面部三维运动定量分析系统的研制及面部三维测量指标研究

发布时间：2018-08-10 21:35

【摘要】：第一部分面部三维运动定量分析系统的研制和精度测量面部三维运动定量分析系统由三部分组成：数据采集部分、数据处理和结果输出部分、固定系统。数据采集部分主要由六台运动捕捉红外矩阵摄像机(采集频率60帧/秒)呈对称“品”字形排列,通过多路同步控制器保证采集的一致性。数据处理部分的核心为定制的面部运动捕捉软件。固定系统包括一个固定头架和一个综合固定椅,两者均可固定于双侧乳突及枕骨突隆。固定系统既能保证与头颅相对静止,又不影响受试者的面部活动。固定系统前端有三个固定标记点,为面部分析的参考坐标系。面部运动测量时,摄像机捕捉面部的反射光线,系统根据标定结果进行坐标转换,软件进行在线计算和离线重建。应用加速度仪进行加速度试验,该面部三维运动定量分析系统的速度测量最大误差为0.0058m/s,平均误差为0.003026m/s,加速测量误差为0.000135m/s2。第二部分面部三维测量指标的影响因素目的：评估面部三维运动定量分析系统在面瘫患者的应用,探讨测量指标的影响因素。方法：连续收集76例单侧面瘫患者。应用三维运动定量分析系统进行静态和动态测量。测量过程：清洁面部→黏贴标记点→受试者动作训练→系统设置和安置。测量结果与年龄作偏相关分析；比较男女间指标的差异；与正常值作析因方差分析。结果：1.健侧静态指标：鼻唇沟长(产0.517)与年龄呈正相关,眦角与年龄呈负相关(γ=-0.424)。除笑角外,余指标在男女间均有统计学差异(P＜0.05),且男性测量值多大于女性。与正常值比较,除笑角外均有统计学差异(P＜0.05)。2.健侧动态指标：眉弓中点(r=-0.428)、上睑中点(=r=-0.495)和鼻旁点(r=-0.471)的最大移动速度与年龄呈负相关,余指标未见明显相关性。男性眉弓中点最大移动距离和速度、用力闭眼上睑中点最大移动速度、轻闭眼上睑中点加速度均大于女性(P＜0.05)。与正常值相比：上睑中点的最大移动速度,眉弓中点最大移动速度,鼻旁点的最大移动速度和加速度,口角的最大移动距离、速度和加速度,眦角变化,均为面瘫组大于正常组(P＜0.05)。3.患侧静态指标：以神经电图(ENoG)作为控制变量作偏相关分析,鼻唇沟长(r=0.538)、左右口角间距(r=0.353)与年龄呈正相关(r=0.538),眦角(r=-0.349)与年龄呈负相关。男女间除睑裂高度外均有统计学差异(P＜0.05)。4.患侧动态指标：眉弓中点(r=-0.332)、用力闭眼上睑中点(r=-0.349)的最大移动速度与年龄呈负相关,余指标与年龄间均无明显相关性。眉弓中点最大移动速度,撅嘴时口角最大移动距离、速度和加速度为男性大于女性(P＜0.05)。结论：(1)面部三维运动定量分析系统可快速评估面瘫情况,评估过程患者无痛苦和创伤。(2)面部静态结构和肌肉运动可随着年龄增长而变化,但在面瘫时未见额外影响。(3)大部分动态和静态指标均为男性大于女性,而面瘫时这种差异可能会减小。(4)面瘫患者的健侧面部运动功能可能会代偿性增加。第三部分各面瘫评价系统相关性及预后评估研究目的：比较面部三维运动定量分析系统与其它面瘫评价系统间的相关性及各系统的预后评估价值。方法：(1)纳入50例单侧面瘫患者。应用House-Brackmann (H-B)分级系统等进行主观评价；应用肌电/诱发电位仪进行瞬目反射和面神经电图检查；应用中文版FaCE量表进行患者自我评价；应用面部三维运动定量分析系统进行三维客观测量,计算患侧与健侧的对称比。三维测量指标综合评分=0.7×D+0.3×V-A,公式中,D为双侧各标记点的最大移动距离百分比平均得分,V为双侧各个标记点的最大移动速度百分比平均得分,A为静态得分。最后比较不同评价系统结果与三维测量指标间的相关性。(2)纳入发病1月以内的急性面瘫患者,对其进行电生理测量、主观评价、三维运动定量分析及患者自我评价(FaCE).对所有患者进行跟踪随访,一月后和半年后复查。根据发病后半年面瘫的恢复情况比较各评价系统对面瘫预后的评估价值。结果：1.动态指标中,各标记点的最大移动距离与主观评价系统间相关性强,最大移动速度次之,而最大移动加速度相关性最差,且部分无统计学意义。三维测量综合评分与各主观评价系统得分之间相关系数分别为-0.630(HB分级)、-0.728(面神经分级量表2.0)、0.697(Sunnybrook评分)和0.617(Fisch评分),P值均＜0.05。FaCE量表仅反映患者面部运动的评价条目与部分指标有相关性。三维动态测量结果与瞬目反射结果的相关性强于同ENoG的相关性。面部三维测量指标与瞬目反射振幅的相关性明显强于潜伏期。与ENoG进行相关性比较,各标记点的最大移动距离指标相关性差异较大。2.共纳入37例单侧面瘫患者,随访期间共有6例(16%)删失。多元回归分析筛选最佳预后评价指标：(1)面瘫发病30内ENoG为最佳预后指标,其预测公式y=113.527-60.558x,公式中y为半年后SFGS评分,x为ENoG估算的面神经变性比例；(2)面瘫31至60天,SFGS得分和三维测量结果为最佳预后指标,其预测公式为y=19.202+0.557x1+0.531x2,x1为SFGS得分,x2为三维测量指标得分。结论：(1)面部三维运动定量分析结果与传统的主观评价结果及电生理检查结果间有较好的相关性,且可以提供更为细致的量化指标。(2)面瘫急性期,电生理检查仍然是预测面瘫预后的最佳指标；面瘫超过1月,联合主观评价和三维运动定量分析可提供最佳的预后评估。
[Abstract]:The first part is the development of the quantitative analysis system for facial 3D motion and its precision measurement. The quantitative analysis system for facial 3D motion consists of three parts: data acquisition, data processing and output, and fixed system. The key part of the data processing section is a customized facial motion capture software. The fixing system includes a fixed headframe and a comprehensive fixed chair, both of which can be fixed to the bilateral mastoid and occipital protuberances. The fixing system ensures relative stillness with the skull without shadow. The front end of the fixing system has three fixed markers for the reference coordinate system of facial analysis. In facial motion measurement, the camera captures the reflected light of the face, the system carries out coordinate transformation according to the calibration results, the software carries out on-line calculation and off-line reconstruction. The maximum error of velocity measurement is 0.0058m/s, the average error is 0.003026m/s, and the acceleration error is 0.000135m/s 2. Part 2: The influencing factors of facial three-dimensional measurement indicators Objective: To evaluate the application of facial three-dimensional motion quantitative analysis system in facial paralysis patients and explore the influencing factors of measurement indicators. Seventy-six patients with unilateral paralysis were collected in succession.Static and dynamic measurements were performed using a three-dimensional motion quantitative analysis system.The measurement process was as follows:clean face_sticking markers_subject movement training_system setting and placement.Partial correlation analysis was made between the measurement results and age.Differences between male and female indicators were compared.Factorial variance scores were calculated with normal values. Results: 1. Static indices of healthy side: length of nasolabial sulcus (0.517) was positively correlated with age, and canthus angle was negatively correlated with age (gamma = - 0.424). Except for laughing angle, the remaining indices were statistically different between men and women (P < 0.05), and the measured values of men were more than those of women. There were statistical differences between healthy side and normal value except laughing angle (P < 0.05). The maximum velocity of eyebrow arch midpoint (r = - 0.428), upper eyelid midpoint (= r = - 0.495) and paranasal point (r = - 0.471) was negatively correlated with age, but no significant correlation was found with other parameters. Constant comparison: the maximum moving speed of the middle point of the upper eyelid, the maximum moving speed and acceleration of the midpoint of the eyebrow arch, the maximum moving speed and acceleration of the paranasal point, the maximum moving distance of the mouth angle, the speed and acceleration, and the change of the canthus angle were all greater in the group of facial paralysis than in the normal group (P < 0.05). 3. Static index of the affected side: EEG was used as the control variable for partial correlation. The results showed that the length of nasolabial sulcus (r = 0.538), the distance between the left and right corners of mouth (r = 0.353) were positively correlated with age (r = 0.538), and the canthus angle (r = - 0.349) was negatively correlated with age. Conclusion: (1) The quantitative analysis system of facial three-dimensional movement can quickly evaluate the situation of facial paralysis, and can evaluate the process of patients without pain and trauma. (2) Facial static knot. (3) Most of the dynamic and static indexes were higher in males than in females, but this difference may be reduced in facial paralysis. (4) The motor function of healthy side of patients with facial paralysis may increase compensatively. Objective: To compare the correlation between facial three-dimensional motion quantitative analysis system and other facial paralysis evaluation systems and the prognostic value of each system.Methods: (1) 50 patients with unilateral paralysis were included. The patients were assessed by electrogram, the Chinese version of FaCE scale was used for self-evaluation, and the facial three-dimensional motion quantitative analysis system was used for three-dimensional objective measurement to calculate the symmetry between the affected side and the healthy side. At last, the correlation between the results of different evaluation systems and three-dimensional measurements was compared. (2) Patients with acute facial paralysis within 1 month of onset were included in the study, and their electrophysiological measurements, subjective evaluation, three-dimensional motion quantitative analysis and patient self-evaluation (FaCE) were performed. All patients were followed up for 1 month and 6 months. The evaluation value of each evaluation system was compared according to the recovery of facial paralysis after the onset of the disease. The correlation coefficients between the three-dimensional measurement score and the subjective evaluation system scores were - 0.630 (HB grade), - 0.728 (Facial Nerve Scale 2.0), 0.697 (Sunnybrook score) and 0.617 (Fisch score). The P values were < 0.05. The correlation between 3-D dynamic measurements and blink reflex was stronger than that of ENoG. The correlation between 3-D facial measurements and blink reflex amplitude was stronger than that of incubation period. Compared with ENoG, the correlation between the maximum moving distance of each marker was different. 2. 37 patients with unilateral paralysis were included. Multivariate regression analysis screened the best prognostic indicators: (1) ENoG was the best prognostic indicator within 30 days of facial paralysis, and its predictive formula y = 113.527-60.558x, in which y was the SFGS score after six months, and X was the ratio of facial nerve degeneration estimated by ENoG; (2) SFGS score and three-dimensional measurement were the best prognostic indicators within 31 to 60 days of facial paralysis. The predictive formula was y = 19.202 + 0.557x1 + 0.531x2, X1 was SFGS score, and X2 was three-dimensional measurement index score. Conclusion: (1) There was a good correlation between quantitative analysis of facial three-dimensional motion and traditional subjective evaluation and electrophysiological examination results, and it could provide more detailed quantitative indicators. (2) In acute facial paralysis, electrical measurements were performed. Physiological examination is still the best index for predicting the prognosis of facial paralysis. If facial paralysis lasts for more than one month, the combination of subjective evaluation and three-dimensional quantitative exercise analysis can provide the best prognostic evaluation.
【学位授予单位】：北京协和医学院
【学位级别】：博士
【学位授予年份】：2016
【分类号】：R745.12

【相似文献】