Original Paper

Head Accelerations during Particle Repositioning Manoeuvres
M.E. Faldon, A.M. Bronstein

Department of Clinical Neuroscience, Division of Neuroscience and Mental Health, Imperial College, London, UK

Address of Corresponding Author

Audiol Neurotol 2008;13:345-356 (DOI: 10.1159/000136153)

 Outline

• Key Words
• Abstract
• Introduction
• Materials and Methods
• Accelerometers
• Angular Velocity Measurement
• Canal Geometry
• cb-Plots
• Results
• Diagnostic Manoeuvres
• Therapeutic Manoeuvres
• Discussion
• Appendix: Model
• References

• Author Contacts
• Article Information
• Publication Details
• Drug Dosage / Copyright

  Key Words

• Benign paroxysmal positional vertigo
• Particle repositioning manoeuvres
• Hallpike manoeuvre
• Epley manoeuvre
• Semont manoeuvre

  Abstract

Benign paroxysmal positional vertigo (BPPV) due to canalithiasis can be treated with particle repositioning manoeuvres, which aim to evacuate trapped particles from the semicircular canals (SCC). The movement of particles within the SCC is affected by gravity as well as by the accelerations of the head during the manoeuvres. Moreover, as experienced by the particles, gravity is indistinguishable from an upward acceleration of the SCC in free space. We used a set of three orthogonal linear accelerometers to measure the net three-dimensional linear acceleration vector acting on the head during the Hallpike manoeuvre and Epley and Semont particle repositioning manoeuvres (which are used to treat posterior canal BPPV). The projection of the net acceleration vector onto the SCC planes showed that both the Epley and Semont manoeuvres approximated to stepwise, 360°, backward rotations in the plane of the targeted posterior canal. Angular velocity measurements however showed that the rotational component during the central stages of these two manoeuvres is opposite in direction. A simple model of head rotations during particle repositioning manoeuvres was created which showed good agreement to the linear acceleration measurements. Analysis of modelled and measured data identified that speed of movement during the Semont manoeuvre should be critical to its clinical success.

Copyright © 2008 S. Karger AG, Basel

 Introduction

Benign paroxysmal positional vertigo (BPPV) is a common disorder that is usually associated with canalithiasis [Hall et al., 1979], when dense particles become trapped in the semicircular canals (SCC) of the vestibular system. If the head is tilted, these particles can fall under gravity, causing abnormal endolymph flow and symptoms of vertigo.

BPPV can often be effectively treated with particle repositioning manoeuvres [Solomon, 2000]. One aim of a particle repositioning manoeuvre is to attain a series of head positions, where in each stage the particles are allowed to fall under gravity and progress around the canal - from their initial location trapped near the cupula, to the exit from the canal into the utricle. Under the principle of equivalence [Kleppner and Kolenkow, 1978], the gravitational field experienced by the particles is indistinguishable from an upward acceleration of the SCC in free space, at a rate g (the acceleration due to gravity = 9.81 m/s²). Particle repositioning manoeuvres also involve head and body rotations. These rotations produce linear accelerations of the SCC which can also affect particle movement. It is the net acceleration of the SCC, due to gravity and rotations, which will determine particle progression around the canal. Accelerations orthogonal to the canal plane will have no effect on the particles, whereas accelerations in the plane of the canal will produce movement of the particles in the opposite direction.

The aim of this work was to measure the accelerations of the SCC during diagnostic and therapeutic manoeuvres for BPPV. In particular, the accelerations acting on the posterior SCC during Epley [Epley, 1992] and Semont [Semont et al., 1988] manoeuvres were investigated. These two manoeuvres are ostensibly quite different: the middle stages of the Epley manoeuvre involve yaw movements of the head on the neck, whereas the middle stage of the Semont manoeuvre comprises a whole body swing in the opposite direction to the initial Hallpike movement. Nevertheless the two manoeuvres have the same objective - to evacuate free-floating particles that have become trapped in the posterior SCC. We wanted to find out how these diverse manoeuvres could have similar effects, and try to identify the critical factors involved.

 Materials and Methods

 Accelerometers

Linear accelerometers have a sensitive axis; they give positive signals for forward accelerations along this axis and negative signals for backward accelerations along the axis. When a linear accelerometer is stationary and aligned so that the sensitive axis points upwards, it gives a positive signal equivalent to an acceleration of 1 g. Linear accelerometers do not distinguish between the acceleration due to gravity and accelerations due to movement. However, this distinction is irrelevant, as it is the net equivalent acceleration of the head that will affect the free-floating particles.

A set of three orthogonal linear accelerometers (Entran Ltd., Watford, UK; EGCS series, ±5 g range, frequency response 0-150 Hz) was used to measure the three-dimensional linear acceleration vector acting on the head during the Hallpike [Dix and Hallpike, 1952] manoeuvre and Epley and Semont particle repositioning manoeuvres. The accelerometers were mounted close together and fixed to the top of the subject's head, using a secure helmet. The subject's head was positioned with the Reid plane horizontal (i.e. a line from the inferior orbital margin to the external auditory meatus was positioned horizontally). Then one accelerometer was aligned with its sensitive direction pointing straight ahead, with positive signals for forward accelerations. A second accelerometer was aligned along the interaural axis, with positive signals for accelerations to the subject's left. The third accelerometer was aligned vertically, with positive signals for upward movement. All of the accelerometer signals were normalized to gravity and then combined to give the net linear acceleration vector (a), in head-fixed coordinates. The acceleration measured at the top of the head was assumed to be a good approximation to the acceleration experienced by the SCC.

 Angular Velocity Measurement

During our experiments, a set of three orthogonal angular rate sensors (Watson Industries, Southampton, UK) was also fixed to the head. These sensors measured the angular velocities of the head about axes coincident with the sensitive directions of the linear accelerometers. The calibrated rate sensor signals were combined to find the angular velocity vector for the head, which could then be projected onto individual canal planes.

 Canal Geometry

In this study, we have assumed a simplified geometry for the SCC. Each canal is assumed to lie in a plane, and the orientations of the canal planes with respect to the head are estimated using the measured values of Blanks et al. [1975]. The vector n defines the direction orthogonal to the canal plane, and the rotation direction for excitation of the canal, according to a right-hand screw. For each canal, we have also estimated the direction of a vector from the centre of the canal to the cupula position (c). This vector defines the 'closed' end of the canal. A third vector is required to define the orientation of the canal unambiguously. The vector b is defined by the cross product of the other two vectors (b = n × c); it is orthogonal to both n and c, and lies in the canal plane (fig. 1).

Fig. 1. SCC geometry. The left-hand side of the figure shows a side view of the right posterior (RP) and right anterior (RA) canals (with the head in the Reid horizontal position) and a top view of the right horizontal (RH) canal. The viewpoint is indicated on the corresponding plan of the right SCC set. n represents the 'normal' vector, orthogonal to the canal plane; represents an arrow coming out of the page, and represents an arrow going into the page. Following the 'right-hand' rule, when the thumb of the right hand is aligned with the direction of n, the fingers curl in the direction of rotational excitation of the canal. c represents the estimated cupula position. b is the third vector required to define the canal orientation (b = n × c). The curved arrow within each SCC indicates the evacuation direction for trapped particles.

Figure 1 illustrates the SCC geometry. The left-hand side of figure 1 shows a side view of the right posterior and right anterior canals, with the head in the Reid horizontal position, and a top view of the right horizontal canal. The viewpoint is denoted by the straight thick arrow on the corresponding plan of the right SCC set. The curved arrow within each SCC indicates the evacuation direction for trapped particles. With the head upright, any free-floating particles in the posterior canal would settle at the lowest point of the canal. In order to evacuate the particles, they need to be moved around the canal, towards the common crus and the opening into the utricle. In the head-upright position, particles trapped in the anterior canal sit at the front of the canal, on the cupula. These particles also need to be moved around the canal towards the common crus, in the direction indicated by the curved arrow.

Since the acceleration vector (a) is measured in head-fixed coordinates, and the canal vectors c and b are also defined with respect to the head, the component of acceleration along each canal vector can be calculated by a simple vector dot product (the projection of the acceleration vector onto the appropriate canal vector).

a_c = a · c

a_b = a · b

Then, a plot of a_b versus a_c represents the time course of the acceleration in the canal plane. This will be called the 'cb-plot'.

 cb-Plots

Figure 2 shows how the acceleration due to gravity could act on particles in the right posterior canal, during a slow backward rotation in the plane of the canal. In the initial position (fig. 2A), the particles sit at the bottom of the canal, having fallen there under the action of gravity, which is equivalent to an acceleration of the SCC in the upward direction (g). The lower part of figure 2 shows the orientation of the acceleration vector g relative to the canal vectors c and b (cb-plot). The circle, plotted on the cb-axes, represents an acceleration amplitude of 1 g (9.81 m/s²). If the subject's head is rotated slowly backwards through 90° (fig. 2B), the particles fall away from the cupula, in the opposite direction to the gravity vector. In the lower part of figure 2B, the cross represents the acceleration due to gravity in the 90° head-back position, the thick line around the unit circle represents the progression of the gravity vector during the manoeuvre from the initial head position. A further backward rotation of 90° produces a head-hanging position (fig. 2C), and allows the particles to fall further around the canal, towards the exit into the utricle. A final 90° rotation (fig. 2D) should be enough to evacuate the particles. The acceleration in the plane of the right posterior canal during this manoeuvre is represented on the cb-plot by a clockwise rotation around the unit circle. This plot only characterizes the acceleration during a very slow head rotation, since any significant head accelerations will combine with gravity, to shift the net acceleration vector away from the unit circle.

Fig. 2. Backward head rotation, through 90° steps, in the plane of the right posterior canal. The middle row represents the orientation of the right posterior canal with respect to gravity (g), and shows the idealized motion of trapped, free-floating particles during each step. The bottom row plots the endpoint of the gravity vector (×) with respect to the c and b axes for the canal. The circle represents an acceleration amplitude of 1 g. The thick line represents the progression of the gravity vector during the manoeuvres. Please note that these plots only characterize the acceleration during a very slow head rotation, since any significant head accelerations will combine with gravity to shift the net acceleration vector away from the unit circle.

In a similar way, free-floating particles in the right anterior canal could be evacuated via a forward rotation in this canal plane (fig. 3). For the anterior canal, the initial forward rotation (from the Reid horizontal position; fig. 3A) does not contribute to the evacuation of the particles, since they remain close to the cupula position. In the 90° face-down position (fig. 3B) the particles can start to fall away from the cupula. Continued forward rotation allows the particles to move around the canal, falling under the action of gravity. The progression of the gravity vector (relative to the anterior canal vectors c and b) during the forward rotation, from face down through to head upright, is shown in the lower part of figure 3. The acceleration in the plane of the canal, during these simple rotation evacuation manoeuvres, is identical for the posterior and anterior SCC (bottom row of fig. 2 and 3, respectively).

Fig. 3. Forward head rotation, through 90° steps, in the plane of the right anterior canal. The middle row represents the orientation of the right anterior canal with respect to gravity (g), and shows the idealized motion of trapped, free-floating particles during each step. Bottom row plots as for figure 2.

A simple rotation evacuation manoeuvre for particles in the right horizontal canal is illustrated in figure 4. The initial position is right ear down (fig. 4A), with the plane of the right horizontal canal aligned with the gravitational vertical. The head is then rotated (in the plane of the horizontal canal) to the subject's left, through 90° steps, to the face-down position. During this rotation, the particles can move around the canal, under the action of gravity, to the exit into the utricle. The progression of the gravity vector, relative to the horizontal canal vectors c and b, is illustrated in the lower part of figure 4. The acceleration in the plane of the right horizontal canal during this evacuation manoeuvre is represented on the cb-plot by a counter-clockwise rotation around the unit circle.

Fig. 4. A simple evacuation manoeuvre for free-floating particles in the right horizontal canal. The initial position is right ear down (A), the head is then rotated to the subject's left, in 90° steps (B-D). The middle row represents the orientation of the right horizontal canal with respect to gravity (g), and shows the idealized motion of trapped, free-floating particles during each step. Bottom row plots as for figure 2.

The head rotations during particle repositioning manoeuvres were modelled (see Appendix) to predict the accelerations in the canal planes, for comparison with the measured data.

 Results

 Diagnostic Manoeuvres

Straight Back (Sagittal) Manoeuvre
Although the Hallpike manoeuvre is the most commonly performed diagnostic manoeuvre for BPPV, the results for the straight back manoeuvre are simpler to interpret. The acceleration measured during a straight back positional manoeuvre (in the sagittal plane) is illustrated in figure 5. The accelerations measured for the right-sided canals are similar to those for the left-sided canals, due to the symmetrical stimulation during this manoeuvre. The stars indicate the initial linear acceleration, in the head-upright position. The progression of the acceleration vector for the posterior canals (left posterior and right posterior) does not follow the unit circle closely because the head rotation is not in the plane of the posterior canals. The acceleration follows the unit circle more closely for the anterior canal plots (left anterior and right anterior) since the anterior canals lie closer to the sagittal plane.

Fig. 5. Measured acceleration in the plane of each SCC, during a straight back positional manoeuvre. The data are plotted on axes representing the canal vectors c and b for each SCC (as shown in fig. 2-4). The circles represent a linear acceleration of 1 g. The stars indicate the initial linear acceleration, in the head-upright position. LP = Left posterior; LA = left anterior; LH = left horizontal; RP = right posterior; RA = right anterior; RH = right horizontal.

Hallpike Manoeuvre
Figure 6 shows the acceleration measured during a Hallpike manoeuvre (straight back) with the head turned 45° to the subject's right. With this head position, the rotation is roughly in the plane of the right posterior and left anterior canals. The plots show that the acceleration tracks the unit circle for these canals.

Fig. 6. Measured acceleration in the plane of each SCC, during a Hallpike manoeuvre with the head turned 45° to the right. LP = Left posterior; LA = left anterior; LH = left horizontal; RP = right posterior; RA = right anterior; RH = right horizontal.

 Therapeutic Manoeuvres

Epley Manoeuvre
Figure 7A shows the acceleration measured during a complete Epley manoeuvre, as performed in our clinics [Lempert et al., 1995]. The plot shows the acceleration in the plane of the right posterior canal, which is the canal being treated. The initial part consists of a Hallpike manoeuvre, with the head turned 45° to the subject's right. At the end of the Hallpike manoeuvre, the subject is in a supine position, with the head hanging and facing to the right. After a delay, the head is quickly turned 90° to the subject's left. With the head held in this position, the subject turns to lie on the left side. The head is then turned a further 90° to the left, so that the subject is looking towards the floor (at an angle of 45° to vertical). After a further delay, the subject sits up to complete the manoeuvre. The plot of a_b versus a_c, during the Epley manoeuvre, approximates to a clockwise rotation around the unit circle, and is very similar to the plot for the simple, 360° backward rotation, evacuation manoeuvre in the plane of the posterior canal, described in figure 2. The Hallpike manoeuvre and sitting-up parts of the Epley manoeuvre are equivalent to the initial and final stages of the simple (fig. 2) head rotation, respectively. The effects of the 90° head turns are to position the right posterior canal so that the free particles can continue to fall around the canal towards the utricle.

Fig. 7.A Measured acceleration during an Epley manoeuvre for the treatment of the right posterior canal. The data are plotted on axes representing the canal vectors c and b for the right posterior canal. B Measured angular velocity in the plane of the right posterior canal during the different stages of an Epley manoeuvre. a = 45° head turn to the subject's right; b = Hallpike manoeuvre; c = first 90° head turn; d = second 90° head turn; e = sit up, and f = head recentred.

Figure 7B shows the measured angular velocity, in the plane of the right posterior canal, during an Epley manoeuvre. It can be seen that during the four main stages of the manoeuvre (the Hallpike, two 90° head turns and sitting up) the angular velocity in the plane of the target canal is always in the same direction.

Semont Manoeuvre
The Semont manoeuvre is also initiated by a Hallpike movement, with the head turned to position the problematic posterior canal in the plane of the movement. However, in this case, the Hallpike manoeuvre is performed by swinging the subject down to lie on the side. So, to treat a right posterior canal, the subject would end up lying on the right side, facing 45° up from the horizontal. From this position, the subject is quickly swung through 180°, to lie on the opposite side, with the face looking 45° down, toward the floor. After a delay, the subject sits up to complete the manoeuvre. The accelerations measured during three Semont manoeuvres, with slow, medium speed and fast 180° swings, are illustrated in figure 8A-C. During the slow 180° swing (duration 3.6 s; fig. 8A), the acceleration traces a counter-clockwise path around the unit circle, reversing the effects of the Hallpike manoeuvre and continuing round until the head is in the face-down position. In this case, any particles that move away from the cupula during the Hallpike manoeuvre return to the cupula during the slow swing and remain trapped. When the Semont manoeuvre is performed with a fast 180° swing (e.g. duration 1.3 s; fig. 8C), the acceleration plot looks more like a complete clockwise rotation of the unit circle, corresponding to the idealized evacuation procedure described in figure 2. So, the speed of the 180° swing is critical in determining the success of the Semont manoeuvre.

Fig. 8.A-C Measured acceleration during three Semont manoeuvres for the treatment of the right posterior canal. The data are plotted on axes representing the canal vectors c and b for the right posterior canal. The duration of the 180° swing was: 3.6 s (A); 1.5 s (B); 1.3 s (C). D Measured angular velocity in the plane of the right posterior canal during the different stages of slow (solid line) and fast (dotted line) Semont manoeuvres. a = 45° head turn to the subject's left; b = Hallpike manoeuvre; c = 180° whole-body swing; d = sit up, and e = head recentred.

The angular velocities in the plane of the right posterior canal, measured during slow and fast Semont manoeuvres, are plotted in figure 8D. During the Semont manoeuvres plotted in figure 8, the angular velocities in the plane of the right posterior canal, at the centre of the swing, were measured as 60°/s, 190°/s and 250°/s, for the slow, medium and fast swings, respectively. These velocities correspond to approximate centripetal accelerations of 0.1, 0.8 and 1.4 g, respectively. Therefore, during the fast 180° swing, the net linear acceleration at the peak of the swing acts in the opposite direction to gravity. Note that the angular velocity during the central stage of the Semont manoeuvre (fig. 8D, c) is in the opposite direction to the angular velocities measured during the middle stages of the Epley manoeuvre (fig. 7B, c and d).

At the beginning and end of the fast 180° swing (fig. 8C) there are regions where the net acceleration acts in approximately the same direction as gravity, but the magnitude of the acceleration is larger than 1 g. At the beginning of the swing, there is a tangential acceleration of the head upward, which adds to the gravitational acceleration. At the end of the swing, there is a deceleration of the tangential head velocity as the subject reaches the face-down position. This is equivalent to an upward acceleration, again enhancing the gravitational acceleration.

 Discussion

In this paper, we both measured and modelled (see Appendix) the linear accelerations acting on the skull during Hallpike, Epley and Semont manoeuvres. A visual comparison of the measurements and modelled data indicates a good agreement between the two [e.g. fig. 7A and 9 (Epley manoeuvre); fig. 8 and 10 (Semont manoeuvre)].

Fig. 9. Model prediction. cb-Plot in the plane of the right posterior canal, during an Epley manoeuvre. The black dots represent 25-Hz data points; the large white dots represent the endpoints of each step. The star indicates the initial position.

Fig. 10. Model prediction. cb-Plots in the plane of the right posterior canal, during three Semont manoeuvres with 180° swings of different speeds. A Slow swing: duration 3.6 s. B Medium swing: duration 1.5 s. C Fast swing: duration 1.3 s.

Both modelled and measured data indicate that when the net acceleration acting in the canal planes is considered, effective particle repositioning manoeuvres approximate to stepwise, 360° rotations in the target canal plane. The rotation direction allows the trapped particles to fall away from the closed end of the canal towards the exit into the utricle. This perspective reveals the true similarities of the Epley and Semont manoeuvres, despite the different head and body movements involved. The finding that Epley and Semont manoeuvres are analogous to backward rotations in the target canal plane is backed up by the fact that successful BPPV treatments based on actual 360° whole-body rotations have been reported [Lempert et al., 1997; Furman et al., 1998; Nakayama and Epley, 2005]. These studies used machines to perform a controlled 360° rotation of the problematic SCC in the earth-vertical plane.

Examination of the separate steps during the repositioning manoeuvres revealed potentially important practical applications. The Semont manoeuvre begins with a 90° backward rotation in the plane of the posterior canal, which is in the appropriate direction for particles to fall away from their trapped position near the cupula. The next step in the Semont manoeuvre is a 180° whole-body swing in the opposite direction. If this 180° swing is performed slowly, then the particles will just reverse direction and return to the trapped position. During the whole-body swing, there is a centripetal acceleration of the SCC, towards the centre of rotation. The magnitude of this centripetal acceleration depends on the head velocity. At the centre of the swing, where the head velocity is greatest, the centripetal acceleration acts downwards, in the opposite direction to the acceleration due to gravity. If the centripetal acceleration is larger than the acceleration due to gravity, then the net acceleration will be in a downward direction - opposite to the gravitational acceleration. So, if the 180° swing is performed quickly enough, then the effect on the posterior canal can be equivalent to a slow backward rotation. The success of the Semont manoeuvre therapy for BPPV is known to depend critically on the speed of the 180° whole-body swing [Radtke et al., 2004]. Analyses of the accelerations acting on the SCC during the Semont manoeuvre suggest that the optimum duration of the 180° swing is less than 1.5 s.

The angular velocity in the plane of the target canal during particle repositioning manoeuvres may also affect particle movement [Furuya et al., 2002]. During a Hallpike manoeuvre in the plane of the right posterior canal, the angular velocity of the canal induces endolymph flow in the same direction as the desired particle movement. This endolymph flow works in synergy with gravity to move the particles around the canal. Figure 7B shows that the angular velocity in the plane of the right posterior canal during a therapeutic Epley manoeuvre generally supports the particle progression induced by the linear acceleration changes. In particular, the angular velocity during the 90° head turns and the sitting-up phase is in the same direction as during the initial Hallpike manoeuvre.

The angular velocity in the plane of the target canal during the 180° swing of the Semont manoeuvre (fig. 8D) is in the wrong direction, and would induce endolymph flow in the opposite direction to the desired particle movement. Again, the speed of the 180° swing is crucial. If this whole-body swing is fast enough, the linear acceleration effects will dominate, as endolymph viscosity [Rajguru et al., 2004] will preclude significant fluid flow.

Measurements and modelling confirm that the speed of rotation is critical for a successful Semont manoeuvre. Similarly, adequate neck extension and flexibility is required for a successful Epley treatment [Viirre et al., 2005]. Clinicians treating BPPV patients should therefore be familiar with both Epley and Semont manoeuvres; a young obese patient with good neck mobility might be better treated with an Epley procedure, whereas a lightweight elderly patient with poor neck mobility would be better treated with a Semont manoeuvre.

In practical applications, a small 3-axis linear accelerometer, mounted on the subject's head, could be used for online evaluation of particle repositioning manoeuvre procedures. Also, considering the critical importance of achieving a fast 180° whole-body swing, during the Semont manoeuvre, it is recommended that the clinician performing this treatment has an assistant behind the patient, who can help control the subject's movement. Further application of the model described herewith could help in devising manoeuvres for rarer forms of BPPV, chiefly anterior canal BPPV, since it is unlikely that any centre would be able to gather sufficient numbers of patients to run a comparative clinical trial.

Afferent nerve responses during gravity-induced intracanalicular particle movement in a live animal model have been reported [Rajguru and Rabbitt, 2007]. These afferent responses were consistent with the magnitude and duration of BPPV due to canalithiasis. Theoretical models have also been used to predict particle movement and canal responses during diagnostic and therapeutic manoeuvres for BPPV [Rajguru et al., 2004; House and Honrubia, 2003]. In the future, it may be useful to combine these biomechanical models with direct recordings of head accelerations during the various particle repositioning manoeuvres.

 Appendix: Model

A simple model of head rotations was created to aid the understanding of the data plots.

For a rigid body, the acceleration of any point P on the body is given by equation 1 [Zappa et al., 2001].

where a_O is the acceleration of a point O on the body, is the angular velocity of the body, and (P - O) is the relative position of P with respect to O. For a fixed-axis rotation about an origin O: the first term on the right-hand side of the equation represents the acceleration due to gravity; the second term represents the tangential acceleration at P, and the third term represents the centripetal acceleration at P.

The particle repositioning manoeuvres were broken down into a series of rotation steps. Each step consisted of a fixed-axis rotation about an origin located at either the hips or the head. The amplitude, duration and rotation axis for each step were chosen to approximate real-life values. The angular acceleration during each swing was assumed to follow a sinusoidal temporal profile. Formulae for the rotation angle, angular velocity and angular acceleration during each swing are given in equations 2-4.

Rotation angle:

where = rotation angle; _max = amplitude of rotation step; t = time variable, and d = duration of rotation step.

The head position and orientation were defined in an earth-fixed right-handed Cartesian coordinate system, with the z-axis pointing upwards, along the gravitational vertical. The canal vectors c and b define the orientation of each SCC in space. In the Reid horizontal position, with the head facing the positive x-direction, the assumed canal vectors for the right SCC are given in table 1. The canal vectors at the start of a positioning manoeuvre were obtained from these defined vectors, by applying the rotation operation that would be required to take the head from the Reid horizontal position to its orientation at the start of the manoeuvre.

Table 1. Canal vectors for the right SCC

The head position, head orientation and SCC orientations over the time course of the manoeuvre were calculated by applying the sequence of defined rotations to the initial conditions. The acceleration vector at each point in time was derived from equation 1. Then the acceleration components along the canal vectors (a_c and a_b) were obtained from vector dot products. Plots of a_b versus a_c over time, for each SCC, could be compared with the measured cb-plots.

The model prediction for the cb-plot in the plane of the right posterior canal during an Epley manoeuvre is shown in figure 9. The durations of each rotation step were taken to approximate those of the measured data. The total Hallpike manoeuvre duration (for a 90° rotation about the hips, followed by a 30° neck extension) was 2.2 s. The head turns were each assumed to take 1.2 s. The return to the upright position took 2.4 s. The model prediction is a fairly good representation of the equivalent data plot (fig. 6).

The model results for three Semont manoeuvres are illustrated in figure 10. All of the manoeuvres began with a 90° Hallpike manoeuvre and ended with a sitting-up movement, each of 2 s duration. The 180° swing in the middle of the manoeuvre was designated to be slow, medium or fast, with durations of 3.6, 1.5 and 1.3 s, respectively. These durations were chosen to approximate those of the data plotted in figure 8. It can be seen that the manoeuvre with the fast swing provides the best approximation to a clockwise rotation of the acceleration around the unit circle. For the slow manoeuvre, the acceleration during the 180° swing tends to follow a counter-clockwise rotation around the unit circle.

  References

1.

Blanks RH, Curthoys IS, Markham CH: Planar relationships of the semicircular canals in man. Acta Otolaryngol 1975;80:185-196.

2.

Dix MR, Hallpike CS: The pathology, symptomatology and diagnosis of certain common disorders of the vestibular system. Proc R Soc Med 1952;45:341-354.

3.

Epley JM: The canalith repositioning procedure: for treatment of benign paroxysmal positional vertigo. Otolaryngol Head Neck Surg 1992;107:399-404.

4.

Furman JM, Cass SP, Briggs BC: Treatment of benign positional vertigo using heels-over-head rotation. Ann Otol Rhinol Laryngol 1998;107:1046-1053.

5.

Furuya M, Suzuki M, Sato H: Experimental study of speed-dependent positional nystagmus in benign paroxysmal positional vertigo. Acta Otolaryngol 2002;123:709-712.

6.

Hall SF, Ruby RRF, McClure JA: The mechanics of benign paroxysmal vertigo. J Otolaryngol 1979;8:151-158.

7.

House MG, Honrubia V: Theoretical models for the mechanisms of benign paroxysmal positional vertigo. Audiol Neurootol 2003;8:91-99.

8.

Kleppner D, Kolenkow RJ: An Introduction to Mechanics (International Student Edition). Singapore, McGraw-Hill, 1978, p 346.

9.

Lempert T, Gresty MA, Bronstein AM: Benign positional vertigo: recognition and treatment. BMJ 1995;311:489-491.

10.

Lempert T, Wolsley C, Davies R, Gresty MA, Bronstein AM: Three hundred sixty-degree rotation of the posterior semicircular canal for treatment of benign positional vertigo: a placebo-controlled trial. Neurology 1997;49:729-733.

11.

Nakayama M, Epley JM: BPPV and variants: improved treatment results with automated, nystagmus-based repositioning. Otolaryngol Head Neck Surg 2005;133:107-112.

12.

Radtke A, von Brevern M, Tiel-Wilck K, Mainz-Perchalla A, Neuhauser H, Lempert T: Self-treatment of benign paroxysmal positional vertigo: Semont maneuver vs Epley procedure. Neurology 2004;63:150-152.

13.

Rajguru SM, Ifediba MA, Rabbitt RD: Three-dimensional biomechanical model of benign paroxysmal positional vertigo. Ann Biomed Eng 2004;32:831-846.

14.

Rajguru SM, Rabbitt RD: Afferent responses during experimentally induced semicircular canalithiasis. J Neurophysiol 2007;97:2355-2363.

15.

Semont A, Freyss G, Vitte E: Curing the BPPV with a liberatory maneuver. Adv Otorhinolaryngol 1988;42:290-293.

16.

Solomon D: Benign paroxysmal positional vertigo. Curr Treat Options Neurol 2000;2:417-428.

17.

Viirre E, Purcell I, Baloh RW: The Dix-Hallpike test and the canalith repositioning maneuver. Laryngoscope 2005;115:184-187.

18.

Zappa B, Legnani G, van den Bogert AJ, Adamini R: On the number and placement of accelerometers for angular velocity and acceleration determination. J Dyn Syst Meas Control 2001;123:552-554.

  Author Contacts

Dr. M.E. Faldon
Department of Clinical Neuroscience, Division of Neuroscience and Mental Health
Imperial College, Charing Cross Hospital
Fulham Palace Road, London W6 8RF (UK)
Tel. +44 20 8846 7349, Fax +44 20 8383 3630, E-Mail m.faldon@imperial.ac.uk

  Article Information

Received: August 3, 2007
Accepted after revision: February 20, 2008
Published online: June 5, 2008
Number of Print Pages : 12
Number of Figures : 10, Number of Tables : 1, Number of References : 18

  Publication Details

Audiology and Neurotology (Basic Science and Clinical Research in the Auditory and Vestibulary Systems and Diseases of the Ear)

Vol. 13, No. 6, Year 2008 (Cover Date: October 2008)

Journal Editor: Harris J.P. (San Diego, Calif.)
ISSN: 1420-3030 (Print), eISSN: 1421-9700 (Online)

For additional information: http://www.karger.com/AUD

  Drug Dosage / Copyright

Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in goverment regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center.