Acta Otolaryngol80: 185-196, 1975
PLANAR RELATIONSHIPS O F THE SEMICIRCULAR CANALS IN MAN
R. H. I. Blanks,' I. S. Curthoys2 and C. H. Markham
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
From the Department of Neurology, University of California School of Medicine, Los Angeles, Calif. 90024, U S A (Received October 25, 1974)
Abstract. Principalcomponent analyses were determined on a series of points measured from the dissected bony labyrinth of ten human skulls, resulting in planar equations for each of the six semicircular canals. Following this, angles were calculated between the ipsilateral canal planes, between opposite synergisticany acting canal planes and between each canal and the Reid stereotaxic planes. Results indicated that pairs of ipsilateral canals were nearly perpendicular, with the exception of the angle formed between the anterior and horizontal canal (mean=l 1I"). Pairs of contralateral synergistic canal planes formed angles of 19" between right and left horizontal canal planes and 23-24' between vertical canal pairs. The horizontal canals formed an angle of 25"with the Reid horizontal plane. Mathematical equations of the semicircular canals were used to predict the optimal head position for rotational and caloric stimulation.
An extensive number of studies have used human subjects to study the stimulus-response characteristics of the semicircular canals. In some of these, angular acceleration was produced by rotation about the major head axes This investigation was supported in part by U.S. Public Health Service Grant NS-06658; by the Health Sciences Computing Facility, University of California, Los Angeles, sponsored by NIH Special Research Resources Grant RR-3; and by a grant from the National Health and Medical Research Council of Australia. ' Max-Planck-Institut fur Himforschung, 6 Frankfurt/M.-Niederrad, West Germany. * Department of Psychology, University of Sydney, Sydney, N.S.W., Australia.
(Melvill Jones et al., 1964; Clark & Stewart, 1968; Oosterveld, 1970; Guedry et al., 1971; see review by 'lark, 1967) producing a combined stimulation of several pairs Of canals. In other studies, however, subjects were a positioned for the purpose Of Single pair Of vertical canals (Van Der ViS, 1958; Decker, 1969). Positioning was based on the assumption that the vertical canals were oriented perpendicular to the horizontal canal and formed 45" angles with the sagittal plane. For the horizontal canals, and using angular acceleration in the earth horizontal plane, descriptions of head position have ranged from a 30" (Van Egmond et al., 1949; Fitzgerald & Hallpike, 1942) to a 20-25" bent forward position (Graybiel et al., 1948). Other investigators have not described their method of positioning the head but have stated that the horizontal canals were aligned in the plane of rotation (F3rown & Crampton, 1%4). Canal Stimulation experiments Using subjects in a variety of head positions raise the questions of which canals are activated by the Stimulus and to what degree. It has been known since the time of Ewald that a single canal will be maximally stimulated when rotated in its Plane and minimally affected when rotated in a plane perpendicular to the canal Acta Otolarvngol80
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
186
R . H . I . Blanks e l al.
plane. It can be concluded that to stimulate individual canals or pairs of canals properly and to control the amount of stimulation to other canal pairs, the position of the canal planes must be known. A complete description of the human semicircular canals in relationship to the stereotaxic reference system and the internal angular relationships between canal planes has not appeared in the literature. Previous studies have been mainly limited to an angular description of a single canal. Spiegel & Sommer (1947) describe the horizontal canal as being elevated 15-30" above the horizontal plane while Girard (1923) gives a value of 25"&7" for the same angle. Other workers, however, have reported smaller angles formed by the horizontal canal with the Frankfurt horizontal plane. For example, Kudo (1965) reported 15737' (sic) and Nishimura (1930) 12.2". A few authors have described the angular relationships of the vertical canals. Colosi & Giannardi (1968) examined the angles formed by the vertical canals and the vertical stereotaxic planes in X-ray material. These angles were also described by Spiegel & Sommer ( 1947) and by Mincker ( 1968). It is clear from this brief review of the literature that more information is required to define more precisely the position of the semicircular canal planes with respect to skull landmarks. This report is concerned with a mathematical description of the human semicircular canal planes with respect to each other and to the stereotaxic coordinate system. PROCEDURES A description of the position of the semicircular canals requires a convenient external frame of reference. For this reason, the Reid stereotaxic coordinate system was employed. This system defines the horizontal plane as one passing through the inferior margin of the orbits (orbitale) and the center points of the two external auditory canals. The Reid Actu Orolaryngol80
stereotaxic frontal and sagittal planes are perpendicular to the horizontal plane and pass through the center of the external auditory canal and mid-sagittal suture respectively. Mathematical conventions proposed by Hixson et al. (1966) to define the stereotaxic system were used throughout this study. For comparison, measurements of the Frankfurt horizontal plane (Ohr-Augen-Ebene) were also taken. This reference system differs from the Reid in that the horizontal plane, and consequently the vertical planes, are defined by the above orbital points and the superior margin of the external auditory canal. Material consisted of ten racially mixed human skulls (Carolina Biological Supply Co., Burlington, N.C.) which were free of gross pathology or bone deformities. Skulls were secured to a Talairach human stereotaxic instrument (Schaltenbrand & Bailey, 1959) by two ear bars placed in the center of the external auditory canal and two eye bars which made contact with the inferior margins of the orbits. These four craniometric points defined the horizontal skull plane with respect to the stereotaxic frame and measuring device. The skull was then secured with four pins (Figs 1 A and B). Micrometer adjustments of each pin permitted accurate realignment of the skull after removal of the eye and ear bars and permitted the removal of the skull for dissection. Exposure of the semicircular canals for measurement required the removal of portions of the mastoid process and middle ear structures. Each canal was exposed laterally about its curvature from the ampullary region to the body of the vestibule. After dissection, each skull was then remounted in the stereotaxic instrument. To insure proper alignment of the skull in relation to the stereotaxic frame and measuring device, both anterior-posterior (A-P) and lateral X-rays were taken of each skull (Figs. 1 A and B). Tube to plate distance was 4.6 meters, producing a magnification of 5%. The X-ray tube was aligned by means of a sighting tube with a front to back target separation of
187
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
Planar relationships of the semicircular canals
Fig. I A. Lateral X-ray of a skull fixed in the stereotaxic frame. The temporal bones have been dissected, altering normal bony landmarks. The measuring device (not shown) is attached to the broad horizontal bar a t the bottom of the picture. The craniometrically defined Reid plane (black line in figure) is defined by black dots placed
on the inferior surface of orbits and at center of external auditory canal. Arrows indicate markers made of metal particles placed at the junction of common crus and vestibule of both labyrinths. In this preparation, springs support the mandible and a latch holds the convexity of the skull. I B. Anterior-posterior view of skull shown in A. Acta Otolaryngol80
R . H . I . Blanks et al.
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
188
Fig. 2 A . A right posterior-1ateral viewI of canal dissection, whiich also illustrates the perspective 01 'Fig. 2 B: and of 3 A, B , C.
Fig. 2 B . Higher magnification view of st:micircular canal dissect.ion.
250 mm. Deviations of less than 0.5" from a true lateral and A-P view (at the level of Reid Horizontal Plane) were produced by this alignment procedure. Acta Otolaryngoi 80
With each skull mounted in the frame, the planes of excursion of an x, y , z micromanipulator (Prior, Farmingdale, N.Y .) were aligned parallel to the stereotaxic planes.
Planar relationships of the semicircular canals
Table I. Mean planar equations f o r each semicircular canal plane defined in the Reid stereotaxic system
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
Insert in figure indicates the axes and polarity of the coordinate system
Semicircular Canal Left horizontal canal.. Left anterior canal.. . . Left posterior canal.. . Right horizontal canal Right anterior canal.. Right posterior canal .
Planar Equation .365x - .158y - , 9 0 5 ~= 0 , 6 5 2 ~- , 7 5 3 ~- .017z = 0 , 7 5 7 ~+ , 5 6 1 ~+ ,3202 = 0 , 3 6 5 ~+ . 158y - ,9052 = 0 , 6 5 2 ~+ .753y - , 0 1 7 ~= 0 , 7 5 7 ~- , 5 6 1 ~+ , 3 2 0 ~= 0
The three dimensional coordinates (to the nearest 0.1 mm) of a series of approximately equally spaced points along each semicircular canal were measured by means of a sharp needle attached to the micromanipulator. Under magnification, from 30 to 103 points were measured on each canal, the difference in number being related to differences in canal diameter (see Fig. 2). Points were taken from the common crus to the ampulla of each vertical canal and from the body of the vestibule to the ampulla in the horizontal canals. Data points were taken from the medial-most portion of the bony vertical canals and from the superior-most portion of the bony horizontal canal. Points within the ampullary regions
189
were not included in subsequent calculations. Translational calculations were performed on the original data to refer each point to the stereotaxic zero. Each point on a canal thus represented a three-dimensional vector in the Reid stereotaxic coordinate system. Data points belonging to each canal were subjected to computer analysis using both a principalcomponent and a multiple regression technique (Blanks et al., 1972). Both methods gave the statistically weighted, best-fitting planar equation passing through all points for each canal and provided virtually identical results. Only those values derived from principalcomponent analysis (Thurstone, 1957) will be reported here. This technique provides the coefficients A, B and C of the generalized equation for a plane Ax+By+Cz+D=O: The D term has been set to zero. Thus, the equations presented are for planes parallel to the nominated canal plane passing through the origin. This has no effect on the computation of angles between planes. Once derived, planar equations for each canal served as input to several computer programs which provided angular interrelationships between the canal planes. RESULTS Craniometric measurements on the skull material revealed a mixture of craniometric shapes. The cephalic index (CI), defined as the ratio of the maximum cranial breadth and the
Table 11. Angles in degrees between pairs of human semicircular canal planes Means, standard deviation (S.D.)and confidence limits for angles between ipsilateral canal planes represent combined right and left values (N=20). Angles between synergistic canal pairs were considered as unique values and were not combined (N=lO) 95% confidence limits
Pairs of canal planes
N
Meant S.D.
Lower
Upper
Horizontal-Anterior Anterior-Posterior Posterior-Horizontal Left anterior-Right posterior Left posterior-Right anterior Left horizontal-Right horizontal
20 20 20 10 10 10
111.76t7.55 86.16L4.72 95.75k4.66 24.56f7.19 23.73 k6.71 19.82 f14.93
108.23 83.95 93.56 19.41 18.93 9.14
115.30 88.38 97.93 29.70 28.53 30.50 Acta Otolaryngol 80
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
190
R . H . I . Blanks et al.
ANGLES I N DEGREES BETWEEN HUMAN CANAL PLANES AND THE REID STEREOTAXIC PLANES Angle between t h e Horizontal Canal Plane & Stereotaxic
Mean -I SD
frontal plane
8, 68.60 2 5.86
Horizontal plane
e2 25.12 2 5.62
Sagittal plane
O3 99.10 2 8.76
ANGLES I N DEGREES BETWEEN HUMAN CANAL PLANES AND THE REID STEREOTAXIC PLANES
95% Confidence Angle between t h e Limits Anterior Canal Mean 2 SD N Lower Upper Plane & Stereotaxic --~ 20 65.85 71.34 0 , 49.32 ? 4.79 20 f r o n t a l plane 20 22.49 27.75 Horizontal plane e2 89.00 4.94 20 20 95.00 103.20 Sagittal plane 0, 41.10 2 4.82 20 N
95% Confidence Limits Lower Upper
_
_
47.08
51.56
86.69
91.32
38.85
43.36
Fig. 3 A-C. Location of the angles (01,02,0,) between the semicircular canal and stereotaxic planes from vector dot-product calculations. Values represent mean from combined right and left sides ( N = 2 0 ) . The right canal
planes are illustrated; location for values on left side are in mirror image locations. Angular values and location for the horizontal canal are shown in A , for anterior canal in B, and for the posterior canal in C.
maximum cranial length, ranged from 70 to 77.5. Seven skulls were of the dolichocephalic or long skull type (7O
PLANAR RELATIONSHIPS O F THE SEMICIRCULAR CANALS IN MAN
R. H. I. Blanks,' I. S. Curthoys2 and C. H. Markham
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
From the Department of Neurology, University of California School of Medicine, Los Angeles, Calif. 90024, U S A (Received October 25, 1974)
Abstract. Principalcomponent analyses were determined on a series of points measured from the dissected bony labyrinth of ten human skulls, resulting in planar equations for each of the six semicircular canals. Following this, angles were calculated between the ipsilateral canal planes, between opposite synergisticany acting canal planes and between each canal and the Reid stereotaxic planes. Results indicated that pairs of ipsilateral canals were nearly perpendicular, with the exception of the angle formed between the anterior and horizontal canal (mean=l 1I"). Pairs of contralateral synergistic canal planes formed angles of 19" between right and left horizontal canal planes and 23-24' between vertical canal pairs. The horizontal canals formed an angle of 25"with the Reid horizontal plane. Mathematical equations of the semicircular canals were used to predict the optimal head position for rotational and caloric stimulation.
An extensive number of studies have used human subjects to study the stimulus-response characteristics of the semicircular canals. In some of these, angular acceleration was produced by rotation about the major head axes This investigation was supported in part by U.S. Public Health Service Grant NS-06658; by the Health Sciences Computing Facility, University of California, Los Angeles, sponsored by NIH Special Research Resources Grant RR-3; and by a grant from the National Health and Medical Research Council of Australia. ' Max-Planck-Institut fur Himforschung, 6 Frankfurt/M.-Niederrad, West Germany. * Department of Psychology, University of Sydney, Sydney, N.S.W., Australia.
(Melvill Jones et al., 1964; Clark & Stewart, 1968; Oosterveld, 1970; Guedry et al., 1971; see review by 'lark, 1967) producing a combined stimulation of several pairs Of canals. In other studies, however, subjects were a positioned for the purpose Of Single pair Of vertical canals (Van Der ViS, 1958; Decker, 1969). Positioning was based on the assumption that the vertical canals were oriented perpendicular to the horizontal canal and formed 45" angles with the sagittal plane. For the horizontal canals, and using angular acceleration in the earth horizontal plane, descriptions of head position have ranged from a 30" (Van Egmond et al., 1949; Fitzgerald & Hallpike, 1942) to a 20-25" bent forward position (Graybiel et al., 1948). Other investigators have not described their method of positioning the head but have stated that the horizontal canals were aligned in the plane of rotation (F3rown & Crampton, 1%4). Canal Stimulation experiments Using subjects in a variety of head positions raise the questions of which canals are activated by the Stimulus and to what degree. It has been known since the time of Ewald that a single canal will be maximally stimulated when rotated in its Plane and minimally affected when rotated in a plane perpendicular to the canal Acta Otolarvngol80
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
186
R . H . I . Blanks e l al.
plane. It can be concluded that to stimulate individual canals or pairs of canals properly and to control the amount of stimulation to other canal pairs, the position of the canal planes must be known. A complete description of the human semicircular canals in relationship to the stereotaxic reference system and the internal angular relationships between canal planes has not appeared in the literature. Previous studies have been mainly limited to an angular description of a single canal. Spiegel & Sommer (1947) describe the horizontal canal as being elevated 15-30" above the horizontal plane while Girard (1923) gives a value of 25"&7" for the same angle. Other workers, however, have reported smaller angles formed by the horizontal canal with the Frankfurt horizontal plane. For example, Kudo (1965) reported 15737' (sic) and Nishimura (1930) 12.2". A few authors have described the angular relationships of the vertical canals. Colosi & Giannardi (1968) examined the angles formed by the vertical canals and the vertical stereotaxic planes in X-ray material. These angles were also described by Spiegel & Sommer ( 1947) and by Mincker ( 1968). It is clear from this brief review of the literature that more information is required to define more precisely the position of the semicircular canal planes with respect to skull landmarks. This report is concerned with a mathematical description of the human semicircular canal planes with respect to each other and to the stereotaxic coordinate system. PROCEDURES A description of the position of the semicircular canals requires a convenient external frame of reference. For this reason, the Reid stereotaxic coordinate system was employed. This system defines the horizontal plane as one passing through the inferior margin of the orbits (orbitale) and the center points of the two external auditory canals. The Reid Actu Orolaryngol80
stereotaxic frontal and sagittal planes are perpendicular to the horizontal plane and pass through the center of the external auditory canal and mid-sagittal suture respectively. Mathematical conventions proposed by Hixson et al. (1966) to define the stereotaxic system were used throughout this study. For comparison, measurements of the Frankfurt horizontal plane (Ohr-Augen-Ebene) were also taken. This reference system differs from the Reid in that the horizontal plane, and consequently the vertical planes, are defined by the above orbital points and the superior margin of the external auditory canal. Material consisted of ten racially mixed human skulls (Carolina Biological Supply Co., Burlington, N.C.) which were free of gross pathology or bone deformities. Skulls were secured to a Talairach human stereotaxic instrument (Schaltenbrand & Bailey, 1959) by two ear bars placed in the center of the external auditory canal and two eye bars which made contact with the inferior margins of the orbits. These four craniometric points defined the horizontal skull plane with respect to the stereotaxic frame and measuring device. The skull was then secured with four pins (Figs 1 A and B). Micrometer adjustments of each pin permitted accurate realignment of the skull after removal of the eye and ear bars and permitted the removal of the skull for dissection. Exposure of the semicircular canals for measurement required the removal of portions of the mastoid process and middle ear structures. Each canal was exposed laterally about its curvature from the ampullary region to the body of the vestibule. After dissection, each skull was then remounted in the stereotaxic instrument. To insure proper alignment of the skull in relation to the stereotaxic frame and measuring device, both anterior-posterior (A-P) and lateral X-rays were taken of each skull (Figs. 1 A and B). Tube to plate distance was 4.6 meters, producing a magnification of 5%. The X-ray tube was aligned by means of a sighting tube with a front to back target separation of
187
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
Planar relationships of the semicircular canals
Fig. I A. Lateral X-ray of a skull fixed in the stereotaxic frame. The temporal bones have been dissected, altering normal bony landmarks. The measuring device (not shown) is attached to the broad horizontal bar a t the bottom of the picture. The craniometrically defined Reid plane (black line in figure) is defined by black dots placed
on the inferior surface of orbits and at center of external auditory canal. Arrows indicate markers made of metal particles placed at the junction of common crus and vestibule of both labyrinths. In this preparation, springs support the mandible and a latch holds the convexity of the skull. I B. Anterior-posterior view of skull shown in A. Acta Otolaryngol80
R . H . I . Blanks et al.
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
188
Fig. 2 A . A right posterior-1ateral viewI of canal dissection, whiich also illustrates the perspective 01 'Fig. 2 B: and of 3 A, B , C.
Fig. 2 B . Higher magnification view of st:micircular canal dissect.ion.
250 mm. Deviations of less than 0.5" from a true lateral and A-P view (at the level of Reid Horizontal Plane) were produced by this alignment procedure. Acta Otolaryngoi 80
With each skull mounted in the frame, the planes of excursion of an x, y , z micromanipulator (Prior, Farmingdale, N.Y .) were aligned parallel to the stereotaxic planes.
Planar relationships of the semicircular canals
Table I. Mean planar equations f o r each semicircular canal plane defined in the Reid stereotaxic system
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
Insert in figure indicates the axes and polarity of the coordinate system
Semicircular Canal Left horizontal canal.. Left anterior canal.. . . Left posterior canal.. . Right horizontal canal Right anterior canal.. Right posterior canal .
Planar Equation .365x - .158y - , 9 0 5 ~= 0 , 6 5 2 ~- , 7 5 3 ~- .017z = 0 , 7 5 7 ~+ , 5 6 1 ~+ ,3202 = 0 , 3 6 5 ~+ . 158y - ,9052 = 0 , 6 5 2 ~+ .753y - , 0 1 7 ~= 0 , 7 5 7 ~- , 5 6 1 ~+ , 3 2 0 ~= 0
The three dimensional coordinates (to the nearest 0.1 mm) of a series of approximately equally spaced points along each semicircular canal were measured by means of a sharp needle attached to the micromanipulator. Under magnification, from 30 to 103 points were measured on each canal, the difference in number being related to differences in canal diameter (see Fig. 2). Points were taken from the common crus to the ampulla of each vertical canal and from the body of the vestibule to the ampulla in the horizontal canals. Data points were taken from the medial-most portion of the bony vertical canals and from the superior-most portion of the bony horizontal canal. Points within the ampullary regions
189
were not included in subsequent calculations. Translational calculations were performed on the original data to refer each point to the stereotaxic zero. Each point on a canal thus represented a three-dimensional vector in the Reid stereotaxic coordinate system. Data points belonging to each canal were subjected to computer analysis using both a principalcomponent and a multiple regression technique (Blanks et al., 1972). Both methods gave the statistically weighted, best-fitting planar equation passing through all points for each canal and provided virtually identical results. Only those values derived from principalcomponent analysis (Thurstone, 1957) will be reported here. This technique provides the coefficients A, B and C of the generalized equation for a plane Ax+By+Cz+D=O: The D term has been set to zero. Thus, the equations presented are for planes parallel to the nominated canal plane passing through the origin. This has no effect on the computation of angles between planes. Once derived, planar equations for each canal served as input to several computer programs which provided angular interrelationships between the canal planes. RESULTS Craniometric measurements on the skull material revealed a mixture of craniometric shapes. The cephalic index (CI), defined as the ratio of the maximum cranial breadth and the
Table 11. Angles in degrees between pairs of human semicircular canal planes Means, standard deviation (S.D.)and confidence limits for angles between ipsilateral canal planes represent combined right and left values (N=20). Angles between synergistic canal pairs were considered as unique values and were not combined (N=lO) 95% confidence limits
Pairs of canal planes
N
Meant S.D.
Lower
Upper
Horizontal-Anterior Anterior-Posterior Posterior-Horizontal Left anterior-Right posterior Left posterior-Right anterior Left horizontal-Right horizontal
20 20 20 10 10 10
111.76t7.55 86.16L4.72 95.75k4.66 24.56f7.19 23.73 k6.71 19.82 f14.93
108.23 83.95 93.56 19.41 18.93 9.14
115.30 88.38 97.93 29.70 28.53 30.50 Acta Otolaryngol 80
Acta Otolaryngol Downloaded from informahealthcare.com by University of Otago on 01/03/15 For personal use only.
190
R . H . I . Blanks et al.
ANGLES I N DEGREES BETWEEN HUMAN CANAL PLANES AND THE REID STEREOTAXIC PLANES Angle between t h e Horizontal Canal Plane & Stereotaxic
Mean -I SD
frontal plane
8, 68.60 2 5.86
Horizontal plane
e2 25.12 2 5.62
Sagittal plane
O3 99.10 2 8.76
ANGLES I N DEGREES BETWEEN HUMAN CANAL PLANES AND THE REID STEREOTAXIC PLANES
95% Confidence Angle between t h e Limits Anterior Canal Mean 2 SD N Lower Upper Plane & Stereotaxic --~ 20 65.85 71.34 0 , 49.32 ? 4.79 20 f r o n t a l plane 20 22.49 27.75 Horizontal plane e2 89.00 4.94 20 20 95.00 103.20 Sagittal plane 0, 41.10 2 4.82 20 N
95% Confidence Limits Lower Upper
_
_
47.08
51.56
86.69
91.32
38.85
43.36
Fig. 3 A-C. Location of the angles (01,02,0,) between the semicircular canal and stereotaxic planes from vector dot-product calculations. Values represent mean from combined right and left sides ( N = 2 0 ) . The right canal
planes are illustrated; location for values on left side are in mirror image locations. Angular values and location for the horizontal canal are shown in A , for anterior canal in B, and for the posterior canal in C.
maximum cranial length, ranged from 70 to 77.5. Seven skulls were of the dolichocephalic or long skull type (7O
