research article partially automated method for...

Research ArticlePartially Automated Method for Localizing StandardizedAcupuncture Points on the Heads of Digital Human Models

Jungdae Kim12 and Dae-In Kang2

1Nano Primo Research Center Advanced Institutes of Convergence Technology Seoul National UniversitySuwon 443-270 Republic of Korea2Pharmacopuncture Medical Research Center Korean Pharmacopuncture Institute Seoul 157-801 Republic of Korea

Correspondence should be addressed to Jungdae Kim tojdkimgmailcom

Received 27 February 2015 Revised 6 May 2015 Accepted 13 May 2015

Academic Editor Vitaly Napadow

Copyright copy 2015 J Kim and D-I Kang This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Having modernized imaging tools for precise positioning of acupuncture points over the human body where the traditionaltherapeutic method is applied is essential For that reason we suggest a more systematic positioning method that uses X-raycomputer tomographic images to precisely position acupoints Digital Korean human data were obtained to construct three-dimensional head-skin and skull surface models of six individuals Depending on the method used to pinpoint the positions ofthe acupoints every acupoint was classified into one of three types anatomical points proportional points and morphologicalpoints A computational algorithm and procedure were developed for partial automation of the positioningThe anatomical pointswere selected by using the structural characteristics of the skin surface and skull The proportional points were calculated fromthe positions of the anatomical points The morphological points were also calculated by using some control points related to theconnections between the source and the target models All the acupoints on the heads of the six individual were displayed on three-dimensional computer graphical image models This method may be helpful for developing more accurate experimental designsand for providing more quantitative volumetric methods for performing analyses in acupuncture-related research

1 Introduction

Regardless of regional differences between western and east-ern medicine a large amount of medical data has beenaccumulated over the centuries in the qualitative forms oftexts and pictures [1] Thanks to the fast central process unitsand vast sizes of computer memories a new representation ofmedical knowledge especially knowledge concerning humananatomy and function can now be made [2] Nowadays thequest to modernize medicine seems to be more demandingon the eastern side than on the western side [3] As oneof the core therapeutic modalities in many Asian countriesacupuncture is still a subject that is actively being studiedby with various modernized tools and advanced techniquessuch as functional magnetic resonance imaging and positronemission tomography [4ndash6]

In countries such as China Japan and Korea acupunc-ture has been practiced for more than 2500 years and hasnow become a global therapeutic method used across theworld Clinical studies have shown promising results forthe efficacy of acupuncture in for example reducing bothpostoperative and chemotherapy nausea and vomiting inadults aswell as postoperative dental pain [7] Although basicresearch on acupuncture has led to considerable progressover the past decades its underlying mechanism is still anabstruse subject [8] Because of the increasing demand forstandardization of acupuncture point locations the WorldHealthOrganizationWesternPacificRegionalOffice initiatedprojects in the early 1990s to reach a consensus on thoselocations The WHO presented the general guidelines foracupuncture point locations in the forms of texts and figuresand it stipulated the methodology for locating acupuncture

Hindawi Publishing CorporationEvidence-Based Complementary and Alternative MedicineVolume 2015 Article ID 483805 12 pageshttpdxdoiorg1011552015483805

2 Evidence-Based Complementary and Alternative Medicine

Table 1 Characteristics of six subjects for the 3D digital CT images

Subject Gender Age (years) Height (cm) Weight (kg) BMI (kgm2) Category1 Male 26 170 66 2284 Normal2 Male 37 175 59 1927 Underweight3 Male 39 170 75 2734 Obese4 Female 30 158 55 2020 Normal5 Female 28 161 48 1852 Underweight6 Female 43 162 64 2401 Overweight

Obese man Underweight manNormal man

Underweight womanNormal womanOverweight woman

Figure 1 Three-dimensional reconstructed surface models of Korean adults with various body types

points on the surface of the human body as well as thelocations of 361 standard acupuncture points This standardestablished by theWHOmay be applied in teaching researchclinical service preparation of publications and academicexchanges involving acupuncture [9]

In more recent studies on acupoint locations researchersbegan to use more advanced equipment such as X-raymachines As a convenient method for locating acupointsthe cun measurement methods have been widely used inthe practice of acupuncture However the traditional cunmeasurement methods have been criticized for their lack ofreliability In one study a comparison of two different locationmethods was done by using dual-energy X-ray absorptiom-etry to measure the soft tissue and the bone mass indepen-dently [10] Another study used computed tomography (CT)to provide a metric description of acupuncture points inthe lumbar region and to give their relation to individualanatomical landmarks and structures [11] Another studyused X-ray radiography to provide experimental evidence tostandardize the location of the acupoint in the hand [12]

Synchrotron radiation phase-contrast X-ray CT was alsoemployed to investigate the three-dimensional (3D) topo-graphic structures of acupuncture points [13] The results ofanother study suggested that biomedical information aboutacupuncture treatment could be visualized in the form of adata-driven 3D acupuncture point system [14]

The accurate localization of acupoints is a key issue inacupuncture research For more precise scientific researchand development on acupuncture therapy having definitionson how to localize acupoints by using a computer-basedpictorial representation of the human body is critical Amethod for localizing the acupoints on the head of a virtualbody by using segmentation and a 3D visualization of theVOXEL-MAN software system was reported [15] Our initialresearch on positioning all the 361 standardized acupuncturepoints was done with the digital data from a healthy Koreanmale with a normal body shape [16] The cross-sectionalimages generated by X-ray CT were used for generating the3D virtual models for the bones and the skinrsquos surface ofthe entire human body The volumetric 3D acupoint model

Evidence-Based Complementary and Alternative Medicine 3

Table 2 Categorized acupoints on the head The acronyms for theacupoint names come from the names of meridians GV (governorvessel) GB (gall bladder) ST (stomach) TE (triple energizer) BL(bladder) CV (conception vessel) LI (large intestine) and SI (smallintestine) The values in the parentheses are the total numbers ofpoints in the categories lowastThese points are not standard acupuncturepoints but are introduced for convenience

Category Acupoints on the head

Anatomicalpoints (34)

GV15 GV16 GV17 GV25 GV26 GV27 (6)GB1 GB2 GB3 GB7 GB8 GB9 GB12GB20 (8)ST1 ST2 ST3 ST4 ST5 ST6 ST7 (7)TE17 TE20 TE21 TE22 TE23 (5)BL1 BL2 (2)CV23 CV24 (2)LI19 LI20 (2)SI17 SI18 (2)Pupillowast Yintanglowast TOPlowast (3)

Proportionalpoints (24)

GV18 GV19 GV20 GV21 GV22 GV23GV24 (7)GB13 GB14 GB15 GB16 GB17 GB18 GB19(7)ST8 (1)BL3 BL4 BL5 BL6 BL7 BL8 BL9 BL10 (8)SI19 (1)

Morphologicalpoints (7)

GB4 GB5 GB6 GB10 GB11 (5)TE18 TE19 (2)

was developed based on projective 2D descriptions of thestandard acupuncture points

According to the general guidelines suggested by theWHO in the western pacific region three methods are usedfor determining acupuncture point locations the anatomicallandmark method the proportional bone (skeletal) mea-surement method and the finger-cun measurement method[7] The anatomical landmark method utilizes some char-acteristics on the surface of the body that may be fixed ormovable such as protuberances or depressions formed byjoints and muscles The proportional bone (skeletal) mea-surement method also uses landmarks on the bodyrsquos surfacethat is primarily joints tomeasure the lengths and thewidthsof various parts of the body The finger-cun measurementmethod refers to the proportional measurement method forlocating acupuncture points based on the size of the fingersof the person to be measured In some cases these methodsare used together for complementary decisions on the sameacupoint

In this study we developed systematic procedures andalgorithms for positioning acupoints on the heads of sixindividuals All the acupoints on the head were categorizedinto one of three types anatomical points proportionalpoints and morphological points The anatomical acupointsare determined by using corresponding anatomical char-acteristics from the skinrsquos surface and bone structure Theproportional acupoints are calculated by using prescribed

x

yFrontBackRight

Left

Up

Down

(0 0 0)(511 0 0)

(0 511 0)

(0 0 300)

(511 511 300)

z

z998400

y998400

x998400

(CMxCMyCMz)

Figure 2 Coordinate systems based on the CT images of the headFor convenient numerical calculations the 119909- 119910- and 119911-axes werechosen to access every voxel from the head skin and skeleton Themodel frame (1199091015840 1199101015840 1199111015840)was introduced by translating (119909 119910 119911) to thecenter of mass of the model and scaling down

proportional numbers between the anatomical points Someremaining acupoints only have descriptive definitions fortheir positions in which case the morphological technique isintroduced among the individuals The traditional methodsfor the acupoint localization are based on measurementsalong the surface of the body with cun Here we do not usecun units which may vary in length for different parts of thebody One of basic assumptions for this study with the head isthat the shape of the human head is approximately spherical

2 Materials and Methods

21 Digital Korean Human Data Digital Korean CT datafor six human beings were obtained from the KoreanInstitute of Science and Technology Information (KISTIhttpdkkistirekr) The CT images of the entire body weretaken from three men and three women whose body sizeswere measured Based on the body mass index (BMI) thesix individuals were categorized as underweight (18 lt BMI lt20 kgm2) normal (20 lt BMI lt 24 kgm2) overweight (24 ltBMI lt 26 kgm2) and obese (BMI gt 26 kgm2) The gendersages heights and weights of the six individuals are shown inTable 1 A stack of CT images with 512 times 512 pixel resolutionwas taken from head to toe in 1mm depth intervals and theresults were saved in the DICOM format

22 Procedure for Surface Reconstruction of 3D HumanModels The procedure for reconstructing the surfaces of theskin and the skull from the stack of 2D images of the sixindividual is basically the same as that used in [16] for asingle person with a normal body shape In brief first theDICOM images are converted to 8-bit BMP format because ofmemory issues Then the images for the surfaces of the skinand bone were put into binary format by using the proper


Table 3 Descriptions of the locations of the anatomical acupuncture points and landmarks lowastThese landmarks are additionally positionedfor localization of the acupoints

Acupointslandmarks Surface models Description of the position for the landmark

GV15 Bone In the depression superior to the spinous process of the second cervical vertebra(C2) on the posterior median line

GV16 Bone Directly inferior to the external occipital protuberanceGV17 Bone External occipital protuberanceGV25 Skin Tip of the noseGV26 Skin Midpoint of the philtrum midlineGV27 Skin Midpoint of the tubercle of the upper lipGB1 Skin and bone Outer canthus of the eyeGB2 Bone Depression between the intertragic notch and the condylar process of the mandibleGB3 Bone Depression superior to the midpoint of the zygomatic arch

GB7 Skin Junction of the vertical line of the posterior border of the temple hairline and thehorizontal line of the apex of the auricle

GB8 Skin Directly superior to the auricular apexGB9 Skin Directly superior to the posterior border of the auricular rootGB12 Bone Depression posteroinferior to the mastoid process

GB20 Bone Inferior to the occipital bone in the depression between the origins of thesternocleidomastoid and the trapezius muscles

ST1 Skin and bone Between the eyeball and the infraorbital margin directly inferior to the pupilST2 Bone In the infraorbital foramen

ST3 Skin Directly inferior to the pupil at the same level as the inferior border of the ala of thenose

ST4 Skin The angle of the mouth

ST5 Bone Anterior to the angle of the mandible in the depression anterior to the masseterattachment

ST6 Bone Angle of the mandible

ST7 Bone Depression between the midpoint of the inferior border of the zygomatic arch andthe mandibular notch

TE17 Skin Posterior to the ear lobe in the depression anterior to the inferior end of themastoid process

TE20 Skin Auricular apex

TE21 Skin and bone In the depression between the supratragic notch and the condylar process of themandible

TE22 Skin Anterior to the auricular root posterior to the superficial temporal arteryTE23 Skin In the depression at the lateral end of the eyebrow (it is superior to GB1)

BL1 Skin and bone In the depression between the superomedial parts of the inner canthus of the eyeand the medial wall of the orbit

BL2 Skin In the depression at the medial end of the eyebrow

CV23 Bone In the anterior region of the neck superior to the superior border of the thyroidcartilage in the depression superior to the hyoid bone on the anterior median line

CV24 Skin In the depression in the center of the mentolabial sulcusLI19 Skin At the same level as the midpoint of the philtrum

LI20 Skin In the nasolabial sulcus at the same level as the midpoint of the lateral border of theala of the nose

SI17 Bone Posterior to the angle of the mandible in the depression anterior to thesternocleidomastoid muscle

SI18 Skin and bone Inferior to the zygomatic bone in the depression directly inferior to the outercanthus of the eye

Pupillowast Skin Center line of the pupilYintanglowast Skin Midpoint between the eyebrowsTOPlowast Skin Top of the head


Yintang

GV17

TE20

TOP

Figure 3 The baselines or planes for the head are determined by using four landmarks Yintang and GV17 (the red line the median plane)and TE20 and TOP (the green line the frontal plane) A transverse plane (the blue line) is determined by using a plane orthogonal to theimaginary line connecting the center of the left and the right TE20 points to the TOP The four landmarks can be positioned by using theiranatomical descriptions on the skin and the skeleton Yintang is the point between the eyebrow and TOP is the highest point of the headTheacupoint GV17 is in the depression superior to the external occipital protuberance and the acupoint TE20 is just superior to the auricularapex on both the left and the right sides

GV17

GV18

GV19

GV20GV21 GV22

GV23GV24

YintangTE20CP15

1515

15 15 10 05

35

GV25GV26GV27

Figure 4 Some proportional acupoints fromGV18 to GV24 on the midsagittal plane that were obtained by dividing the angle between GV17and Yintang by appropriate proportionality constants based on the guidelines of the WHO for standard acupuncture positioning


ST8

GB4

GB5

GB6

GB7

GB9

TE20

GB12 TE17

GB10

GB11

TE19

TE18

14

14

141413

1313

13 13

13

(a)

P

Q

RX

Standard source Target

Q998400

P998400

L equiv (P998400minus X) minus (P minus R) + (Q

998400minus X) minus (Q minus R)

(b)

Figure 5 Morphological acupoints on the right side of the head (a) and the distance formula for the unknown position 119883 with the knownpositions 11987510158401198761015840 from the target model and the 119875119876 and 119877 from the standard source model The position119883 is adjusted in such a way that thedistance 119871 is minimized

threshold values for the pixel intensity 10 for the skin and110 for the skull Unnecessary holes were filled by using a3Dbinary dilation subroutine and boundarieswere extractedfrom the objects in the binary images The boundary for theskinrsquos surface or for the bonersquos surface was isolated and a3D dataset was made for the skinrsquos surface or for the bonersquossurface by using the Marching Cube algorithm through theboundary Actually the Marching Cube algorithm can beapplied to the 8-bit images directly However a procedure forbinarizing and deleting unnecessary parts is essential in orderto obtainmodel data that are more compactThe 3D image ofthe surface can be made smoother by averaging the normalvectors at every vertex during the triangulation and by usingOpenGL software (httpswwwopenglorg) to present the3D surfacemodels as shown in Figure 1 for the skinrsquos surfaceThe two surfacemodels one from the skin and the other fromthe skull of an individual can be combined in such a way thattheir centers of mass coincide The skull can be seen throughthe skin by computer-graphically giving it some opacity

23 Definition for the Reference Frames and the AnatomicalPlanes If any point in the 3D digital models is to bedescribed proper reference frames should be introduced Anatural frame called the stack frame (119909 119910 119911) can be fixed byusing the three indices (119894 119895 119896) along the width length andheight of the stack of 2D images as shown in Figure 2 Formore a model-dependent frame the model frame (1199091015840 1199101015840 1199111015840)can be obtained by translating and scaling the stack frame asfollows

(1199091015840

1199101015840

1199111015840

) = 120572 (119909 minusCM119909 119910 minusCM

119910 119911 minusCM

119911) (1)

where the position of the center of mass position for a givenstack of binary images 119868bin

119894119895119896 is CM

119909= sum119894119895119896119868bin119894119895119896119894 CM

119910=

sum119894119895119896119868bin119894119895119896119895 CM

119911= sum119894119895119896119868bin119894119895119896119896 and we have set 120572 = 4512

for convenience

The anatomical planes for the head can be defined bychoosing some obvious points anatomically The sagittalplane that is the midsagittal plane or median plane moreprecisely is determined by fixing three points in the headYintang (midpoint between the eyebrows) TOP (top of thehead) and the acupoint GV17 (external occipital protuber-ance in the back of the head) as shown in Figure 3 Anypoint 119875 which resides in the sagittal plane should satisfy thefollowing equation

[(GV17minusYintang) times (TOPminusYintang)]

sdot (119901 minusYintang) = 0(2)

where all the points in the equation are considered asvectors with triplet values corresponding to the vectorrsquos threecomponents and the cross and the dot in the equation meanthe cross product and the scalar product respectively Acoronal plane in the head can be defined by choosing threepoints on the head two acupoints TE20 at the auricular apexjust above the left and the right ears and the TOP Any point119875 in the coronal plane should satisfy the condition

[(TE20left minusTOP) times (TE20right minusTOP)] sdot (119901 minusTOP)

= 0(3)

The transverse plane at any level can be defined in such away that the plane is perpendicular to the common axis thatresides in the sagittal and the coronal planes simultaneously

24 Categorization for Acupuncture Points All the acupointson the head can be categorized into one of three typesdepending on how they are located as shown in Table 2In addition to the standard acupuncture points three morepoints Pupil Yintang and TOP are introduced for position-ing the acupointsThe anatomical acupoints have anatomicaldescriptions that aremore or less clear based on the structures


GB17GB16GB15

GB14GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8GB18

BL9

BL10

GB19

GB20

GB7

(a)GB17GB16

GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2

BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8 GB18

BL9

BL10

GB19

GB20

GB7

(b)

Figure 6 Continued


GB17GB16GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4

BL3

BL2

BL1

GV22GV23

GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8

GB18

BL9

BL10

GB19

GB20

GB7

(c)

Figure 6 Various views of the 3Dmodel of an (a) obese a (b) normal and an (c) underweight man with positioned acupoints on their heads

of the skin or the skull Brief descriptions on the acupoints onthe head are presented in Table 3 The corresponding surfacemodels for their positioning are also given in the table

For the proportional acupoints we first introduce themidpoint between the left and the right TE20 as follows

TE20CP equiv(TE20left + TE20right)

2

(4)

This imaginary point is introduced just for arithmetic con-venience In order to obtain the conditions for positioningthe proportional acupoints let us define the angle 120579

0made by

the three points Yintang TE20CP and GV17 with the vertexpoint being TE20CP as follows

1205790= cosminus1(Yintang minus TE20CP) sdot (GV17 minus TE20CP)1003816100381610038161003816Yintang minus TE20CP100381610038161003816

1003816|GV17 minus TE20CP|

(5)

We can also calculate two angles depending on the point 119875which is one of the voxels for the surface of the skin on thehead If the point119875 is on the skinrsquos surface in the sagittal planewe define the angle as follows

1205791 = cosminus1 (119901 minus TE20CP) sdot (GV17 minus TE20CP)1003816100381610038161003816119901 minus TE20CP100381610038161003816


(6)

If the point 119875 is on the skinrsquos surface in the transverse planewe define the angle as follows



(7)

The conditions for the proportional acupoints are displayedin Table 4 and the pictorial descriptions of the conditions areshown in Figure 4

According the standard acupuncture positioning in [9]some acupoints are described on the curved lines betweentwo acupoints as shown in Figure 5(a) and Table 5 Forexample GB4 GB5 and GB6 are positioned along thecurved line between ST8 and GB7 In this case we use themorphological method for positioning the acupoints GB4GB5 and GB6 by using the previously-known points ST8and GB7 as two control points Precise information aboutpositioning the morphological acupoints should be preparedfrom a ldquostandardrdquo source We took that ldquostandardrdquo sourceto be the 3D model of the normal man as displayed inFigure 5(a) and we obtained the positions of the anatomicaland proportional acupoints in advance Figure 5(b) showsthat the target point 119883 between the two control points 1198751015840and 1198761015840 can be calculated in such a way that the sum ofthe distances between the two control points is minimized


ST8

GB4

GB5

GB6GB7

GB9

GB10

GB11

GB12

TE20

TE17

TE19TE18

Figure 7 Morphological acupoints and the lines connecting the two control points through them The upper left corner shows themorphological acupoints on the head of anormal man they exactly match the same positionings on our ldquostandardrdquo source model for themorphological technique In the other models the shapes of the curves are shown to be slightly different depending on the individuals

compared to the distances of the two control points from thestandard source

3 Results

All the acupoints on the head above the neck were calculatedfollowing the above procedure We positioned the acupointson the head models taken from the 3D digital CT imagesFigures 6(a) 6(b) and 6(c) show various views of theheads of the obese the normal and the underweight manrespectively The total of 65 standard acupoints on the headcould be classified into 34 anatomical points 24 proportionalpoints and 7 morphological points The positions of theanatomical acupoints were described and those descriptionsare summarized in Table 3 Based on the descriptions theanatomical acupoints were positioned manually and theirpositions were saved for further calculations on positioningthe proportional and the morphological acupoints

The analytic and pictorial descriptions of the proportionalacupoints are presented in Table 4 and Figure 4 respectivelyIf we assume that the shape of the surface of the headis spherical the distance between consecutive acupoints inthe midsagittal planes can be approximately determined byusing the dividing the angle between the two points witha vertex into smaller angles The acupoints from GV17 to

GV24 for example are positioned by dividing the anglebetween the GV17 and Yintang with the vertex TE20CPinto smaller subangles with proportional values as shownin Figure 4 In the numerical code for positioning theproportional acupoints proper small intervals are introducedfor every condition in Table 4 If a pixel point on the skinrsquossurface satisfied the condition within the interval that pixelpoint was included and an average point which was taken tobe the position for the proportional acupoint was obtainedTraditionally all the acupoints are classified according to themeridians to which they belong

The results for the morphological acupoints for the sixsubjects are shown in Figure 7 The figures are displayedfor the right sides of the heads and the three lines in eachhead connect the two control points with the morphologicalacupoints along the curves The morphological acupointsGB4 GB5 and GB6 for example reside on the curveconnecting the control points of ST8 and GB7 Our computercode for the morphological technique has been checked tostart positioning themorphological acupoints with themodelof the normal man (the panel in the upper left corner inFigure 7) By adopting the model of the normal man as thestandard source model we confirmed that the source pointsprecisely match the target points in the case of the normalman The same code was applied to the other models toobtain the morphological acupoints Figure 7 shows lines


Table 4 Proportional acupoints The left and the right sides ofthe acupoints are determined by the positions of 119875 whether theacupoints are located on the left and the right side of the 119910119911plane respectively daggerThe point AUX is introduced for arithmeticconvenience

Conditions Calculatedpoints

12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19

with slightly different shapes that depend on the surfacestructures of the heads of the other two men (the obese man

and the underweightman) and those of the three women (theoverweight woman the normal woman and the underweightwoman)

4 Discussion

Recently considerable efforts have been made to understandthe characteristics of acupuncture points [17 18] In our previ-ous study acupuncture point positioning for the entire bodyof a single person was done by using the 3D digital modelof the normal man The reference points or the landmarkswere positioned based on the standard descriptions of theacupoints and the formulae for the proportionalities betweenthe acupoints and the reference points were presented every-where on the body We found that the 37 of the 361standardized acupoints on the entire bodywere automaticallylinked to the reference pointsThe reference points accountedfor 11 of the 361 acupoints and the remaining acupoints(52) were positioned point-by-point by using 3D computergraphics libraries [16] In this study we increased the numberof subjects to six and confined the positioning area to theirheads For the locations of some acupoints we used themorphological technique in which topographical constraintsbetween acupoints were considered among the individualsMany morphological approaches to handling medical imagedata have developed for use in basic research as well as inclinical studies in various fields [19ndash22]

Currently substantial interrater variability in acupointlocation and intrarater variability within the clinical settingexist because the optimal location of the point therapeuticallymay deviate from the location of the standard textbookpoint and may vary from treatment to treatment That is inaddition to the anatomical proportional and morphologicalconsiderations for locating a point a clinical palpatoryconsideration may also exist which may involve modifyingthe textbook location of the point based upon the textureor the tension of the tissue as detected by the practitioneror based upon the ldquoAshirdquo tenderness felt by the patientOur approach can also accommodate the localization of anyadditional points on the body thatmight be clinically relevantfor treatment

The traditional methods for acupoint localization arebased on measurements along the surface of a body withcun the traditional Chinese measure which varies in lengthfor different parts of the body and for different directionsof the longitude (vertical) and latitude (horizontal) With acontrasting approach to this work authors of [15] introducedthree projection planes on the head for a 2D acupointdescription system For a definite viewing direction all visiblepoints on the bodyrsquos surface have one projection plane that isperpendicular to that viewing direction and the visible pointshave corresponding projection points in that projectionplane Each projection point may be back-projected to thebodyrsquos surface to obtain the corresponding 3D coordinatesThis project matching between the 2D and the 3D images isnecessary to satisfy the traditional descriptions on position-ing acupoints


Table 5 Seven acupoints categorized as morphological points with the two corresponding control points The values in the parentheses arethe ldquostandard positionsrdquo of the acupoints on the left and the right sides of the head taken from the 3D digital model of the normal manThesevalues were used as the standard for calculating the morphological acupoints in the other models

Acupoints Control points

GB4 Left (minus0539 0274 minus0912)Right (0560 0274 minus0912) ST8 Left (minus0447 0278 minus1026)

GB5 Left (minus0610 0272 minus0759) Right (0457 0278 minus1026)Right (0628 0272 minus0759) GB7 Left (minus0641 0155 minus0474)

GB6 Left (minus0639 0239 minus0594) Right (0651 0155 minus0474)Right (0655 0239 minus0594)

GB10 Left (minus0584 minus0266 minus0467) GB9 Left (minus0641 minus0089 minus0714)Right (0541 minus0266 minus0467) Right (0613 minus0089 minus0714)

GB11 Left (minus0556 minus0266 minus0220) GB12 Left (minus0529 minus0138 0027)Right (0541 minus0266 minus0220) Right (0529 minus0138 0027)

TE18 Left (minus0555 minus0159 minus0128) TE17 Left (minus0546 minus0004 0027)Right (0555 minus0134 minus0128) Right (0566 minus0004 0027)

TE19 Left (minus0604 minus0159 minus0284) TE20 Left (minus0628 minus0004 minus0459)Right (0585 minus0159 minus0284) Right (0623 minus0004 minus0459)

In conclusion we propose a partially automated methodand procedure for localizing the standardized acupuncturepoints on the heads of 3D digital CT human models takenfrom six living individuals In the future we expect to be ableto develop for practical clinical use a fully automated methodfor positioning acupuncture points on the entire body of anindividual

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors used the Digital Korean data that were pro-duced and distributed by the Catholic Institute for AppliedAnatomy College of Medicine Catholic University of Koreaand the Korean Institute of Science and Technology Informa-tion

References

[1] K H Hohne B Pflesser A Pommert M Riemer R Schubertand U Tiede ldquoA new representation of knowledge concerninghuman anatomy and functionrdquo Nature Medicine vol 1 no 6pp 506ndash511 1995

[2] A Pommert K H Hohne B Pflesser et al ldquoCreating a high-resolution spatialsymbolic model of the inner organs based onthe visible humanrdquoMedical Image Analysis vol 5 no 3 pp 221ndash228 2001

[3] Q Xu R Bauer B M Hendry et al ldquoThe quest for moderni-sation of traditional Chinese medicinerdquo BMC Complementaryand Alternative Medicine vol 13 article 132 2013

[4] H Liu X Shen H Tang J Li T Xiang and W Yu ldquoUsingMicroPET imaging in quantitative verification of the acupunc-ture effect in ischemia stroke treatmentrdquo Scientific Reports vol3 article 1070 2013

[5] H Liu J-Y Xu L Li B-C Shan B-B Nie and J-QXue ldquoFMRI evidence of acupoints specificity in two adjacentacupointsrdquo Evidence-Based Complementary and AlternativeMedicine vol 2013 Article ID 932581 5 pages 2013

[6] V Napadow N Makris J Liu N W Kettner K K Kwongand K K S Hui ldquoEffects of electroacupuncture versus manualacupuncture on the humanbrain asmeasured by fMRIrdquoHumanBrain Mapping vol 24 no 3 pp 193ndash205 2005

[7] National Institutes of Health (NIH) ldquoConsensus developmentpanel on acupuncture acupuncturemdashNIH consensus confer-encerdquoThe Journal of the AmericanMedical Association vol 280no 17 pp 1518ndash1524 1998

[8] V Napadow A Ahn J Longhurst et al ldquoThe status and futureof acupuncturemechanism researchrdquo Journal of Alternative andComplementary Medicine vol 14 no 7 pp 861ndash869 2008

[9] WHO ldquoWorldHealthOrganization (WHO) standard acupunc-ture point locationrdquo in WHO Standard Acupuncture PointLocations in theWestern Pacific Region pp 1ndash14WHOWesternPacific Region Geneva Switzerland 2008

[10] D H W Groenemeyer L Zhang S Schirp and J BaierldquoLocalization of acupuncture points BL25 and BL26 using com-puted tomographyrdquo Journal of Alternative and ComplementaryMedicine vol 15 no 12 pp 1285ndash1291 2009

[11] H-J Park Y Chae M-Y Song et al ldquoA comparison betweendirectional and proportional methods in locating acupuncturepoints using dual-energy X-ray absorptiometry in Koreanwomenrdquo American Journal of Chinese Medicine vol 34 no 5pp 749ndash757 2006

[12] D Zhang X Yan X Zhang et al ldquoSynchrotron radiation phase-contrast X-ray CT imaging of acupuncture pointsrdquo Analyticaland Bioanalytical Chemistry vol 401 no 3 pp 803ndash808 2011

[13] S Koo S Kim Y Kim S Kang and S Choi ldquoMeasuring thelocation of PC8 acupuncture point using X-ray radiography inhealthy adultsrdquo Korean Journal of Oriental Medicine vol 16 pp123ndash126 2010 (Korean)

[14] I-S Lee S-H Lee S-Y Kim H Lee H-J Park and YChae ldquoVisualization of the meridian system based on biomedi-cal information about acupuncture treatmentrdquo Evidence-Based


Complementary and Alternative Medicine vol 2013 Article ID872142 5 pages 2013

[15] L Zheng B Qin T Zhuang U Tiede and K H HohneldquoLocalization of acupoints on a head based on a 3D virtualbodyrdquo Image and Vision Computing vol 23 no 1 pp 1ndash9 2005

[16] J Kim and D-I Kang ldquoPositioning standardized acupuncturepoints on the whole body based on X-ray computed tomogra-phy imagesrdquoMedical Acupuncture vol 26 no 1 pp 40ndash49 2014

[17] J Kim K-H Bae K-S Hong S-C Han and K-S SohldquoMagnetic resonance imaging and acupuncture a feasibilitystudy on the migration of tracers after injection at acupointsof small animalsrdquo Journal of Acupuncture and Meridian Studiesvol 2 no 2 pp 152ndash158 2009

[18] J Kim D-I Kang K-S Soh and S Kim ldquoAnalysis on post-mortem tissues at acupuncture points in the image datasets ofvisible human projectrdquo Journal of Alternative and Complemen-tary Medicine vol 18 no 2 pp 120ndash129 2012

[19] B Golosio A Brunetti and S R Amendolia ldquoA novel mor-phological approach to volume extraction in 3D tomographyrdquoComputer Physics Communications vol 141 no 2 pp 217ndash2242001

[20] T L Weng S J Lin W Y Chang and Y N Sun ldquoVoxel-based texturemapping formedical datardquoComputerizedMedicalImaging and Graphics vol 26 no 6 pp 445ndash452 2002

[21] E Stindel J L Briard P Merloz et al ldquoBone morphing 3Dmorphological data for total knee arthroplastyrdquoComputerAidedSurgery vol 7 no 3 pp 156ndash168 2002

[22] D C Barber and D R Hose ldquoAutomatic segmentation of med-ical images using image registration diagnostic and simulationapplicationsrdquo Journal of Medical Engineering amp Technology vol29 no 2 pp 53ndash63 2005

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom


Table 1 Characteristics of six subjects for the 3D digital CT images

Subject Gender Age (years) Height (cm) Weight (kg) BMI (kgm2) Category1 Male 26 170 66 2284 Normal2 Male 37 175 59 1927 Underweight3 Male 39 170 75 2734 Obese4 Female 30 158 55 2020 Normal5 Female 28 161 48 1852 Underweight6 Female 43 162 64 2401 Overweight

Obese man Underweight manNormal man

Underweight womanNormal womanOverweight woman

Figure 1 Three-dimensional reconstructed surface models of Korean adults with various body types

points on the surface of the human body as well as thelocations of 361 standard acupuncture points This standardestablished by theWHOmay be applied in teaching researchclinical service preparation of publications and academicexchanges involving acupuncture [9]

In more recent studies on acupoint locations researchersbegan to use more advanced equipment such as X-raymachines As a convenient method for locating acupointsthe cun measurement methods have been widely used inthe practice of acupuncture However the traditional cunmeasurement methods have been criticized for their lack ofreliability In one study a comparison of two different locationmethods was done by using dual-energy X-ray absorptiom-etry to measure the soft tissue and the bone mass indepen-dently [10] Another study used computed tomography (CT)to provide a metric description of acupuncture points inthe lumbar region and to give their relation to individualanatomical landmarks and structures [11] Another studyused X-ray radiography to provide experimental evidence tostandardize the location of the acupoint in the hand [12]

Synchrotron radiation phase-contrast X-ray CT was alsoemployed to investigate the three-dimensional (3D) topo-graphic structures of acupuncture points [13] The results ofanother study suggested that biomedical information aboutacupuncture treatment could be visualized in the form of adata-driven 3D acupuncture point system [14]

The accurate localization of acupoints is a key issue inacupuncture research For more precise scientific researchand development on acupuncture therapy having definitionson how to localize acupoints by using a computer-basedpictorial representation of the human body is critical Amethod for localizing the acupoints on the head of a virtualbody by using segmentation and a 3D visualization of theVOXEL-MAN software system was reported [15] Our initialresearch on positioning all the 361 standardized acupuncturepoints was done with the digital data from a healthy Koreanmale with a normal body shape [16] The cross-sectionalimages generated by X-ray CT were used for generating the3D virtual models for the bones and the skinrsquos surface ofthe entire human body The volumetric 3D acupoint model









GB4 GB5 GB6 GB10 GB11 (5)TE18 TE19 (2)




x

yFrontBackRight

Left

Up

Down

(0 0 0)(511 0 0)

(0 511 0)

(0 0 300)

(511 511 300)

z

z998400

y998400

x998400

(CMxCMyCMz)

































Yintang

GV17

TE20

TOP


GV17

GV18

GV19

GV20GV21 GV22

GV23GV24

YintangTE20CP15

1515

15 15 10 05

35

GV25GV26GV27



ST8

GB4

GB5

GB6

GB7

GB9

TE20

GB12 TE17

GB10

GB11

TE19

TE18

14

14

141413

1313

13 13

13

(a)

P

Q

RX


Q998400

P998400



(b)




(1199091015840

1199101015840

1199111015840

) = 120572 (119909 minusCM119909 119910 minusCM

119910 119911 minusCM

119911) (1)


119894119895119896 is CM

119909= sum119894119895119896119868bin119894119895119896119894 CM

119910=

sum119894119895119896119868bin119894119895119896119895 CM


for convenience






= 0(3)




GB17GB16GB15

GB14GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8GB18

BL9

BL10

GB19

GB20

GB7

(a)GB17GB16

GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2

BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8 GB18

BL9

BL10

GB19

GB20

GB7

(b)

Figure 6 Continued


GB17GB16GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4

BL3

BL2

BL1

GV22GV23

GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8

GB18

BL9

BL10

GB19

GB20

GB7

(c)





2

(4)


0made by




(5)




(6)




(7)




ST8

GB4

GB5

GB6GB7

GB9

GB10

GB11

GB12

TE20

TE17

TE19TE18



3 Results








12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19



4 Discussion

















Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
























GB4 GB5 GB6 GB10 GB11 (5)TE18 TE19 (2)




x

yFrontBackRight

Left

Up

Down

(0 0 0)(511 0 0)

(0 511 0)

(0 0 300)

(511 511 300)

z

z998400

y998400

x998400

(CMxCMyCMz)

































Yintang

GV17

TE20

TOP


GV17

GV18

GV19

GV20GV21 GV22

GV23GV24

YintangTE20CP15

1515

15 15 10 05

35

GV25GV26GV27



ST8

GB4

GB5

GB6

GB7

GB9

TE20

GB12 TE17

GB10

GB11

TE19

TE18

14

14

141413

1313

13 13

13

(a)

P

Q

RX


Q998400

P998400



(b)




(1199091015840

1199101015840

1199111015840

) = 120572 (119909 minusCM119909 119910 minusCM

119910 119911 minusCM

119911) (1)


119894119895119896 is CM

119909= sum119894119895119896119868bin119894119895119896119894 CM

119910=

sum119894119895119896119868bin119894119895119896119895 CM


for convenience






= 0(3)




GB17GB16GB15

GB14GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8GB18

BL9

BL10

GB19

GB20

GB7

(a)GB17GB16

GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2

BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8 GB18

BL9

BL10

GB19

GB20

GB7

(b)

Figure 6 Continued


GB17GB16GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4

BL3

BL2

BL1

GV22GV23

GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8

GB18

BL9

BL10

GB19

GB20

GB7

(c)





2

(4)


0made by




(5)




(6)




(7)




ST8

GB4

GB5

GB6GB7

GB9

GB10

GB11

GB12

TE20

TE17

TE19TE18



3 Results








12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19



4 Discussion

















Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of











































Yintang

GV17

TE20

TOP


GV17

GV18

GV19

GV20GV21 GV22

GV23GV24

YintangTE20CP15

1515

15 15 10 05

35

GV25GV26GV27



ST8

GB4

GB5

GB6

GB7

GB9

TE20

GB12 TE17

GB10

GB11

TE19

TE18

14

14

141413

1313

13 13

13

(a)

P

Q

RX


Q998400

P998400



(b)




(1199091015840

1199101015840

1199111015840

) = 120572 (119909 minusCM119909 119910 minusCM

119910 119911 minusCM

119911) (1)


119894119895119896 is CM

119909= sum119894119895119896119868bin119894119895119896119894 CM

119910=

sum119894119895119896119868bin119894119895119896119895 CM


for convenience






= 0(3)




GB17GB16GB15

GB14GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8GB18

BL9

BL10

GB19

GB20

GB7

(a)GB17GB16

GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2

BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8 GB18

BL9

BL10

GB19

GB20

GB7

(b)

Figure 6 Continued


GB17GB16GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4

BL3

BL2

BL1

GV22GV23

GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8

GB18

BL9

BL10

GB19

GB20

GB7

(c)





2

(4)


0made by




(5)




(6)




(7)




ST8

GB4

GB5

GB6GB7

GB9

GB10

GB11

GB12

TE20

TE17

TE19TE18



3 Results








12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19



4 Discussion

















Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















ST8

GB4

GB5

GB6

GB7

GB9

TE20

GB12 TE17

GB10

GB11

TE19

TE18

14

14

141413

1313

13 13

13

(a)

P

Q

RX


Q998400

P998400



(b)




(1199091015840

1199101015840

1199111015840

) = 120572 (119909 minusCM119909 119910 minusCM

119910 119911 minusCM

119911) (1)


119894119895119896 is CM

119909= sum119894119895119896119868bin119894119895119896119894 CM

119910=

sum119894119895119896119868bin119894119895119896119895 CM


for convenience






= 0(3)




GB17GB16GB15

GB14GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8GB18

BL9

BL10

GB19

GB20

GB7

(a)GB17GB16

GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2

BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8 GB18

BL9

BL10

GB19

GB20

GB7

(b)

Figure 6 Continued


GB17GB16GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4

BL3

BL2

BL1

GV22GV23

GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8

GB18

BL9

BL10

GB19

GB20

GB7

(c)





2

(4)


0made by




(5)




(6)




(7)




ST8

GB4

GB5

GB6GB7

GB9

GB10

GB11

GB12

TE20

TE17

TE19TE18



3 Results








12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19



4 Discussion

















Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















GB17GB16GB15

GB14GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8GB18

BL9

BL10

GB19

GB20

GB7

(a)GB17GB16

GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4 BL3

BL2

BL1

GV22GV23GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8 GB18

BL9

BL10

GB19

GB20

GB7

(b)

Figure 6 Continued


GB17GB16GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4

BL3

BL2

BL1

GV22GV23

GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8

GB18

BL9

BL10

GB19

GB20

GB7

(c)





2

(4)


0made by




(5)




(6)




(7)




ST8

GB4

GB5

GB6GB7

GB9

GB10

GB11

GB12

TE20

TE17

TE19TE18



3 Results








12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19



4 Discussion

















Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















GB17GB16GB15

GB14 GB13

ST8

GB4GB5

GB6TE23 TE20

GB1ST1ST2

ST3ST4

ST5

LI19

LI20SI18

BL6BL5BL4

BL3

BL2

BL1

GV22GV23

GV24

GV25GV26GV27

CV24

CV23

TE22TE21SI19GB2

ST6SI17

GB9GB8

GB10

BL7BL8

GB18

BL9

BL10

GB19

GB20

GB7

(c)





2

(4)


0made by




(5)




(6)




(7)




ST8

GB4

GB5

GB6GB7

GB9

GB10

GB11

GB12

TE20

TE17

TE19TE18



3 Results








12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19



4 Discussion

















Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















ST8

GB4

GB5

GB6GB7

GB9

GB10

GB11

GB12

TE20

TE17

TE19TE18



3 Results








12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19



4 Discussion

















Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of



















12057911205790=

15125 119875 = GV18

12057911205790=

30125 119875 = GV19

12057911205790=

45125 119875 = GV20

12057911205790=

60125 119875 = GV21

12057911205790=

75125 119875 = GV22

12057911205790=

85125 119875 = GV23

12057911205790=

90125 119875 = GV24

12057911205790=

110125 119875 = AUXdagger

12057921205790=

225125 119875 = ST8

12057921205790=

15125 119875 = GB13

119875119909= Pupil

119909 119875119911= AUX

119911 119875 = GB1412057921205790=

025125 119875 = BL3

12057921205790=

05125 119875 = BL4

119875119909= Pupil

119909 119875119911= GV24

119911 119875 = GB15

119875119909= Pupil

119909 119875119911= GV23

119911 119875 = GB16

119875119909= Pupil

119909 119875119911=

12lowast GV22

119911+

12lowast GV23

119911 119875 = GB17

119875119909= BL4

119909 119875119911= GV23

119911 119875 = BL5

119875119909= BL4

119909 119875119910=

13lowast GV21

119910+

23lowast GV22

119910 119875 = BL6

119875119909= BL4

119909 119875119910=

13lowast GV20

119910+

23lowast GV21

119910 119875 = BL7

119875119909= BL4

119909 119875119910=

13lowast GV19

119910+

23lowast GV20

119910 119875 = BL8

119875119909= GV17

119909minus

1315lowast (BL4

119909minus GV24

119909) 119875119911= GV17

119911 119875 = BL9

119875119909= GB17

119909 119875119911= BL7

119911 119875 = GB18

119875119909= GB20

119909 119875119911= GV17

119911 119875 = GB19

119875119909=

12lowast (GV16

119909+ GB20

119909) 119875119911= GV15

119911 119875 = BL10

119875 =

12lowast (TE21 + GB2) 119875 = SI19



4 Discussion

















Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of





























Acknowledgments


References






























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
















research article partially automated method for...

Documents