Research On Surgical Route For Removal Of Impacted Wisdom Teeth Based On Knowledge Grap

As people’s standard of living rises,maintaining oral health has become increasingly important to people.Among them,impacted wisdom teeth,especially mandibular impacted wisdom teeth,have a high incidence rate and many postoperative complications.There is no systematic theoretical basis for surgical details,such as incision size,incision angle,etc.,and the details are still formulated based on the doctor’s experience or the actual situation during the operation.At the same time,the knowledge map has strong relational expression ability,and can handle complex and diverse association analysis.Therefore,this topic utilizes the advantages of knowledge atlas to apply knowledge atlas technology to the recommendation of surgical plans for impacted wisdom tooth extraction.The main research results of this paper are as follows:(1)Aiming at the problem that the classical BiLSTM model is insufficient in extracting local spatial features,this paper proposes a model imConvNet based on convolutional neural network,which is combined with the BiLSTM model and achieves better results than a single model.For the problem of less training corpus,this article uses BERT to pre-train the Chinese model,which greatly improves the effectiveness of named entity recognition.Finally,we constructed the BERTimConvNet-BiLSTM-CRF model.Through training and testing on the Yidu-S4 K dataset,and designing comparative experiments,it can be seen that the performance of the model we designed is superior to other models,and has achieved good results in Chinese medical named entity recognition tasks.(2)There is no relevant research data on stomatology,especially in the field of impacted wisdom teeth.Based on 207 clinical records and CBCT imaging data obtained from Guiyang Stomatological Hospital,this article has compiled the original corpus in the field of impacted wisdom teeth;At the same time,using the Doccano annotation tool,we conducted a solid annotation task and constructed a dataset of impacted wisdom teeth.After that,a knowledge map of impacted wisdom teeth was constructed using the relevant knowledge of the knowledge map.(3)By selecting a user based collaborative filtering algorithm,the recommendation of surgical plans for extraction of impacted wisdom teeth is achieved.Then,using the Qt development framework,a graphical user interface was designed based on user needs,and the research on the recommended path for surgical procedures was ultimately completed.