請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/16618
標題: | 數位影像分析於臨床牙醫學的應用:第一部分:牙根表面積和齒槽骨高度的評估 第二部分:正顎手術後顎面軟硬組織的變化 Digital Imaging Analysis in Clinical Dentistry: Part I. Evaluation of Root Surface Area and Alveolar Ridge Height Part II. Maxillofacial Alteration after Orthognathic Surgery |
作者: | Ju-Hung Pan 潘汝弘 |
指導教授: | 郭生興 |
關鍵字: | 牙根表面積,下顎上升枝矢向分裂截骨術,下顎上升枝垂直截骨術,立體攝影測量法,測顱分析,近心段夾角, Tooth root face area,Sagittal split ramus osteotomy (SSRO),Intraoral vertical ramus osteotomy (IVRO),Stereophotogrammetry,Radiographic cephalometry,Proximal segment angulation, |
出版年 : | 2011 |
學位: | 博士 |
摘要: | 第一部分
目的:(1)以數學模型,找出牙根表面積和x光投影的關係。(2)以牙根長度或牙根投影面積估計牙科X光影像當中的牙根表面積。(3)發現牙根表面積和真實牙根厚度的關聯性,或是和牙科數位X光影像計算出的牙根厚度的關聯性。 材料及方法:(1)將牙根橫切面簡化為不同離心率的橢圓,並以每隔1度分別投影,得到來自90個方向的牙根投影資料,用於估計每個橢圓的周長,並和已知的周長比較。其次,由和橢圓長軸平行的中央光線的投影資料估計其周長,並和所有可能來自同樣的投影資料的周長比較。(2)取8顆已拔下之單牙根之牙進行三維掃瞄,直接測量牙根表面積。原始測量的點資料再進行三維重建,得到牙根之長度(估計牙根表面積),投影面積(估計牙根表面積),及真實牙根表面積。根據這三種方法,在不同牙根長度分別評估有齒槽骨支持的牙根所占的比例的精確度。(3)另外在其數位X光影像當中,分別計算牙根長度,牙根投影面積,和牙根的數位X光影像當中的各點像素值,並分別藉以估計牙根表面積比例,計算並比較誤差。(4)牙根厚度可以直接測量或由數位X光影像估計二種方法得到。比較二種由牙根厚度來估計牙根周長的方法。 結果:(1)在橢圓中,所有估計的周長都小於真實的周長。離心率減少時,最大誤差也減少。即使有相同的投影資料,當不對稱因子增加時,誤差也會增加。(2)以牙根長度估計牙根表面積,最大平均誤差是9.58%,其95%信賴區間皆正值,顯示會高估了具齒槽骨之支持比例,並有顯著差異。由CEJ向根尖每隔1 mm便分別以牙根長度,牙根投影面積進行一次估計時發現,以牙根長度估計直到在牙骨質牙釉質界線下方8 mm時,都還呈現顯著差異,而以投影面積估計時,只有在牙骨質牙釉質界線下方2 mm內,才呈現顯著差異。(3)以牙根長度估計的平均誤差為7.9%,以牙根投影面積估計的平均誤差為1.3%。(4)直接測量厚度和估計厚度的差平均在兩組當中最大值分別是-4.94% (±0.07%)和23.02% (±1.12%),而其最大的95%信賴區間分別是 (-2.82%, -1.87%) 和(-5:42%, 2.22%)。 討論和結論:(1)橢圓形的物體可以由橢圓的投影資料來近似。所以單牙根的牙根表面積可以使用投影資料而有臨床上的精確度。(2)但當邊緣骨破壞超過牙骨質牙釉質界線2 mm時,以牙根投影面積估計牙根表面積便不會有顯著差異。牙根表面積比例的函數可能是牙科數位X光影像系統的優點。(3)三維坐標量測儀可進行三維掃瞄,是一種非破壞性測量牙根表面積的方法。只使用長度資料來估計會造成顯著地不精確。單牙根牙齒的牙根表面積可以由數位牙科X光影像的投影資料估計而能有很高的精確度,所以牙科數位X光影像提供一個有效估計牙根支持比的可能性。(4)當有厚度資料時,單牙根之牙根表面積估計誤差小於5%。理論上,投影的放射線影像可以算出牙根的厚度,但估計周長的誤差可能會超過20%。如果牙科數位X光影像的估計厚度可以再更精確的話,前述誤差便可最小化。 第二部分 目的:探討正顎手術對顎面軟硬組織的變化。比較SSRO和IVRO的異同,包括下顎近心段和遠心段的位移變化,中線輪廓的變化,顎面軟組織變化的差異。以及單純下顎手術及雙顎手術顎面軟組織變化的差異。 病人和方法:於2003-2010年,在於國立臺灣大學醫學院附設醫院口腔顎面外科接受過正顎手術的病人。在術前一個月內(T1),術後一周內(T2),術後至少半年後(T3)分別取得側面及正面的測顱放射線影像,及部分病人的三維影像。 結果:(1)全體上升枝角之測量標準誤差為0.08º±0.05º。(2)T1時SSRO和IVRO的總上升枝角(TRA)並沒有差別。T2時兩組皆立即明顯地增加,且在IVRO組增加的量比在SSRO組要大。T2-T3時,IVRO組的TRA呈現劇減,但SSRO組則維持穩定。總的來說,IVRO組的TRA變化反而並不夠顯著,但SSRO組在術後半年後則維持和術前顯著不同。T3時二者的TRA並無差別。(3)上升枝傾斜度角(RIA)的變化:T1時兩組沒有差別,T2時則有明顯差異,但T3時卻又沒有差別。SSRO在T1-T2-T3測量,沒有差異。但IVRO則是先增(即上升枝近心段向後旋轉,P < 0.01)後減(即上升枝近心段向前回復,P < 0.05)。(4)condylion位置變化:在SSRO組當中,condylion由T1到T3時的位移不顯著。在IVRO組當中,condylion水平方向先向前後向後總體來說是向前。垂直方向不顯著。(5)就SSRO而言,RIA變化對於術後遠心段在垂直方向的回復關聯性最高。但就IVRO而言,condylion的位移變化,無論是水平或垂直方向,對於術後遠心段的回復關聯性最高。(6)對於側向位移而言,SSRO的同側RA和側移有關聯,對側則沒有關聯性。IVRO則未逹到顯著的關聯性。(7)測顱分析的軟組織變化主要在下臉部各點。(8)三維影像校正誤差約在0.2到0.3 mm之間。測量誤差則是0.11±0.55 mm。兩種使用的投射光色塊的尺寸對重建的三維模型的線性測量並沒有顯著的差異。但照相的角度卻會顯著地影響長度測量的結果。(9)雙顎手術之三維影像變化:在環口部位都是呈現後縮,但頦部則因向前重置。(13)單顎手術之三維影像變化:下唇以下部位十分明顯,尤其是menton的後縮最明顯,而上唇即便也有後縮,但變化則較不明顯且變異度較大,沿中線往上的各點變化就更少了。這個結果和測顱分析的結果一致。 討論:測顱要在一個客觀可重複取像的位置,其次要有實際的長度或影像的放大率。研究採用整個上升枝的外緣來代表近心段的位置顯示有很好的精準度。即便手術復原癒合重塑過程SSRO和IVRO有所不同,但術後中長期而言,二組TRA以及RIA都沒有差別。就SSRO而言,RIA變化對於術後遠心段在垂直方向的回復關聯性最高。但就IVRO而言,condylion的位移變化,無論是水平或垂直方向,對於術後遠心段的回復關聯性最高。在二維影像方法當中,只能侷限於特定平面的變化,其他平面的變化則無法得到。所以顏面軟組織的變化,就有賴於三維影像。立體攝影測量法(binocular stereophotogrammetry),其精確度決定在校正(calibration)和對應(correspondence)上。先藉由校正的過程,以直接線性轉換法計算出影像系統有關的參數。而校正的誤差由三維校正板坐標誤差來評估。然後以這些參數代入實際人臉影像,計算出其三維坐標。一個好的三維重建模型,原始影像必須要有良好的影像品質,必須有一個好的取像角度,不要有陰影,或是投射光點有不連續情形發生,並且準確地標定臉上的標定點。在臉部的曲率增加時,就要適當地增加取樣密度(sample density),以維持精確度。 Part I Purpose: (1) To correlate the root surface area with radiographic projection by a simplified mathematical model. (2) To evaluate the accuracy of supported single-root surface ratio estimated from the length and projected area of the tooth, using digital dental radiographs. (3) To find how the surface area of a single root is related to the true thickness data and the calculated thickness data from digital dental radiographs. Methods: (1) Cross-sections of a single-tooth root were simulated using ellipses with different eccentricities. Projection data from 90 directions at 1-degree intervals were obtained to estimate circumferences and were compared with the known circumference. Circumference was estimated from projection data derived from the projection of an ellipse with the central ray parallel to the long axis. The estimated circumference was compared with possible circumferences resulting from this projection data. (2) Eight extracted, single-root teeth were thee-dimensionally digitized using a contact technique for direct surface area measurement. The data were then processed, and length, projection darea, and true surface area of the root at a designated length were obtained. Based on these three measurements, the accuracy of the supported surface area ratio measurement at different lengths of the root was evaluated. (3) In addition, root length, projection area, and pixel values were then measured on digital radiographs. The accuracy of the ratio estimation of supported surface area from linear, area, and pixel values was calculated and compared. (4) The true thickness of the root was measured. The estimated circumference data were calculated from both the measured thickness and the thickness estimated from the digital image and then compared. Results: (1) All estimated circumferences are under-estimated. The largest error in each case decreased rapidly as the eccentricity decreased. The larger the asymmetric factor, the more error. (2) The largest mean error from linear estimation was 9.58%. The 95% confidence intervals were all positive, indicating that linear measurements overestimated supported ratio significantly. When analyzing the supported ration of tha alveolar bone receding from the cemento-enamel junction (CEJ) toward the apex of the root at each mm, linear estimation showed significant differences down to 8 mm, while area estimation showed significant differences only up to 2 mm. (3) The mean error from linear estimation was 7.9%; the mean error from area estimation was 1.3%. (4) The largest circumference difference mean for measured thickness and for estimated thickness was 24.94% (±0.07%) and 23.02% (±1.12%) respectively. The largest 95% confidence interval for difference means for measured thickness and for estimated thickness was (-2.82%, -1.87%) and (-5:42%, 2.22%), respectively. Conclusions: (1) The circumference of an elliptical object can be approximated from the projection data of this ellipse. Therefore, the surface area of a single tooth root may be estimated with clinically useful accuracy from the projection data. (2) A three-dimensional digitizing device could be used as a non-destructive method of measuring root surface area. Digital dental radiographs provide the potential for estimating the ratio of supported root surface efficiently. (3) When the thickness data are available, the surface area of a single-root tooth can be estimated to an error of less than 5%. Part II Purpose: To demonstrate the maxillofacial alteration after orthognathic surgery (OGS) by comparing sagittal split ramus osteotomy (SSRO) and intraoral vertical ramus osteotmy (IVRO). In addition, qualitative analysis of three-dimensional (3-D) imaging after OGS were also done. Patients and methods: (1) From 2003 to 2010, the patients who underwent OGS at the Department of Oral and Maxillofacial Surgery, National Taiwan University Hospital (NTUH). Lateral and frontal cephalometric radiographs were taken with 1 month before surgery (T1), immediately after the operation (within 1week, T2), and at the time of completion of postoperative orthodontic treatment (at least 6 month postoperatively, T3). The alteration of total ramus angles (TRA) and ramus inclination angle (RIA), the displacement of condylion, hard and soft tissue profile, side-shift, and the correlation between them were analyzed by different times in a group and between groups. (2) A stone model study was performed for the evaluation of the self-developed 3-D imaging system using modified stereophotogrammetry, which consists two cameras for synchronous capturing the image of the face, one projector for structured lighting, and a desktop or laptop computer for controlling the pattern of the structured light and processing the calibration and the correspondence by direct linear transform and disparity mapping algorithm. Soft tissue landmarks were located with 3-D coordinates, the displacement vector of the individual landmark was traced, and linear measurements between the two landmarks were recorded. Clinical OGS cases were presented as qualitative analysis of soft tissue. Results: (1) No difference in TRA between SSRO and IVRO groups in T1. In T1-T2, the TRAs increased significantly in the two groups, and more increase in TRA was noted after IVRO than SSRO. In T2-T3, TRA significantly decreased in IVRO group but remained relatively stable for SSRO. No difference in TRAs between SSRO and IVRO groups in T3. No difference in RIA between two groups in T1 and T3. It was significant difference between two groups in T2. The RIA in IVRO increased (T1-T2) and then decreased (T2-T3). Condylion kept relatively stable in SSRO group. However, condylion moved forward (T1-T2) and then backward (T2-T3). Overall, condylion moved forward (T1-T3). In SSRO, only ΔRIAT1-T2 correlated with the vertical displacement of B and pogonion in T2-T3. In IVRO, displacement of condylion correlated with displacement of distal segment. The ipsilateral ramus angle correlates with the side-shift of the distal segment in SSRO. (2) The calibration error is 0.2~0.3 mm, the measurement error is 0.11±0.55 mm. No difference in linear measurements between the two patterns of structured light. However, different orientations for capturing made significant difference in linear measurements (P < 0.05). In the patient study, in two-jaw OGS, the soft tissue profile overlying the forehead, eyes, and zygoma were minimally changed. Postoperatively, the perioral area was setback, the chin was advanced, and the lower portion of paranasal area was also slightly advanced. In one-jaw OGS, the soft tissue profile overlying the forehead, eyes, paranasal area, and zygoma were minimally changed. The region below the lower lip was setback, especially for soft tissue menton. The upper lip altered more or less. The presenting in 3D imaging agreed with that in the cephalometric radiography. Conclusion: High precision was gained when gonion was replaced with the lateral border of ramus (proximal segments) due to difficulty locating postoperatively. In 2-D imaging, such as cephalometry, it is limited to mid-sagittal plane or bi-porionic plane. Therefore, 3-D imaging is necessary for the evaluation of maxillofacial alteration after odthognathic surgery. The essentials for a good 3-D model are good original image quality, suitable capturing orientation, minimal shading from anatomic structures, free from discontinuity of structured light, accurately locating the facial landmarks. In order to keep highly accurate and precised, it is necessary to increase the sampling density on the steep curvature. The 3-D imaging system allows 3-D evaluation of the facial profile of patients with before, during, and after corrective treatment. Advances in 3-D imaging technology have made the development of practical applications in medicine, forensic, and anthropology more simple, convenient, inexpensive, and popular. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/16618 |
全文授權: | 未授權 |
顯示於系所單位: | 臨床牙醫學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-100-1.pdf 目前未授權公開取用 | 3.32 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。