by the present embodiment can be applied to hardware devices such as a controller, a personal computer or a server. Such as an ARM (advanced RISC machines) controller, an FPGA (field Programmable Gate array) controller, an SoC (System on chip) controller, a DSP (digital Signal processing) controller, or an MCU (micro controller Unit) controller; the Personal computer is, for example, a desktop computer, a notebook computer, a tablet computer, a smart phone, a smart television, a Personal Digital Assistant (PDA for short), or the like; the server may be arranged on one or more entity servers according to various factors such as functions, loads, and the like, or may be formed by a distributed or centralized server cluster, which is not limited in this embodiment.
In step S11, jaw CT data of a plurality of sampled persons within a preset population range is acquired.
Preferably, the predetermined population range referred to in this embodiment includes the population meeting the predetermined acquisition requirement. The preset acquisition requirements include, but are not limited to: the same family, the same age group, the same height range, no history of skull injury and the like. It should be understood that, in the present embodiment, the sampled persons within the preset crowd range are selected to obtain more similar jawbone CT data, so that the collected jawbone CT data has a better reference meaning.
For example, 111 normal Han adult skull CT data were collected for the same hospital visit, including 43 males and 68 females, with an average age of 24.3 years and an age range of 24.3 + -6.1 years. All sampled persons meet the preset requirements of jaw data acquisition: the appearance of the face is natural and symmetrical, the dentition in the mouth is complete without crowding, the first molar teeth on two sides are in neutral occlusion relation, the sagittal facial appearance accords with the type I bone-face type, no obvious history of maxillofacial trauma exists, and patients who deny the history of orthodontic treatment are denied. The above examples are for illustration only and are not intended to limit the scope of the present invention.
In an optional implementation manner of this embodiment, a spiral CT machine is used for acquiring images of the skull, and the scan parameters may be set as follows: bulb voltage: 120kV, bulb current: 60mA, the thickness of the scanning layer: 0.625 mm; the above parameters are only referenced and are not intended to limit the scope of the present invention; the skull CT data collected in this embodiment may be in a digital imaging and Communications in medicine format, i.e., a data format for digital medical imaging and Communications, but the data format is not limited in this embodiment.
In step S12, based on the jaw CT data, a plurality of jaw characteristic points characteristic of the maxilla surface and the mandible surface of each of the sampled persons are established.
In an optional implementation manner of the embodiment, the jaw CT data is imported into Proplan CMF 3.0 digital design software for computer three-dimensional modeling and image segmentation. Based on the theory of 'four-section jaw repair and reconstruction', the characteristic anatomical landmark points of the surfaces of the upper jaw and the lower jaw are established.
The present embodiment relates to 16 anatomical landmark points (9 maxilla and 7 mandible) in total, and the anatomical landmark points are respectively marked as shown in fig. 2: upper alveolar ridge point a (maxillary central incisor alveolar ridge vertex), right intraorbital side point B1 (intersection of right intraorbital side wall and bottom wall), left intraorbital side point B2 (intersection of left intraorbital side wall and bottom wall), right intraorbital side point C1 (intersection of right intraorbital side wall and bottom wall), left intraorbital side point C2 (intersection of left intraorbital side wall and bottom wall), right zygomatic arch outer lower point D1 (lowermost point of right zygomatic suture), left zygomatic arch outer lower point D2 (lowermost point of left zygomatic suture), right mesial alveolar edge point E1 (mesial alveolar ridge vertex of right maxillary 6 teeth), left mesial alveolar edge point E2 (mesial alveolar ridge vertex of left maxillary 6 teeth), anterior mandibular point a (mesial edge point), right chin side chin edge point B1 (intersection of perpendicular bisector of mandibular intersection angle long axis of mandible 3, 4 and lower edge of left maxillary), left side chin edge point B2 (2), 4 intersection point of vertical bisector of long axis intersection angle of tooth body and inferior maxilla, right mandibular angle point c1 (intersection point of angular bisector of inferior maxilla inferior border and tangent of ascending branch and inferior maxilla), left mandibular angle point c2 (intersection point of inferior maxilla inferior border and tangent of ascending branch and inferior maxilla inferior border), right condylar lateral point d1 (right lateral condylar outmost point), and left condylar lateral point d2 (left lateral condylar outmost point).
The right ear point PoR, the left ear point PoL, the right orbital point OrR and the left orbital point OrL in the figure are used for fitting and determining a Frankfurt plane, the head position correction of the model is completed by marking the Frankfurt plane, then the 16 jaw bone characteristic points are comprehensively positioned and marked on the reconstructed 3D head bone model from a plurality of two-dimensional tomographic images of a horizontal plane, a coronal plane and a sagittal plane, and finally the characteristic point coordinate information of each virtual jaw bone model is derived for algorithm analysis through the coordinate extraction function of software. In the Proplan CMF 3.0 software, distance and angle between characteristic points are measured for the maxilla and mandible of all persons to be sampled, and the left and right measurements are averaged for variables with symmetry (such as AB1 and AB 2). Statistical analysis can be performed using MATLAB software, and the results of t-test are significant with differences of P < 0.05.
In step S13, a correlation between the jaw bone feature points is obtained through a machine learning algorithm to perform jaw bone defect reconstruction based on the correlation between the jaw bone feature points.
It should be understood that the machine learning algorithm related to this embodiment is a research hotspot in the field of computers in the context of big data at present, and it uses an algorithm logic to simulate a link mode of a multilayer topology network between neurons of a human brain, so as to implement a layer-by-layer analysis capability of a program on original information, even if the machine has a certain learning capability. Specifically, a supervised machine learning algorithm, an unsupervised machine learning algorithm, a semi-supervised machine learning algorithm, or a reinforcement learning machine learning algorithm may be used, and the embodiment is not limited.
In an implementation manner of this embodiment, the three-dimensional coordinate information of all jaw bone feature points of the upper and lower jaw bone surface features corresponding to each sampled person is categorized into a group of data; the method comprises the following steps: using a portion of the group data as a training set for training the machine learning algorithm; and taking the residual grouped data as a test set for testing the trained machine learning algorithm, and counting the difference between the output value and the actual value of the machine learning algorithm.
Preferably, the test set can be used to simulate an extreme jaw bone missing condition, output a coordinate predicted value of a feature point corresponding to the jaw bone missing part, and count a difference between the coordinate predicted value and a coordinate actual value; wherein the extreme mandibular loss condition comprises bilateral complete loss of maxilla or complete loss of mandible.
Specifically, taking 111 collected cases as an example of normal han adult skull CT data which are treated in a certain hospital, three-dimensional coordinate information N (Xi, Yi, Zi), i being 1, 2, … …, 16 of each group of 16 jaw bone feature points is imported into a self-programming machine learning algorithm environment; training of a machine learning algorithm is performed based on the principle of vector similarity matching by regarding the 16 × 3-48 specific three-dimensional coordinate values as a high-dimensional vector. First, 94 groups of data (85% of total cases) out of 111 data groups were randomly selected as a "training set", and the other 17 groups of data (15% of total cases) were selected as a "test set". Inputting the training set into a compiled machine learning algorithm to obtain respective jaw bone characteristic point coordinate data sets of male and female characters; and then testing the reserved test set. Two extreme cases of defects were simulated, namely bilateral complete loss of the maxilla or complete loss of the mandible. And respectively inputting the coordinate information of 7 lower jaws or 9 upper jaws feature points into an algorithm, obtaining a feature point coordinate predicted value corresponding to a jaw missing part after vectorized matching operation, and finally counting the error distance between the feature point coordinate predicted value and an actual coordinate.
In this embodiment, after performing optimization training on the machine learning algorithm through 94 sets of male and female jaw bone feature point data, another 17 sets of test set data are substituted, the case that all the feature points of the upper and lower half jaw bones are missing is predicted, the distance error between the predicted coordinate and the original coordinate of each set of 16 feature points is calculated, and the total average distribution of the prediction errors of each point is 3.452 ± 0.727 mm. Wherein the precision of the near-middle alveolar crest points E1 and E2 of the first two-sided molar is the highest (the average error value is 2.382 +/-0.839 mm), and the precision of the corner points c1 and c2 of the second two-sided mandible is the lowest (the average error value is 4.475 +/-1.507 mm).
To verify the effectiveness of the machine learning-based jaw bone defect reconstruction method of the present embodiment, a real case is hereinafter demonstrated. The diagnosis case of a 38-year-old male is examined before operation: 46 to 32 tooth bodies are absent, alveolar ridges in related areas are low and flat, and mandible on the buccal side has obvious swelling feeling; CT suggests the occupation of the right mandible and the lesion over the midline affects the contralateral anterior dental area. Considered as "postoperative recurrence of ameloblastoma of the right mandible. Therefore, by using the jaw bone defect reconstruction method provided by the embodiment, 3D modeling is performed through ProplanCMF 3.0, segmental resection of the mandible from the lower right 7 near middle to the lower left 4 near middle is planned, and contemporaneous fibula repair is performed; the coordinate information of the healthy maxillary feature points of the case is introduced into a machine learning algorithm, and the coordinate parameters of the missing points a, b1 and b2 are obtained through matching prediction, so that preoperative design is performed.
In one embodiment, the method for reconstructing a defect of a jaw bone further comprises: and carrying out statistical analysis on the jaw bone characteristic points to obtain the difference of the jaw bone characteristic points of different genders on the linear variable and the angle variable respectively.
Specifically, the three-dimensional reconstruction is performed on the collected 111 normal adult skull CT images, and the measurement is performed on the virtual jaw bone model, so that the result shows that the jaw bone feature points of different genders have obvious statistical difference (P <0.05) in linear variables, and in terms of angle variables, the jaw bone feature points of different genders have no obvious gender difference among the angle variables except for the jaw bone angle b1ab2 (male average value 136.06 °, female average value 132.18 °, P ═ 0.003< 0.05). The jaw angle < b1ab2 is an angle formed by a right side submental point b1, a middle submental point a and a left side submental point b 2.
In order to facilitate understanding by those skilled in the art, the following description will be made in conjunction with tables for measuring jaw bone variables of males and females. Table 1 shows the measured values of the linear variables of the jaws of men and women, and table 2 shows the measured values of the angular variables of the jaws of men and women.
TABLE 1
TABLE 2
Therefore, in the embodiment, the machine learning algorithm is used for analyzing and researching the appearance characteristics of the normal jaw bone, and for clinical observation that male and female jaw bones have difference in appearance size, jaw bone characteristic data are firstly measured in groups according to gender so as to check the influence degree of the difference. The results of measurement and analysis indicate that the jaw bones of men and women do have statistical difference in dimension, but the jaw bones of men and women do not have significant gender correlation on the angle parameters except for the angle b1ab2, namely, the shape difference of the jaw bones with different genders is mainly reflected in the difference of small dimensions without the heterogeneity of obvious structural curvature.
Therefore, in the embodiment, the personalized features with complex jaw bone appearance are simplified into the three-dimensional coordinate set of 16 groups of points, the three-dimensional coordinate set is regarded as 1 group of high-dimensional vectors, scaling correction is performed through similar vector modulo processing, then the hundreds of groups of data are subjected to learning training by using a machine learning algorithm, the spatial distribution rule among the jaw bone appearance feature points is obtained, and the preoperative scheme with complex defects is planned.
In summary, the jaw bone defect reconstruction method based on machine learning provided by the embodiment utilizes machine learning to restore jaw bone feature points, can provide an accurate and personalized scheme in a complex jaw bone reconstruction process, and solves the clinical problem that the reconstruction of a large-range jaw bone defect case across a midline can be relied on only by experience and without reference.
Example two:
fig. 3 is a schematic structural diagram of a jaw defect reconstruction device based on machine learning according to an embodiment of the present invention. The jaw bone defect reconstruction apparatus of the present embodiment includes a data acquisition module 31, a feature point establishing module 32, and a defect reconstruction module 33.
The data acquisition module 31 is used for acquiring jawbone CT data of a plurality of sampled persons in a preset crowd range; the feature point establishing module 32 is configured to establish a plurality of jaw bone feature points characteristic of the maxilla surface and the mandible surface of each of the sampled persons based on the jaw bone CT data; the defect reconstruction module 33 is configured to obtain a correlation between the jaw feature points through a machine learning algorithm, so as to perform jaw defect reconstruction based on the correlation between the jaw feature points.
In an optional implementation manner of this embodiment, the jaw defect reconstruction apparatus further includes a statistical analysis module 34, which is configured to perform statistical analysis on the jaw feature points to obtain differences between the jaw feature points of different genders in the linear variable and the angular variable respectively.
It should be noted that the jaw bone defect reconstruction device in this embodiment is similar to the jaw bone defect reconstruction method in the previous embodiment, and therefore, the description thereof is omitted. In addition, it should be understood that the division of each module of the above apparatus is only a logical division, and the actual implementation may be wholly or partially integrated into one physical entity, or may be physically separated. And these modules can be realized in the form of software called by processing element; or may be implemented entirely in hardware; and part of the modules can be realized in the form of calling software by the processing element, and part of the modules can be realized in the form of hardware. For example, the data acquisition module may be a processing element separately set up, or may be implemented by being integrated in a chip of the apparatus, or may be stored in a memory of the apparatus in the form of program code, and a processing element of the apparatus calls and executes the functions of the data acquisition module. Other modules are implemented similarly. In addition, all or part of the modules can be integrated together or can be independently realized. The processing element described herein may be an integrated circuit having signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.
For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when one of the above modules is implemented in the form of a Processing element scheduler code, the Processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
Example three:
fig. 4 is a schematic structural diagram of an electronic terminal according to an embodiment of the present invention. This example provides an electronic terminal, includes: a processor 41, a memory 42, a communicator 43; the memory 42 is connected to the processor 41 and the communicator 43 through a system bus and performs communication with each other, the memory 42 is used for storing computer programs, the communicator 43 is used for communicating with other devices, and the processor 41 is used for operating the computer programs, so that the electronic terminal performs the steps of the jaw bone defect reconstruction method based on machine learning.
The above-mentioned system bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The system bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus. The communication interface is used for realizing communication between the database access device and other equipment (such as a client, a read-write library and a read-only library). The Memory may include a Random Access Memory (RAM), and may further include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory.
The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the device can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component.
Example four:
the present embodiment provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the machine learning-based jaw bone defect reconstruction method in the foregoing embodiments.
Those of ordinary skill in the art will understand that: all or part of the steps for implementing the above method embodiments may be performed by hardware associated with a computer program. The aforementioned computer program may be stored in a computer readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
In summary, the present application provides a jaw bone defect reconstruction method, apparatus, terminal and medium based on machine learning, the present invention utilizes machine learning to perform jaw bone feature point reduction, can provide an accurate and personalized scheme in a complex jaw bone reconstruction process, and solves the clinical problem that cross-centerline large-scale jaw bone defect case reconstruction can be relied on only by experience and without reference. In addition, the jaw bone reconstruction strategy is provided for a specific group in a targeted manner based on jaw bone CT data in a preset crowd range, so that the blood supply of a bone segment is ensured, and the planting site is kept, and the jaw bone reconstruction strategy is closer to the facial appearance of the specific group. Therefore, the application effectively overcomes various defects in the prior art and has high industrial utilization value.
The above embodiments are merely illustrative of the principles and utilities of the present application and are not intended to limit the application. Any person skilled in the art can modify or change the above-described embodiments without departing from the spirit and scope of the present application. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical concepts disclosed in the present application shall be covered by the claims of the present application.
Claims (10)
1. A jaw bone defect reconstruction method based on machine learning is characterized by comprising the following steps:
collecting jaw CT data of a plurality of sampled persons in a preset crowd range;
establishing a plurality of jaw bone feature points characteristic of the upper jaw bone surface and the lower jaw bone surface of each sampled person based on the jaw bone CT data;
and obtaining the correlation between the jaw bone characteristic points through a machine learning algorithm so as to reconstruct the jaw bone defect based on the correlation between the jaw bone characteristic points.
2. The method of claim 1, wherein the jaw bone defect reconstruction method further comprises:
and carrying out statistical analysis on the jaw bone characteristic points to obtain the difference of the jaw bone characteristic points of different genders on the linear variable and the angle variable respectively.
3. The method of claim 1, wherein the predetermined population comprises a population meeting predetermined collection requirements; the preset acquisition requirement comprises: any one or more of the same group, the same age group, the same height range and the history of cranialess injury.
4. The method according to claim 1, wherein three-dimensional coordinate information of all jaw bone feature points of the upper and lower jaw bone surface features corresponding to each of the sampled persons is classified into a group of data; the method comprises the following steps:
using a portion of the group data as a training set for training the machine learning algorithm;
and taking the residual grouped data as a test set for testing the trained machine learning algorithm, and counting the difference between the output value and the actual value of the machine learning algorithm.
5. The method of claim 4, further comprising:
simulating an extreme jaw bone missing condition by using the test set, outputting a coordinate predicted value of a characteristic point corresponding to a jaw bone missing part, and counting the difference between the coordinate predicted value and a coordinate actual value; wherein the extreme mandibular loss condition comprises bilateral complete loss of maxilla or complete loss of mandible.
6. The method of claim 1, wherein the jaw bone feature points comprise: upper alveolar edge point, bilateral orbital medial point, bilateral orbital lateral point, bilateral zygomatic arch lateral inferior point, bilateral mesiodial alveolar edge point, anterior mental point, bilateral submental point, bilateral mandibular angular point, bilateral condylar lateral point.
7. A jaw bone defect reconstruction device based on machine learning, comprising:
the data acquisition module is used for acquiring jaw CT data of a plurality of sampled persons in a preset crowd range;
a feature point establishing module for establishing a plurality of jaw bone feature points of the upper jaw bone surface and the lower jaw bone surface of each sampled person based on the jaw bone CT data;
and the defect reconstruction module is used for acquiring the correlation among the jaw bone characteristic points through a machine learning algorithm so as to reconstruct the jaw bone defect based on the correlation among the jaw bone characteristic points.
8. The apparatus of claim 7, further comprising:
and the statistical analysis module is used for performing statistical analysis on the jaw bone characteristic points so as to acquire the difference of the jaw bone characteristic points of different genders on the linear variable and the angle variable respectively.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the machine learning-based jaw bone defect reconstruction method according to any one of claims 1 to 7.
10. An electronic terminal, comprising: a processor and a memory;
the memory is used for storing a computer program;
the processor is configured to execute the memory-stored computer program to cause the terminal to perform the machine learning-based jaw bone defect reconstruction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010288381.3A CN111563953A (en) | 2020-04-14 | 2020-04-14 | Jaw bone defect reconstruction method, device, terminal and medium based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010288381.3A CN111563953A (en) | 2020-04-14 | 2020-04-14 | Jaw bone defect reconstruction method, device, terminal and medium based on machine learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111563953A true CN111563953A (en) | 2020-08-21 |
Family
ID=72072996
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010288381.3A Pending CN111563953A (en) | 2020-04-14 | 2020-04-14 | Jaw bone defect reconstruction method, device, terminal and medium based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111563953A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113133802A (en) * | 2021-04-20 | 2021-07-20 | 四川大学 | Bone surgery line automatic positioning method based on machine learning |
| CN114261095A (en) * | 2022-03-03 | 2022-04-01 | 西安博恩生物科技有限公司 | AI-based orthopedic 3D printing method and device |
| CN114550278A (en) * | 2022-04-28 | 2022-05-27 | 中汽研汽车检验中心（天津）有限公司 | Method, equipment and storage medium for determining head and face feature point positions of collision dummy |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103919631A (en) * | 2014-04-28 | 2014-07-16 | 四川大学 | Manufacturing method for jaw defect individual restoration |
| CN108829232A (en) * | 2018-04-26 | 2018-11-16 | 深圳市深晓科技有限公司 | The acquisition methods of skeleton artis three-dimensional coordinate based on deep learning |
| CN108920893A (en) * | 2018-09-06 | 2018-11-30 | 南京医科大学 | A kind of cranio-maxillofacial bone and soft tissue form prediction method based on artificial intelligence |
-
2020
- 2020-04-14 CN CN202010288381.3A patent/CN111563953A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103919631A (en) * | 2014-04-28 | 2014-07-16 | 四川大学 | Manufacturing method for jaw defect individual restoration |
| CN108829232A (en) * | 2018-04-26 | 2018-11-16 | 深圳市深晓科技有限公司 | The acquisition methods of skeleton artis three-dimensional coordinate based on deep learning |
| CN108920893A (en) * | 2018-09-06 | 2018-11-30 | 南京医科大学 | A kind of cranio-maxillofacial bone and soft tissue form prediction method based on artificial intelligence |
Non-Patent Citations (3)
| Title |
|---|
| 刘剑楠: "基于机器学习的颌骨解剖标记点相关性分析及临床应用", 《2019第一届全国口腔颌面头颈肿瘤学术大会论文汇编》 * |
| 李佐军: "《大数据的架构技术与应用实践的探究》", 30 April 2019 * |
| 李建军 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113133802A (en) * | 2021-04-20 | 2021-07-20 | 四川大学 | Bone surgery line automatic positioning method based on machine learning |
| CN114261095A (en) * | 2022-03-03 | 2022-04-01 | 西安博恩生物科技有限公司 | AI-based orthopedic 3D printing method and device |
| CN114261095B (en) * | 2022-03-03 | 2022-05-27 | 西安博恩生物科技有限公司 | AI-based orthopedic 3D printing method and device |
| CN114550278A (en) * | 2022-04-28 | 2022-05-27 | 中汽研汽车检验中心（天津）有限公司 | Method, equipment and storage medium for determining head and face feature point positions of collision dummy |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Singare et al. | Rapid prototyping assisted surgery planning and custom implant design | |
| Schendel et al. | 3-dimensional facial simulation in orthognathic surgery: is it accurate? | |
| CN111563953A (en) | Jaw bone defect reconstruction method, device, terminal and medium based on machine learning | |
| CN109767841B (en) | Similar model retrieval method and device based on craniomaxillofacial three-dimensional morphological database | |
| Tilotta et al. | Construction and analysis of a head CT-scan database for craniofacial reconstruction | |
| Nakao et al. | Automated planning with multivariate shape descriptors for fibular transfer in mandibular reconstruction | |
| Yuan et al. | Personalized design technique for the dental occlusal surface based on conditional generative adversarial networks | |
| Chougule et al. | Clinical case study: spine modeling for minimum invasive spine surgeries (MISS) using rapid prototyping | |
| Vitković et al. | The parametric model of the human mandible coronoid process created by method of anatomical features | |
| Stojkovic et al. | Reverse modelling of human long bones using T-splines-case of tibia | |
| Tian et al. | A dual discriminator adversarial learning approach for dental occlusal surface reconstruction | |
| Le et al. | Medical reverse engineering applications and methods | |
| Tian et al. | Efficient tooth gingival margin line reconstruction via adversarial learning | |
| Yuan et al. | Tooth segmentation and gingival tissue deformation framework for 3D orthodontic treatment planning and evaluating | |
| Moncayo-Matute et al. | Description and application of a comprehensive methodology for custom implant design and surgical planning | |
| Li et al. | CT image-based computer-aided system for orbital prosthesis rehabilitation | |
| Nassef et al. | Extraction of human mandible bones from multi-slice computed tomographic data | |
| Xiong et al. | Virtual reconstruction of midfacial bone defect based on generative adversarial network | |
| Mitić et al. | Reconstruction of the missing part in the human mandible | |
| Motie et al. | Application of Artificial Intelligence in Diagnosing Oral and Maxillofacial Lesions, Facial Corrective Surgeries, and Maxillofacial Reconstructive Procedures | |
| Jokanović et al. | State of the Art of Artificial Intelligence in Dentistry and Its Expected Future | |
| Singare et al. | The use of 3D printing technology in human defect reconstruction-a review of cases study | |
| Liu et al. | Study on digital data processing techniques for 3d medical model | |
| Joseph et al. | Three dimensional reconstruction models for medical modalities: a comprehensive investigation and analysis | |
| Chabanas et al. | Computer aided planning for orthognatic surgery |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |