本研究的目的是提出一個以3D頭型資料庫來設計呼吸半面罩的方法，目標是達到呼吸面罩貼合介面曲線與臉部的緊密。本研究所應用的人體計測資料庫總共270受測者，其中包括135位男性及135位女性。在勞工人口中依比例抽樣，將身高分成極小、小、中、大、極大五個級距及體重分為胖、中、瘦三種級距。由Gemini人體掃描儀的六個測頭取得頭部的表面資料。
設計方法包含兩個階段：擷取3D貼合介面曲線以及建構呼吸面罩的面罩體。而擷取3D貼合介面曲線又分為三個步驟：先依據臉長分群、畫出2D面罩投影曲線，以及取得3D貼合介面曲線。在取得所有受測者的臉長後，發現尺寸值分佈在82.6 mm~ 115.8 mm 之間，差距範圍達33.2 mm，所以我們將受測者分為兩群：“size S”為82.6 mm ~98.2 mm ，“size L”為98.2 mm~ 115.8 mm。在各群體中，我們將在臉長此尺寸值最接近該群體平均值的受測者作為代表人，再以此代表人的臉型來決定2D投影曲線。2D投影曲線為一封閉曲線，從代表人正面來看，曲線是從鼻中點為起點，之後往左右及向下穿越下眼框區的部位，沿著頰區的咬肌前緣最後在下頦區的地方以下頦點為曲線終點。接下來，將此2D投影的封閉曲線投影至每位受測者的臉上，可以得到3D的貼合介面曲線。並且，各群體中的3D貼合介面曲線會在鼻樑點與下頦點的地方予以重疊。從側面來看，所有這些重疊的曲線可以表示不同的臉型深度，曲線散佈的程度表示臉型在深度上的差異的程度，可以作為下一步要尺碼分群的依據。根據以往的經驗，在鼻樑骨及下頦的區域可以容許10 mm以內的差異，而臉頰的區域可以容許15 mm的差異，以此作為尺碼分群的尺度標準。根據此標準，研究結果為將原本的”size S”及 “size L”群體中因為有輪廓較深陷的臉型及輪廓較淺的臉型，又再各自細分兩群體，以“SS”、“SD”、“LS”以及“LD”來表示。此外，在各群體中，最深陷的輪廓曲線即為該群體的尺碼標準人。
第二階段為建構呼吸面罩的面罩體，依據標準人的貼合介面曲線以及中間矢狀面輪廓線。面罩體的中間矢狀面輪廓線是一條簡單的曲線，從鼻中點的地方向上提高10mm、鼻尖點的地方向上提高10mm，以及下頦點的地方高出10mm。接下來，我們建構面罩體的水平參考曲線，有一條曲線通過鼻尖點高10mm處，另外兩條則是通過貼合介面曲線在鼻中點及下頦點的地方。之後，照著以上類似的方法，每隔10mm建構一條水平輪廓線，所有的這些水平輪廓線是用來建構出一個連續的3D呼吸面罩的面罩體表面。面罩體與臉部之間的空間稱為滯留腔。此滯留腔在SS、SD、LS以及 LD四種尺碼中滯留腔體積大小分別為213.3 cm3、235.8 cm3、256.2 cm3及285.1 cm3。這些體積都小於市售的呼吸面罩滯留腔體積（平均值294.3 cm3，標準差28.5 cm3）。

This study proposes a design method for half-face respiratory masks based on an existing 3-D head database, aiming at the air-tightness of faceseal. The database contains a total of 270 subjects, 135 males and 135 females, which were drawn from Taiwanese worker population based on the body size distribution of two previous anthropometrical surveys. The distribution was classified into 5 staure heights and 3 body weights. The 3-D surface of the head was measured using a six-head Gemini scanner.
The design method consists of two phases: extraction of 3-D faceseal contour and contruction of mask body. The extraction of 3-D faceseal contour proceeded in 3 steps: sizing on face-length, drawing of 2-D contour, and extraction of 3-D faceseal contour. Since the face-lengths range between 82.6 mm and 115.8 mm, with a range difference of 33.2 mm, so they were classified into two size groups: “size S” is from 82.6 mm to 98.2 mm and “size L” is from 98.2 mm to 115.8 mm. Within each size group, the one whose face-length is the most close to the mean was chosen as the representatives samples to determine the representative 2-D contour. The representative 2-D contour is a closed contour which is drawn on the front view of the face of the representative sample, its top begins with mid-nasal point, then it curves outward (to the right and left) and downward and traverses the infraorbital regions, and then to the outer-skirt of bunasolabial sulcus of the buccol regions, and finally along the chin and closes at mentum. Then, this 2-D contour was projected onto each subject to obtain a 3-D faceseal contour. All the 3-D faceseal contours within each size group were registered and superimposed on the menton and mid-nasal. In side view, all of these superimposed faceseal contours exhibits in different depths. The dispersion in depth is a measure of the degree of face difference, and is used for further sizing consideration. Based on past experiences, a difference of 10 mm difference in the bonny nasal and mentum area, and 15 mm difference in the cheek fleshy area are considered as the cutoff criteria for size grouping. Based on these criteria, both size A and size B were required to further classified into two groups, namely “SS”, “SD”, and “LS”, “LD”, in respected to the shallower faces and the deeper faces. Within each size group, the deepest one was chosen as the final representative faceseal contour.
The construction of mask body is based on the faceseal and the mid-saggittal contour of the typical sample. The central ridge of the mask was determined based on the mid-saggittal contour of the face. It is a simple curved line elevates above the mid-saggittal contoure with an elevations of 10 mm, 20 mm and 10 mm each at mid-nasal, nasal tip and menton. Next, a horizontal reference contour of the mask was drawn, which passes the 20 mm elevation point at nasal tip and with two end points terminate on the faceseal contour. Then in a similar manner, a series of horizontal contours were drawn at 10 mm interval. All horizontal contours were further interpolated to form a continuous 3-D surface. This surface is the 3-D shape of the mask body. The space between the mask body and the face is the dead space. The dead space volume each for SS, SD, LS, and LD size was 213.3 cm3、235.8 cm3、256.2 cm3 and 285.1 cm3respectively, which is smaller than most existing masks (with a mean of 294.3 cm3 and S.D of 28.5 cm3).

(1.) Mary K. Salazar, Catherine Connon, Timothy K. Takaro, Nancy Beaudet, Scott Barnhart, 2001;”An Evaluation of Factors affecting Hazardous Waste workers’ Use of Respiratory Protective Equipment.”, AIHA Journal 62:236-245
(2.) Riedar K. Oestentad, H. Kenneth Dillion, and Lara L. Perkins,1990;”Distribution of Faceseal Leak Sites on a Half-Mask Respirator and their Association with Facial Dimensions.” , AIHA Journal 51:285-290
(3.) Riedar K. Oestentad, Jimmy L. Perkins, and Vernon E. Rose, 1990;”Identification of Faceseal Leak Sites on a Half-Mask Respirator.”, AIHA Journal 51:280-284
(4.) Don-Hee Han, Kook-lyeol Choi, 2003;”Facial Dimension and Predictors of Fit for Half-Mask Respirators in Koreans”, AIHA Journal 64:815-822
(5.) David M. Caretti, Jeffrey A. Whitley, 1998;”Exercise performance during inspiratory resistance breathing under exhaustive constant load work”, Ergonomics, Vol.41, No.4, 501-511
(6.) EN 149: 1991, “Respiratory protective devices- Filtering half masks to protect against particles – Requirements, testing, marking”
(7.) Joomyung RHI, Donhee HAN, “A Design Process for Half - face piece Respirators with Face Digitizing Method”
(8.) William J. Brazile, Roy M. Buchan, Del R. Sandfort, Walter Melvin, Janet A. Johnson, Michael Charney,1998; “Respirator Fit and Facial Dimension of Two minority Group”, Appl. Occup. Environ. Hyg. 13(4)
(9.) Hyunwook KIM, Don-Hee HAN, Young-Man ROH, Kangyoon KIM and Yong-Gyu PARK, 2003;”facial Anthropometric dimension of Koreans and their Associations with Fit of Quarter-Mask Respirators“, Industrial Health, 41, 8-18
(10.) Joyce E. Rebar, Arthur T. Johnson, Estelle Russek-Cohen, David M. Caretti, and William H. Scott, 2004;”Effect of Differing Facial Characteristics on Breathing Resistance inside a Respirator Mask”, Journal of Ocuupational and Environmental Hygiene, 1: 343-348
(11.) Makiko Kouchi and Masaaki Mochimaru, 2004;”Analysis of 3D face forms for proper sizing and CAD of spectacle frames”, Ergonomics, Nov. vol. 47, No. 14, 1499-1516
(12.) Arthur T. Johnson, William H. Scott, Christopher G. Lausted, Karen M. Coyne, Manjit S. Sahota, Monique M. Johnson, 2000;’’Effect of External Dead Volume on Performance While Wearing a Respirator ”, AIHA Journal 61:678-684
(13.) NIOSH, 1973;”Selection of Respirator Test Panels Representative of U.S Adult Facial Sizes”
(14.) 陳萬春、葉文裕、陳友剛, 1999; “佩戴者臉型對半面體呼吸防護具面體密合係數影響探討”, 行政院勞工委員會勞工安全衛生研究所
(15.) “如何正確使用呼吸防護具”, 行政院勞工委員會勞工安全衛生處