CN117954051A - Method, device and storage medium for acquiring dental defect treatment scheme based on artificial intelligence - Google Patents
Method, device and storage medium for acquiring dental defect treatment scheme based on artificial intelligence Download PDFInfo
- Publication number
- CN117954051A CN117954051A CN202410347088.8A CN202410347088A CN117954051A CN 117954051 A CN117954051 A CN 117954051A CN 202410347088 A CN202410347088 A CN 202410347088A CN 117954051 A CN117954051 A CN 117954051A
- Authority
- CN
- China
- Prior art keywords
- task
- model
- dental
- conclusion
- treatment
- 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
- 238000011282 treatment Methods 0.000 title claims abstract description 174
- 230000007547 defect Effects 0.000 title claims abstract description 44
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 23
- 230000006698 induction Effects 0.000 claims abstract description 9
- 230000008439 repair process Effects 0.000 claims description 68
- 230000003239 periodontal effect Effects 0.000 claims description 17
- 238000012549 training Methods 0.000 claims description 17
- 210000004262 dental pulp cavity Anatomy 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 11
- 238000011269 treatment regimen Methods 0.000 claims description 11
- 238000000605 extraction Methods 0.000 claims description 10
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 230000014759 maintenance of location Effects 0.000 claims description 9
- 206010043183 Teething Diseases 0.000 claims description 8
- 238000010521 absorption reaction Methods 0.000 claims description 8
- 230000036346 tooth eruption Effects 0.000 claims description 8
- 230000010354 integration Effects 0.000 claims description 6
- 210000000988 bone and bone Anatomy 0.000 claims description 5
- 238000003745 diagnosis Methods 0.000 claims description 5
- 210000001738 temporomandibular joint Anatomy 0.000 claims description 4
- 239000007943 implant Substances 0.000 claims description 3
- 239000000523 sample Substances 0.000 claims 5
- 210000002569 neuron Anatomy 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 230000007717 exclusion Effects 0.000 description 2
- 238000002513 implantation Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 208000024827 Alzheimer disease Diseases 0.000 description 1
- 208000035874 Excoriation Diseases 0.000 description 1
- 206010028980 Neoplasm Diseases 0.000 description 1
- 208000006011 Stroke Diseases 0.000 description 1
- 208000008312 Tooth Loss Diseases 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 239000002253 acid Substances 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000001055 chewing effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 208000002925 dental caries Diseases 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 210000000214 mouth Anatomy 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000035790 physiological processes and functions Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000009323 psychological health Effects 0.000 description 1
- 238000011127 radiochemotherapy Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000001356 surgical procedure Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 210000001519 tissue Anatomy 0.000 description 1
- 230000008733 trauma Effects 0.000 description 1
Abstract
The invention provides a method, a device and a storage medium for acquiring a dental defect treatment scheme based on artificial intelligence. The method comprises the following steps: the method comprises the steps of obtaining clinical examination information of the suffering teeth, inputting the clinical examination information of the suffering teeth into a trained multitask model, respectively outputting a yes or no suffering teeth task conclusion by taking a task as a unit, inputting the suffering teeth task conclusion into a logic induction model, and outputting to obtain an suffering teeth treatment scheme. In this way, a comprehensive and accurate dental defect treatment scheme can be obtained, accurate personalized medical treatment is realized, and medical resources are saved.
Description
Technical Field
The invention relates to the technical field of dental defect treatment design, in particular to a method, a device and a storage medium for acquiring a dental defect treatment scheme based on artificial intelligence.
Background
Dental defects are common and frequently occurring diseases of the stomatology, and the incidence rate is about 24% -53%. The cause of the tooth defect is complex and various, one of the most main causes is caries, trauma, abrasion and acid erosion, stress, developmental deformity and the like, and the tooth defect is also a common cause. Dental defects often have adverse effects on chewing, development, periodontal tissue, aesthetics, psychological health, even general health, etc. If not treated in time, it may eventually lead to loss of teeth. The reserved treatment of dental defects is of critical importance, including maintaining alveolar bone height, physiological function of teeth, reserving tooth Zhou Benti receptors, reducing the risk of alzheimer's disease and cerebral stroke, and the like. Common treatment means include filling, inlay, veneering, full crown, piling nuclear crown, implantation, partial denture, full denture and the like, and professional treatment such as tooth body, periodontal, surgery, orthodontic and the like is combined when other symptoms are combined. Although various treatment means exist for the existing tooth defect, the clinical treatment of the tooth defect lacks unified standards, and the difficulty still exists in how to select a more reasonable and more suitable scheme for patients.
Disclosure of Invention
To overcome at least some of the problems with the related art, the present invention provides a method for obtaining a dental defect treatment regimen based on artificial intelligence, comprising: acquiring clinical examination information of the suffering teeth; inputting the clinical examination information of the suffering teeth into a trained multitask model, wherein the multitask model outputs a yes or no suffering teeth task conclusion respectively; and inputting the dental task conclusion to a logic generalization model, and outputting to obtain a dental treatment scheme.
Further, the multi-task model includes a plurality of first task models related to tooth removal, the plurality of first task models including: observing a task model, and outputting whether an observed dental task conclusion is obtained; the task model is pulled out, and a task conclusion of the patient teeth is output whether to pull out or not; and the task model is reserved, and whether the task conclusion of the suffering tooth is reserved or not is output.
Further, the multitasking model further comprises a plurality of second task models related to plucking, the plurality of second task models comprising: the planting task model outputs a dental task conclusion whether to plant or not; the movable denture task model outputs a dental task conclusion whether the movable denture is installed or not; the total denture task model outputs a dental task conclusion whether the total denture is installed or not; and the fixed bridge repair task model outputs a dental task conclusion of whether the fixed bridge is repaired or not.
Further, the multitasking model further comprises a plurality of third task models relating to retention, the plurality of third task models comprising: the dental treatment task model outputs a dental treatment task conclusion whether the dental treatment is carried out; the repair treatment task model outputs a dental task conclusion whether to repair treatment or not; the periodontal treatment task model outputs a dental task conclusion of whether periodontal treatment is performed; and the orthodontic treatment task model outputs a dental task conclusion of whether orthodontic treatment is performed.
Further, the extraction task model, the dental treatment task model, and the periodontal treatment task model are implemented using a tree-model-based integration algorithm (XGBoost), and the observation task model, the retention task model, the implant task model, the removable denture task model, the total denture task model, the fixed bridge repair task model, the repair treatment task model, and the orthodontic treatment task model are implemented using a multi-layer perceptron (MLP).
Further, the logical generalization model is to: receiving and analyzing the dental task conclusions output by the plurality of first task models; when the observed dentosis task conclusion is yes, generating the dentosis treatment scheme according to the observed dentosis task conclusion; when the task conclusion of the tooth suffering from the tooth is yes, analyzing a plurality of second task models related to the tooth extraction, and generating the tooth suffering treatment scheme according to the task conclusion of the tooth suffering from the tooth, which is output by the task model and the plurality of second task models; and when the task conclusion of the reserved suffering teeth is yes, analyzing the plurality of third task models related to the reserved teeth, and generating the treatment scheme of the suffering teeth according to the task conclusion of the reserved teeth, which is output by the task models and the plurality of third task models.
Further, the multi-task model further includes a plurality of fourth task models related to dental treatment, the plurality of fourth task models including: the dental filling task model outputs a dental filling task conclusion of whether the dental filling is carried out; the root canal treatment task model outputs a dental task conclusion of whether the root canal treatment is performed; and the root tip operation task model outputs a dental task conclusion of whether the root tip operation is performed.
The multitasking model further includes a plurality of fifth task models related to repair treatments, the plurality of fifth task models including: removing the prosthesis task model, and outputting a dental task conclusion whether to remove the prosthesis; re-cementing the task model, and outputting a dental task conclusion whether to be re-cemented; the veneering repair task model outputs a dental task conclusion whether veneering repair is performed; the inlay repairing task model outputs a dental task conclusion whether inlay repairing is carried out; the crown repair task model outputs a dental task conclusion whether crown repair is performed; and the stump nuclear crown repair task model outputs a dental task conclusion of whether the stump nuclear crown repair is performed.
Further, generating the teething treatment regimen according to the teething task conclusions output by the reservation task model and the plurality of third task models includes: when the dentosis task conclusion of whether the dental treatment is yes, analyzing the plurality of fourth task models related to the dental treatment, and generating a dentosis treatment scheme according to the dentosis task conclusion output by the reserved task model, the plurality of third task models and the plurality of fourth task models; and when the judging result of the dentate task of the repair treatment is yes, analyzing the plurality of fifth task models related to the repair treatment, and generating the dentate treatment scheme according to the dentate task conclusion output by the reserved task model, the plurality of third task models and the plurality of fifth task models.
Further, the patient clinical examination information includes at least one of: the tooth position of the affected tooth, whether the opposite tooth is missing, whether a filling body exists, whether a restoration exists, the number of the remained teeth, whether a false tooth exists, a lingual section, a labial section, a mesial wall section, a distal mesial wall section, whether hidden fissures exist, root canal treatment, root tip Zhou Bingbian, extraroot absorption, root length, clinical crown height, crown root ratio, alveolar bone absorption degree, adjacent tooth inclination, loosening degree, missing of the opposite tooth, occlusion condition of the affected tooth, oral auxiliary function, temporomandibular joint TMJ, oral hygiene condition and tooth defect diagnosis.
Further, the method further comprises: training the multitasking model, the training comprising: acquiring clinical examination information of a plurality of groups of sample suffering teeth and corresponding task conclusions of a plurality of groups of sample suffering teeth; and taking the clinical examination information of the plurality of groups of sample suffering teeth as input, taking the task conclusion of the plurality of groups of sample suffering teeth as output, and training the multi-task model.
The present invention further provides an apparatus for obtaining a dental defect treatment regimen based on artificial intelligence, comprising a memory and a processor coupled to the memory, the processor being configured to perform the steps of the method described above based on instructions stored in the memory.
Further, the present invention provides a storage medium having stored thereon an executable program which, when called, performs the steps of the method described above.
In addition, the invention also provides a device for acquiring the dental defect treatment scheme based on artificial intelligence, which comprises the following components: the information acquisition module is used for acquiring clinical examination information of the suffering teeth; the multi-task model is used for receiving the clinical examination information of the suffering teeth and respectively outputting a yes or no conclusion of the task of the suffering teeth by taking the task as a unit, wherein the multi-task model is a trained artificial intelligent model; and the logic induction model is used for receiving the dental task conclusion and outputting a dental treatment scheme.
Further, the multi-task model includes a plurality of first task models related to tooth removal, the plurality of first task models including: observing a task model, and outputting whether an observed dental task conclusion is obtained; the task model is pulled out, and a task conclusion of the patient teeth is output whether to pull out or not; and the task model is reserved, and whether the task conclusion of the suffering tooth is reserved or not is output.
Further, the multitasking model further comprises a plurality of second task models related to plucking, the plurality of second task models comprising: the planting task model outputs a dental task conclusion whether to plant or not; the movable denture task model outputs a dental task conclusion whether the movable denture is installed or not; the total denture task model outputs a dental task conclusion whether the total denture is installed or not; and the fixed bridge repair task model outputs a dental task conclusion of whether the fixed bridge is repaired or not.
Further, the multitasking model further comprises a plurality of third task models relating to retention, the plurality of third task models comprising: the dental treatment task model outputs a dental treatment task conclusion whether the dental treatment is carried out; the repair treatment task model outputs a dental task conclusion whether to repair treatment or not; the periodontal treatment task model outputs a dental task conclusion of whether periodontal treatment is performed; and the orthodontic treatment task model outputs a dental task conclusion of whether orthodontic treatment is performed.
Further, the extraction task model, the dental treatment task model, and the periodontal treatment task model are implemented using a tree-model-based integration algorithm (XGBoost), and the observation task model, the retention task model, the implant task model, the removable denture task model, the total denture task model, the fixed bridge repair task model, the repair treatment task model, and the orthodontic treatment task model are implemented using a multi-layer perceptron (MLP).
Further, the logical generalization model is to: receiving and analyzing the dental task conclusions output by the plurality of first task models; when the observed dentosis task conclusion is yes, generating the dentosis treatment scheme according to the observed dentosis task conclusion; when the task conclusion of the tooth suffering from the tooth is yes, analyzing a plurality of second task models related to the tooth extraction, and generating the tooth suffering treatment scheme according to the task conclusion of the tooth suffering from the tooth, which is output by the task model and the plurality of second task models; and when the task conclusion of the reserved suffering teeth is yes, analyzing the plurality of third task models related to the reserved teeth, and generating the treatment scheme of the suffering teeth according to the task conclusion of the reserved teeth, which is output by the task models and the plurality of third task models.
Further, the multi-task model further includes a plurality of fourth task models related to dental treatment, the plurality of fourth task models including: the dental filling task model outputs a dental filling task conclusion of whether the dental filling is carried out; the root canal treatment task model outputs a dental task conclusion of whether the root canal treatment is performed; and the root tip operation task model outputs a dental task conclusion of whether the root tip operation is performed.
The multitasking model further includes a plurality of fifth task models related to repair treatments, the plurality of fifth task models including: removing the prosthesis task model, and outputting a dental task conclusion whether to remove the prosthesis; re-cementing the task model, and outputting a dental task conclusion whether to be re-cemented; the veneering repair task model outputs a dental task conclusion whether veneering repair is performed; the inlay repairing task model outputs a dental task conclusion whether inlay repairing is carried out; the crown repair task model outputs a dental task conclusion whether crown repair is performed; and the stump nuclear crown repair task model outputs a dental task conclusion of whether the stump nuclear crown repair is performed.
Further, the logic generalization model generating the teething treatment plan according to the teething task conclusions output by the reservation task model and the plurality of third task models comprises: when the dentosis task conclusion of whether the dental treatment is yes, analyzing the plurality of fourth task models related to the dental treatment, and generating a dentosis treatment scheme according to the dentosis task conclusion output by the reserved task model, the plurality of third task models and the plurality of fourth task models; and when the judging result of the dentate task of the repair treatment is yes, analyzing the plurality of fifth task models related to the repair treatment, and generating the dentate treatment scheme according to the dentate task conclusion output by the reserved task model, the plurality of third task models and the plurality of fifth task models.
Further, the patient clinical examination information includes at least one of: the tooth position of the affected tooth, whether the opposite tooth is missing, whether a filling body exists, whether a restoration exists, the number of the remained teeth, whether a false tooth exists, a lingual section, a labial section, a mesial wall section, a distal mesial wall section, whether hidden fissures exist, root canal treatment, root tip Zhou Bingbian, extraroot absorption, root length, clinical crown height, crown root ratio, alveolar bone absorption degree, adjacent tooth inclination, loosening degree, missing of the opposite tooth, occlusion condition of the affected tooth, oral auxiliary function, temporomandibular joint TMJ, oral hygiene condition and tooth defect diagnosis.
Further, the multitasking model is used to: in the training process, a plurality of groups of sample suffering tooth clinical examination information are used as input of the multi-task model, a plurality of groups of sample suffering tooth task conclusions corresponding to the plurality of groups of sample suffering tooth clinical examination information are used as output of the multi-task model, and the multi-task model is trained.
In addition, the invention also provides a device for acquiring the dental defect treatment scheme based on artificial intelligence, which comprises the following components: the acquisition module is used for acquiring clinical examination information of the suffering teeth; the first input module is used for inputting the clinical examination information of the suffering teeth into a trained multitask model, and the multitask model respectively outputs a yes or no suffering teeth task conclusion by taking a task as a unit; and the second input module is used for inputting the dental task conclusion to a logic induction model and outputting to obtain a dental treatment scheme.
According to the embodiment of the invention, the clinical examination information of the suffering teeth is obtained, the clinical examination information of the suffering teeth is input into a trained multitask model, the multitask model respectively outputs the yes or no suffering teeth task conclusion in a task unit, and the suffering teeth task conclusion is input into a trained logic induction model and is output to obtain the suffering teeth treatment scheme. In this way, a comprehensive and accurate dental defect treatment scheme can be obtained, accurate personalized medical treatment is realized, and medical resources are saved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description, serve to explain the principles of the invention.
Fig. 1 shows a flow chart of a method for obtaining a dental defect treatment regimen based on artificial intelligence in accordance with an embodiment of the present invention.
FIG. 2 illustrates a logic generalization flowchart of a logic generalization model according to an embodiment of the present invention.
FIG. 3 shows a flow diagram for training a multitasking model according to an embodiment of the invention.
FIG. 4 illustrates a task model of a multi-layer perceptron implementation, in accordance with an embodiment of the present invention.
Fig. 5 illustrates a schematic diagram of an apparatus for obtaining a dental defect treatment regimen based on artificial intelligence in accordance with an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the spirit of the present disclosure will be clearly described in the following drawings and detailed description, and any person skilled in the art, after having appreciated the embodiments of the present disclosure, may make alterations and modifications by the techniques taught by the present disclosure without departing from the spirit and scope of the present disclosure.
The exemplary embodiments of the present invention and the descriptions thereof are intended to illustrate the present invention, but not to limit the present invention. In addition, the same or similar reference numerals are used for the same or similar parts in the drawings and the embodiments.
The terms "first," "second," …, etc. as used herein do not denote a particular order or sequence, nor are they intended to limit the invention, but rather are merely used to distinguish one element or operation from another in the same technical term.
As used herein, the terms "comprising," "including," "having," "containing," and the like are intended to be inclusive and mean an inclusion, but not limited to.
As used herein, "and/or" includes any or all combinations of such things.
Reference herein to "a plurality" includes "two" and "more than two"; the term "plurality of sets" as used herein includes "two sets" and "more than two sets".
Certain words used to describe the invention will be discussed below or elsewhere in this specification to provide additional guidance to those skilled in the art in describing the invention.
Currently, although there are a variety of treatments for dental defects, there is a lack of unified standards for the treatment of clinical dental defects and difficulties in selecting a more rational, more patient-friendly regimen.
The embodiment of the invention provides a method for acquiring a dental defect treatment scheme based on artificial intelligence, which can obtain a comprehensive and accurate dental defect treatment scheme by using a trained multitask model and a logic induction model, realizes accurate personalized medical treatment and saves medical resources.
Fig. 1 shows a flow chart of a method for obtaining a dental defect treatment regimen based on artificial intelligence in accordance with an embodiment of the present invention. As shown in FIG. 1, the method for obtaining a dental defect treatment scheme based on artificial intelligence according to the embodiment of the invention comprises the following steps S11-S13.
S11, acquiring clinical examination information of the affected teeth.
First, patient dental history information is acquired. And secondly, standardizing the obtained information of the suffering dental medical record to obtain clinical examination information of suffering dental.
The patient dental history information may include several items. In one embodiment, the inventor researches refer to the teaching materials of eight years of northern da oral cavity, 3 rd edition of oral restoration science, modern oral restoration science standard diagnosis and treatment manual, electronic dental defect case templates of northern da oral restoration department and related researches, extracts the suffering dental medical record information into a plurality of items shown in the following table 1 as suffering dental clinical examination information, and standardizes the suffering dental medical record information by referring to the standardized mode in the following table 1.
TABLE 1
Thus, in one embodiment, the patient clinical examination information includes at least one of: the tooth position of the affected tooth, whether the opposite tooth is missing, whether a filling body exists, whether a restoration exists, the number of the remained teeth, whether a false tooth exists, a lingual section, a labial section, a mesial wall section, a distal mesial wall section, whether hidden fissures exist, root canal treatment, root tip Zhou Bingbian, extraroot absorption, root length, clinical crown height, crown root ratio, alveolar bone absorption degree, adjacent tooth inclination, loosening degree, missing of the opposite tooth, occlusion condition of the affected tooth, oral auxiliary function, temporomandibular joint TMJ, oral hygiene condition and tooth defect diagnosis.
The clinical examination information of the affected teeth standardized in the above manner can be used for subsequent input.
Those skilled in the art know that standardized clinical examination information of the patient's tooth can also be directly used as the patient medical record information in the clinical examination process of the patient.
S12, inputting the information of the clinical examination of the suffering teeth into a trained multitask model, wherein the multitask model respectively outputs a yes or no conclusion of the task of the suffering teeth by taking the task as a unit.
In the embodiment of the invention, a multi-task model is set.
In particular, the multi-task model comprises a plurality of first task models related to tooth removal. The plurality of first task models may include: observing a task model, and outputting whether an observed dental task conclusion is obtained; the task model is pulled out, and a task conclusion of the patient teeth is output whether to pull out or not; and the task model is reserved, and whether the task conclusion of the suffering tooth is reserved or not is output.
Further, the multitasking model also includes a plurality of second task models related to plucking. The plurality of second task models may include: the planting task model outputs a dental task conclusion whether to plant or not; the movable denture task model outputs a dental task conclusion whether the movable denture is installed or not; the total denture task model outputs a dental task conclusion whether the total denture is installed or not; and the fixed bridge repair task model outputs a dental task conclusion of whether the fixed bridge is repaired or not.
Further, the multi-tasking model further comprises a plurality of third tasking models related to retention. The plurality of third task models may include: the dental treatment task model outputs a dental treatment task conclusion whether the dental treatment is carried out; the repair treatment task model outputs a dental task conclusion whether to repair treatment or not; the periodontal treatment task model outputs a dental task conclusion of whether periodontal treatment is performed; and the orthodontic treatment task model outputs a dental task conclusion of whether orthodontic treatment is performed.
Further, the multi-task model also includes a plurality of fourth task models related to dental treatment. The plurality of fourth task models includes: the dental filling task model outputs a dental filling task conclusion of whether the dental filling is carried out; the root canal treatment task model outputs a dental task conclusion of whether the root canal treatment is performed; and the root tip operation task model outputs a dental task conclusion of whether the root tip operation is performed.
Further, the multitasking model further comprises a plurality of fifth task models related to repair treatments. The plurality of fifth task models includes: removing the prosthesis task model, and outputting a dental task conclusion whether to remove the prosthesis; re-cementing the task model, and outputting a dental task conclusion whether to be re-cemented; the veneering repair task model outputs a dental task conclusion whether veneering repair is performed; the inlay repairing task model outputs a dental task conclusion whether inlay repairing is carried out; the crown repair task model outputs a dental task conclusion whether crown repair is performed; and the stump nuclear crown repair task model outputs a dental task conclusion of whether the stump nuclear crown repair is performed.
In one embodiment, the extraction task model, the dental treatment task model, the crown repair task model, the post-nuclear crown repair task model, and the periodontal treatment task model are implemented using a tree model-based integration algorithm (XGBoost), and the remaining task models are implemented using a multi-layer perceptron (MLP).
The training process of the task model will be described in detail later.
S13, inputting the dental task conclusion to a logic generalization model, and outputting to obtain a dental treatment scheme.
In the embodiment of the invention, the logic relationship exists among the tasks.
In one implementation, the logic generalization model is configured to receive and analyze the dental task conclusions output by the plurality of first task models; generating a dental treatment scheme according to the observed dental task conclusion when the observed dental task conclusion is yes; when the task conclusion of the suffering teeth, which is about the pulling out, is yes, analyzing the plurality of second task models related to the pulling out, and generating a suffering teeth treatment scheme according to the task conclusion of the suffering teeth, which is output by the task model and the plurality of second task models; and when the task conclusion of the reserved suffering teeth is yes, analyzing the third task models related to the reserved teeth, and generating a treatment scheme of the suffering teeth according to the task conclusion of the reserved teeth, which is output by the task models and the third task models.
In one implementation, generating the teething treatment regimen from the teething task conclusions output by the retention task model and the plurality of third task models includes: when the dentosis task conclusion of whether the dental treatment is yes, analyzing the plurality of fourth task models related to the dental treatment, and generating a dentosis treatment scheme according to the dentosis task conclusion output by the reserved task model, the plurality of third task models and the plurality of fourth task models; and when the judging result of the dentate task of the repair treatment is yes, analyzing the plurality of fifth task models related to the repair treatment, and generating the dentate treatment scheme according to the dentate task conclusion output by the reserved task model, the plurality of third task models and the plurality of fifth task models.
FIG. 2 illustrates a logic generalization flowchart of a logic generalization model according to an embodiment of the present invention. As shown in fig. 2, the logical generalization model includes three levels in total: the layer is left, the treatment layer 1 and the treatment layer 2. In the removing layer, firstly, judging the removing and treating of the teeth, including observing, removing and retaining, wherein mutual exclusion is carried out among the three conditions, if the removing is carried out, the observation or the retaining is not carried out, and vice versa. After the selection of the deputy layer is finished, the treatment layer 1 is reached, and the selection is correspondingly carried out in the treatment layer 1 according to the selection result of the deputy layer, such as: the removal layer is selected for removal, and the treatment scheme possibly given later comprises implantation, removable denture, total denture or fixed bridge repair, and the relationship among the four can be simultaneously present in the treatment scheme. Treatment layer 1 is selected to be at treatment layer 2 after the end, such as: selecting a tooth treatment in the treatment layer 1, wherein the treatment layer 2 can be selected from tooth filling, root canal treatment or root tip operation as a treatment scheme; another example is: if the retention layer is selected and the repair treatment is selected in the treatment layer 1, then one of the repair body removal, the re-cementing, the veneering repair, the inlay repair, the crown repair or the stake core crown repair can be selected as a treatment scheme in the treatment layer 2.
So far, through logical generalization of the logical generalization model to the dental task conclusion, a comprehensive and accurate dental defect treatment scheme can be obtained, accurate personalized medical treatment is realized, and medical resources are saved.
FIG. 3 shows a flow diagram for training a multitasking model according to an embodiment of the invention. As shown in fig. 3, the method for training the multitasking model according to the embodiment of the present invention includes the following steps S31-S32.
S31, obtaining clinical examination information of a plurality of groups of sample suffering teeth and corresponding task conclusions of a plurality of groups of sample suffering teeth.
Firstly, sample suffering tooth clinical examination information and sample suffering tooth task conclusion are collected.
In the embodiment of the invention, the dental cases related to the special department dental defect of a hospital after 2015 are collected,
Inclusion criteria were as follows:
-18 years old;
-diagnosing a tooth defect;
complete clinical and imaging examinations;
The doctor's job is called the secondary altitude and above.
The exclusion criteria were as follows:
poor overall condition and inability to cooperate with therapy;
-severe opening is limited;
-large area maxillofacial defects;
-before and after tumor chemoradiotherapy.
Through the preliminary screening process, an initial data set is obtained, and contains 27953 cases, which contain 27953 groups of sample medical record information (sample clinical examination information) and a plurality of groups of sample dental task conclusions corresponding to the 27953 groups of sample medical record information.
And secondly, standardizing the acquired sample medical record information and the sample dental task conclusion.
Clinical medical records are typically recorded in natural language, which is not straightforward for machine learning. Therefore, it is necessary to divide the case into several items and standardize it as a label, and record it in a form easy for machine learning.
For sample medical record information, in one implementation, the sample medical record information may be collected and standardized according to table 1 above, to obtain sample clinical examination information of the patient's teeth.
For sample dentate task conclusions, in one embodiment, corresponding to each dentate task model, sample dentate task conclusions may be collected and normalized according to Table 2 below.
TABLE 2
As described above, 20 task models are set in total in the embodiment of the invention, and 20 dental task conclusions are correspondingly output. However, the 20 task models and the corresponding dentate task conclusions are only one embodiment of the present invention, and those skilled in the art should understand that more or fewer task models and corresponding dentate task conclusions may be set depending on the number of hospital treatment protocols.
It is also known to those skilled in the art that table 2 may be written as 0 instead of 1.
As above, a plurality of sets of sample dental clinical examination information for training and a plurality of sets of sample dental task conclusions are prepared.
S32, taking the clinical examination information of the plurality of groups of sample suffering teeth as input, taking the task conclusion of the plurality of groups of sample suffering teeth as output, and training the multi-task model.
In one implementation, as described above, 20 task models are provided.
There are many artificial intelligence models currently used in the medical field, including traditional models such as linear regression, logistic regression, support vector machines, etc., and newer neural network models such as multi-layer perceptrons, convolutional neural networks, etc.
In the initial stage of training, the embodiment of the invention selects an integration algorithm (XGBoost) based on a tree model, a Bayes model (Bayes), a multi-layer perceptron (MLP), a Support Vector Machine (SVM) and a random forest model. After testing, the accuracy of the tree model-based integrated algorithm and the multi-layer perceptron is found to be superior to that of the other three traditional models. The tree model-based integrated algorithm is better in five tasks, namely a dental treatment task, a periodontal treatment task, a crown repair task, a stake nuclear crown repair task and an extraction task, and better in the performance of the multi-layer perceptron in the rest tasks.
Therefore, in the embodiment of the invention, the five models of the dental treatment task model, the periodontal treatment task model, the crown repair task model, the post-nuclear crown repair task model and the extraction task model are realized by adopting an integration algorithm (XGBoost) based on a tree model, and the other task models are realized by adopting a multi-layer perceptron (MLP).
Of course, those skilled in the art will appreciate that other artificial intelligence models may be employed to implement the task models.
In the embodiment of the invention, each task model is trained respectively.
Specifically, for any one task model, multiple groups of sample dental clinical examination information are used as input, and dental task conclusions corresponding to any one task model in multiple groups of sample dental task conclusions are used as output to train the any one task model.
For example, for an observation task model, multiple groups of sample suffering tooth clinical examination information are used as input, and whether the suffering tooth task conclusion corresponding to the observation task model is observed or not in the multiple groups of sample suffering tooth task conclusions is used as output, so that the observation task model is trained, and a trained observation task model is obtained.
Aiming at five task models, namely a dental treatment task model, a periodontal treatment task model, a crown repair task model, a stake nuclear crown repair task model and an extraction task model, which are realized by adopting a tree model-based integrated algorithm (XGBoost), the task models are completed by adopting models provided in a sklearn library in the embodiment of the invention. Different parameter settings are obtained for different task models respectively.
The determined multi-layer perceptron (MLP) model is shown in fig. 4 for other task models implemented using MLP. In fig. 4, the input layer is the first layer of the neural network for receiving the raw data. At the input layer, the data is passed to the hidden layer after preprocessing and normalization. The normalization layer is generally used for normalizing input data to accelerate training and improve stability of the model. By normalizing the data to a unified standard, the scale differences between different features can be reduced, making the model easier to learn. The ReLU layer is a commonly used activation function that functions to non-linearly transform the output of neurons, enabling the network to learn more complex feature representations. The formula of the ReLU function is f (x) =max (0, x), and when the input is a positive number, the value of the positive number is output; when the input is negative, the output is 0. The fully connected layer is a special layer for converting the feature layer into a predicted output. In the fully connected layer, each neuron is connected to all neurons of the previous layer and the outputs of the previous layer are weighted and summed. At the Sigmoid output layer, the Sigmoid function is a commonly used activation function for mapping the output of neurons between 0 and 1. In the output layer, the output of the fully connected layer is typically converted into a probability distribution form using Sigmoid function as an activation function for prediction of multi-classification problems or bi-classification problems.
Those skilled in the art will appreciate that for the output of each task model, a positive conclusion may be set to output 1 and a negative conclusion to output 0. Of course, a positive conclusion may be set to output 0 and a negative conclusion to output 1.
By respectively training each task model, a trained multi-task model can be obtained.
In the embodiment of the invention, as described above, 27953 groups of sample suffering tooth clinical examination information and corresponding groups of sample suffering tooth task conclusions are collected, all data are randomly divided into a training set and a testing set according to the proportion of 8:2, 80% of the data are used as samples to train each task model, and finally 20% of the data are used as the testing set to test the trained task model, and the accuracy rate of the output result reaches more than 90% through comparison.
According to the specific embodiment of the invention, the clinical examination information of the suffering teeth is obtained, the clinical examination information of the suffering teeth is input into a trained multitask model, the multitask model respectively outputs a yes or no suffering teeth task conclusion by taking a task as a unit, the suffering teeth task conclusion is input into a trained logic induction model, and an suffering teeth treatment scheme is obtained through output. In this way, a comprehensive and accurate dental defect treatment scheme can be obtained, accurate personalized medical treatment is realized, and medical resources are saved.
Embodiments of the present invention also provide an apparatus for obtaining a dental defect treatment regimen based on artificial intelligence, comprising a memory and a processor coupled to the memory, the processor configured to execute the method for obtaining a dental defect treatment regimen based on artificial intelligence in any of the embodiments of the present invention based on instructions stored in the memory.
The memory may be a system memory or a fixed nonvolatile storage medium, and the system memory may store an operating system, an application program, a boot loader, a database, and other programs.
Embodiments of the present invention also provide a computer readable storage medium, such as a memory comprising a computer program executable by a processor to perform the method of obtaining a dental defect treatment regimen based on artificial intelligence in any of the embodiments of the present invention.
The embodiment of the invention also provides a device for acquiring the dental defect treatment scheme based on artificial intelligence, which comprises the following steps: the acquisition module is used for acquiring clinical examination information of the suffering teeth; the first input module is used for inputting the clinical examination information of the suffering teeth into a trained multitask model, and the multitask model respectively outputs a yes or no suffering teeth task conclusion by taking a task as a unit; and the second input module is used for inputting the dental task conclusion to a logic induction model and outputting to obtain a dental treatment scheme.
As shown in fig. 5, an embodiment of the present invention further provides an apparatus for obtaining a dental defect treatment plan based on artificial intelligence, including: the information acquisition module is used for acquiring clinical examination information of the suffering teeth; the multi-task model is used for receiving the clinical examination information of the suffering teeth and respectively outputting a yes or no conclusion of the task of the suffering teeth by taking the task as a unit, wherein the multi-task model is a trained artificial intelligent model; and the logic induction model is used for receiving the dental task conclusion and outputting a dental treatment scheme.
Details of each module and model in this embodiment may refer to the above method embodiments, and are not described herein.
The foregoing is merely illustrative of the embodiments of this invention and any equivalent and equivalent changes and modifications can be made by those skilled in the art without departing from the spirit and principles of this invention.
Claims (12)
1. A method for obtaining a dental defect treatment regimen based on artificial intelligence, comprising:
Acquiring clinical examination information of the suffering teeth;
Inputting the clinical examination information of the suffering teeth into a trained multitask model, wherein the multitask model respectively outputs a yes or no conclusion of the task of the suffering teeth by taking the task as a unit; and
Inputting the dental task conclusion to a logic induction model, outputting to obtain a dental treatment scheme,
Wherein the multitasking model comprises a plurality of first task models related to tooth removal, the plurality of first task models comprising:
Observing a task model, and outputting whether an observed dental task conclusion is obtained;
the task model is pulled out, and a task conclusion of the patient teeth is output whether to pull out or not; and
And (5) reserving the task model, and outputting whether the reserved task conclusion of the suffering tooth is reserved or not.
2. The method of claim 1, wherein the multitasking model further comprises a plurality of second task models related to plucking, the plurality of second task models comprising:
the planting task model outputs a dental task conclusion whether to plant or not;
the movable denture task model outputs a dental task conclusion whether the movable denture is installed or not;
the total denture task model outputs a dental task conclusion whether the total denture is installed or not; and
And outputting a dental task conclusion whether the fixed bridge is repaired or not by the fixed bridge repair task model.
3. The method of claim 2, wherein the multitasking model further comprises a third plurality of task models relating to retention, the third plurality of task models comprising:
The dental treatment task model outputs a dental treatment task conclusion whether the dental treatment is carried out;
the repair treatment task model outputs a dental task conclusion whether to repair treatment or not;
the periodontal treatment task model outputs a dental task conclusion of whether periodontal treatment is performed; and
And outputting a dental task conclusion whether orthodontic treatment is performed or not by the orthodontic treatment task model.
4. The method of claim 3, wherein the extraction task model, the dental treatment task model, and the periodontal treatment task model are implemented using a tree-model-based integration algorithm (XGBoost), and wherein the observation task model, the retention task model, the implant task model, the active denture task model, the total denture task model, the fixed bridge repair task model, the repair treatment task model, and the orthodontic treatment task model are implemented using a multi-layer perceptron (MLP).
5. A method according to claim 3, wherein the logical generalization model is for:
receiving and analyzing the dental task conclusions output by the plurality of first task models;
When the observed dentosis task conclusion is yes, generating the dentosis treatment scheme according to the observed dentosis task conclusion;
when the task conclusion of the tooth suffering from the tooth is yes, analyzing a plurality of second task models related to the tooth extraction, and generating the tooth suffering treatment scheme according to the task conclusion of the tooth suffering from the tooth, which is output by the task model and the plurality of second task models;
And when the task conclusion of the reserved suffering teeth is yes, analyzing the plurality of third task models related to the reserved teeth, and generating the treatment scheme of the suffering teeth according to the task conclusion of the reserved teeth, which is output by the task models and the plurality of third task models.
6. The method of claim 5, wherein the step of determining the position of the probe is performed,
The multi-task model further includes a plurality of fourth task models related to dental treatment, the plurality of fourth task models including:
the dental filling task model outputs a dental filling task conclusion of whether the dental filling is carried out;
The root canal treatment task model outputs a dental task conclusion of whether the root canal treatment is performed; and
The root tip operation task model outputs a dental task conclusion of whether the root tip operation is performed;
The multitasking model further includes a plurality of fifth task models related to repair treatments, the plurality of fifth task models including:
removing the prosthesis task model, and outputting a dental task conclusion whether to remove the prosthesis;
Re-cementing the task model, and outputting a dental task conclusion whether to be re-cemented;
The veneering repair task model outputs a dental task conclusion whether veneering repair is performed;
The inlay repairing task model outputs a dental task conclusion whether inlay repairing is carried out;
The crown repair task model outputs a dental task conclusion whether crown repair is performed; and
And outputting a dental task conclusion whether the pile nuclear crown is repaired or not by the pile nuclear crown repair task model.
7. The method of claim 6, wherein the step of providing the first layer comprises,
Generating the teething treatment plan according to the teething task conclusions output by the reservation task model and the plurality of third task models comprises:
when the dentosis task conclusion of whether the dental treatment is yes, analyzing the plurality of fourth task models related to the dental treatment, and generating a dentosis treatment scheme according to the dentosis task conclusion output by the reserved task model, the plurality of third task models and the plurality of fourth task models;
And when the judging result of the dentate task of the repair treatment is yes, analyzing the plurality of fifth task models related to the repair treatment, and generating the dentate treatment scheme according to the dentate task conclusion output by the reserved task model, the plurality of third task models and the plurality of fifth task models.
8. The method of claim 1, wherein the patient clinical examination information comprises at least one of: the tooth position of the affected tooth, whether the opposite tooth is missing, whether a filling body exists, whether a restoration exists, the number of the remained teeth, whether a false tooth exists, a lingual section, a labial section, a mesial wall section, a distal mesial wall section, whether hidden fissures exist, root canal treatment, root tip Zhou Bingbian, extraroot absorption, root length, clinical crown height, crown root ratio, alveolar bone absorption degree, adjacent tooth inclination, loosening degree, missing of the opposite tooth, occlusion condition of the affected tooth, oral auxiliary function, temporomandibular joint TMJ, oral hygiene condition and tooth defect diagnosis.
9. The method according to claim 1, wherein the method further comprises: training the multitasking model, the training comprising:
acquiring clinical examination information of a plurality of groups of sample suffering teeth and corresponding task conclusions of a plurality of groups of sample suffering teeth; and
And taking the clinical examination information of the plurality of groups of sample suffering teeth as input, taking the task conclusion of the plurality of groups of sample suffering teeth as output, and training the multi-task model.
10. An apparatus for obtaining a dental defect treatment plan based on artificial intelligence, comprising a memory and a processor coupled to the memory, the processor configured to perform the steps in the method of any of claims 1-9 based on instructions stored in the memory.
11. A storage medium having stored thereon an executable program which when invoked performs the steps of the method according to any one of claims 1-9.
12. An apparatus for obtaining a dental defect treatment plan based on artificial intelligence, comprising:
The information acquisition module is used for acquiring clinical examination information of the suffering teeth;
The multi-task model is used for receiving the clinical examination information of the suffering teeth and respectively outputting a yes or no conclusion of the task of the suffering teeth by taking the task as a unit, wherein the multi-task model is a trained artificial intelligent model; and
The logic generalization model is used for receiving the dental task conclusion and outputting a dental treatment scheme,
Wherein the multitasking model comprises a plurality of first task models related to tooth removal, the plurality of first task models comprising:
Observing a task model, and outputting whether an observed dental task conclusion is obtained;
the task model is pulled out, and a task conclusion of the patient teeth is output whether to pull out or not; and
And (5) reserving the task model, and outputting whether the reserved task conclusion of the suffering tooth is reserved or not.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410347088.8A CN117954051A (en) | 2024-03-26 | 2024-03-26 | Method, device and storage medium for acquiring dental defect treatment scheme based on artificial intelligence |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410347088.8A CN117954051A (en) | 2024-03-26 | 2024-03-26 | Method, device and storage medium for acquiring dental defect treatment scheme based on artificial intelligence |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117954051A true CN117954051A (en) | 2024-04-30 |
Family
ID=90792419
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410347088.8A Pending CN117954051A (en) | 2024-03-26 | 2024-03-26 | Method, device and storage medium for acquiring dental defect treatment scheme based on artificial intelligence |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117954051A (en) |
-
2024
- 2024-03-26 CN CN202410347088.8A patent/CN117954051A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Agrawal et al. | Artificial intelligence in dentistry: past, present, and future | |
Ahmed et al. | Artificial intelligence techniques: analysis, application, and outcome in dentistry—a systematic review | |
Tandon et al. | Present and future of artificial intelligence in dentistry | |
Revilla-León et al. | Artificial intelligence models for tooth-supported fixed and removable prosthodontics: A systematic review | |
KR20220056234A (en) | Methods, systems and devices for on-the-fly automated design of custom dental objects | |
CN113679500B (en) | AI algorithm-based caries and dental plaque detection and distribution method | |
Milošević et al. | A comprehensive exploration of neural networks for forensic analysis of adult single tooth X-Ray images | |
Bhat et al. | A comprehensive survey of deep learning algorithms and applications in dental radiograph analysis | |
Deleat-Besson et al. | Automatic segmentation of dental root canal and merging with crown shape | |
Jaiswal et al. | Multi oral disease classification from panoramic radiograph using transfer learning and XGBoost | |
Altalhi et al. | The Impact of Artificial Intelligence on Dental Implantology: A Narrative Review | |
CN117954051A (en) | Method, device and storage medium for acquiring dental defect treatment scheme based on artificial intelligence | |
CN112201349A (en) | Orthodontic operation scheme generation system based on artificial intelligence | |
KR102448169B1 (en) | Method and apparatus for predicting orthodontic treatment result based on deep learning | |
CN112420171B (en) | Method and system for classifying maxillary sinus floor bone based on artificial intelligence | |
Jokanović et al. | State of the Art of Artificial Intelligence in Dentistry and Its Expected Future | |
El Joudi et al. | Review of the role of Artificial Intelligence in dentistry: Current applications and trends | |
Roy et al. | Artificial intelligence in dentistry and its future | |
Thulaseedharan et al. | Deep learning based object detection algorithm for the detection of dental diseases and differential treatments | |
Sukegawa et al. | Multi-task deep learning model for classification of dental implant brand and treatment stage using dental panoramic radiograph images. Biomolecules. 2021; 11 | |
Sornam | Artificial Intelligence in Orthodontics-An exposition | |
WO2021240207A1 (en) | Predicting, diagnosing, and proposing treatment plans for oral and dental diseases in humans and animals with artificial intelligence, machine learning and deep learning techniques | |
Kommalapati et al. | Artificial neural networking in applied dentistry | |
Khairkar et al. | Future Trend: Review of Artificial Intelligence Applications in Prosthodontics | |
Kunt | Dental caries detection from bitewing X-ray images |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination |