牙科治疗声品质评价研究

发布时间：2018-06-25 18:35

  本文选题：用户调研 + 牙科治疗声音　； 参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：牙科焦虑症是牙科治疗中常见心理状态,严重影响患者治疗体验、就医及时性和治疗效果。调查表明,牙科治疗过程中产生的噪声是造成牙科焦虑症的重要因素之一。本文从牙科治疗声音入手,通过挖掘出患者对牙科治疗声音的感受评价与理解,选定研究范围,设计实验采集牙科治疗过程中的患者实际听到的声音,对采集到的声音进行客观分析和主观评价,最终建立牙科治疗声品质评价的BP神经网络模型,以预测牙科治疗声音所致焦虑程度。首先,基于UGC平台文本信息挖掘及语料分析,了解到有过牙科治疗经验的患者对牙科治疗声音的焦虑及恐惧的痛点所在;然后对口腔科相关从业人员的访谈,从专业角度了解这些导致焦虑的声音来源;根据所得到的信息,设计调查问卷,确定了患者对牙科治疗声音不适感的来源及程度,并分析了其他因素是否相关。针对上述调查结果确定研究的内容。其次,以对振动作用于牙齿时以空气、骨骼、血液、肌肉为介质的声音传播而导致的听觉特性为依据,设计了用羊头代替人头并在头内嵌入声音采集设备的声音采集实验,采集了专业牙医以专业的治疗手法用三种牙科治疗器械打磨该羊头的牙齿时的声音数据,将声音数据制作成40个声音样本并对其进行了响度、尖锐度、粗糙度、抖动强度、语言清晰度五个心理声学属性的客观分析,并选择了合适的心理声学属性算法。接着,利用声音样本进行主观评价实验。选择26名有牙科治疗经验并对治疗声音存在不适感的被测者进行声音样本听音实验,根据感受对声音样本进行牙科焦虑度评价。对评价结果进行筛选后,考察牙科焦虑度与各心理声学属性参数的相关性,确定主观评价群体以及导致牙科焦虑的心理声学属性。最后,建立牙科治疗声品质的神经网络评价模型,以导致牙科焦虑的三个心理声学属性抖动强度、响度、AI指数为输入,主观评价牙科焦虑度为输出,通过不同算法在建立牙科治疗声音品质评价模型的误差比较,确定利用BP神经网络建模,并确定神经网络的主要参数和结构。最终建立具有预测功能的牙科治疗声音所致焦虑度的神经网络模型,从声音角度提出缓解牙科焦虑症状的方法建议,并阐述该神经网络模型如何应用于优化牙科治疗声品质。
[Abstract]:Dental anxiety disorder is a common psychological state in dental treatment, which seriously affects the experience, timeliness and effect of treatment. The investigation shows that the noise produced during dental treatment is one of the important factors causing dental anxiety. This article starts with the dental treatment sound, through excavates the patient to the dental treatment sound feeling appraisal and the understanding, selects the research scope, designs the experiment to collect the dental treatment process the patient actually hears the sound, The objective analysis and subjective evaluation of the collected sound were carried out, and the BP neural network model of sound quality evaluation for dental treatment was established to predict the degree of anxiety caused by the sound in dental treatment. First of all, based on the UGC platform text information mining and corpus analysis, we know the pain point of anxiety and fear of dental treatment voice in patients with dental treatment experience, and then interview the relevant practitioners of stomatology. According to the information obtained, a questionnaire was designed to determine the source and extent of patients' voice discomfort in dental treatment, and to analyze whether other factors were relevant. According to the above investigation results, the content of the study is determined. Secondly, based on the auditory characteristics caused by the sound propagation of air, bone, blood and muscle when the vibration acts on the teeth, a sound acquisition experiment is designed, in which the sheep head replaces the human head and the sound acquisition equipment is embedded in the head. The sound data of professional dentists who used three kinds of dental instruments to grind the teeth of the sheep head were collected. The sound data were made into 40 sound samples and were made into loudness, acuity, roughness and shaking intensity. Objective analysis of five psychoacoustics attributes of language articulation and selection of appropriate psychoacoustic attributes algorithm. Then, the subjective evaluation experiment is carried out with sound samples. 26 subjects with dental treatment experience were selected to conduct sound sample listening experiments and to evaluate dental anxiety of sound samples according to their feelings. After the evaluation results were screened, the correlation between dental anxiety and the parameters of each psychoacoustical attribute was investigated, and the subjective evaluation group and the psychoacoustic attributes that led to dental anxiety were determined. Finally, a neural network evaluation model of dental treatment sound quality was established. The three psychoacoustic attributes of dental anxiety, such as the intensity of jitter, the loudness and AI index, and the subjective evaluation of dental anxiety were taken as the input, and the subjective evaluation of dental anxiety was taken as the output. By comparing the errors of different algorithms in establishing a sound quality evaluation model for dental treatment, the BP neural network is used to model the model, and the main parameters and structure of the neural network are determined. Finally, a neural network model with predictive function for anxiety degree caused by sound in dental treatment was established, and the methods to alleviate dental anxiety symptoms were proposed from the sound point of view, and how the neural network model was applied to optimize the sound quality of dental treatment was expounded.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP183;R78
，

本文编号：2067075