Journal of Oral Rehabilitation, 1993, Volume 20, pages 3 3 - 4 3 ' ' •"•.

: - < 'v --

The recording and analysis of EMG and jaw

tracking. II. reproducibility of jaw tracking
p . G . T . H O W E L L , S. E L L I S , C . W . L . J O H N S O N , I.B WATSON*
and I . K L I N E B E R G Department of Prosthetic Dentistry, University of Sydney and *Dental
School, University of Glasgow

Summary
In 10 male and 10 female adults jaw movements and muscle activity were monitored
during chewing. Subjects had a Class I occlusion and fluent unrestricted jaw movements;
none exhibited signs or symptoms of craniomandibuiar disorders. Recordings were
made on two occasions separated by 2 weeks. Within each recording session the
subjects carried out a number of tasks under direction, including eating nuts and gum.
The jaw movements were monitored with a Sirognathograph® which had been interfaced
to a computer for data logging and analysis.
A computer program identified each chewing stroke and after scrutiny a data base
was constructed of measurements from a subset of 4447 suitable strokes. Mean values
of 10 parameters selected to describe jaw movements during chewing were entered
into an ANOVA. This indicated that the major contribution to variance was due to the
subject from whom the data was recorded. This intersubject variability is well known,
but can its separation from the other contributions to the variation within the data
be assessed? To provide an estimate of the spread of data the variance for each
parameter was calculated, transformed and these values entered into an analysis
of variance.
All subject differed significantly in their chewing movements from one another.
But 15 of the 20 subjects showed a greater degree of consistency than the other five:
they exhibited good inter- and intra-recording reproducibility. The food being eaten
made a major contribution to the variability. The timing of the chewing strokes was
also more consistent than the dimensions of the chewing envelope.

Introduction
Jaw movements in individuals have been known to differ widely; reports (Hildebrand,
1931; Jankelson etal, 1953; Kydd & Sandier, 1961; Bates etal, 1975a & b; Proschel,
1987) attest to this. Both occlusal scheme (Beyron, 1964; Gibbs et al, 1971; Gillings
etal, 1973; Neill & Howell, 1986; Michler etal, 1987; Proschel, 1987), and the sex of
the subject (Howell, 1987) have been shown to effect mandibular displacement. George
(1983) suggested that to provide some degree of consistency within the measurements
of jaw movements, a standardized occlusal relationship should be adopted.
A number of recent studies have been critical of claims by manufactures of
tracking devices, concerning the possible diagnostic and prognostic value of jaw
tracking measurement (Feine, Hutchins & Lund 1988; Lund & Widmer 1989). These

Correspondence: Professor Iven Klineberg, Department of Prosthetic Dentistry, University of Sydney,

2 Chalmers Street, Sydney, NSW 2010, Australia.

33
34 P. G.T. Howell et al. -•_. ^^^^^^^i:. :

are reasonable concerns. However it is also reasonable to hypothesise that the accurate
monitoring of jaw movement may provide unique evidence of jaw muscle and jaw
joint function and that such information may be of diagnostic and prognostic value.
Management of craniomandibuiar disorders (CMDs) is still largely based on clinical
assessment, and where a diagnosis cannot be made with confidence, diagnostic ap-
pliances (occlusal splints) are often fitted in the hope that they may help the problem.
It is important for clinicians to confidently diagnose psychogenic, muscular or articular
CMDs in order for management strategies to be appropriate and predictable. Objective
assessments or special tests, such as specialised imaging, analysis of study casts etc, in
conjunction with history and clinical examination are known to assist diagnosis. Jaw
tracking also may be valuable to assess jaw muscle and jaw joint function.
In order for jaw tracking to be acceptable, it must be carried out in a standardised
manner, and specific movements measured, which have been shown in controlled
clinical recordings to be reproducible. As with all human movement there is a broad
range of intra- and inter-individual values. It is accepted then, that within a group
of individuals there is a considerable degree of variability of chewing movements.
However, it has yet to be established what degree of variability exists for an individual
when they are observed from one day to the next or over a greater time scale. If an
alteration was made to an individual's teeth (such as by restorations) during this
intermediary period, would it be possible to determine whether any difference in the
pattern of chewing movements was due to this specific change in the system, or
whether it was due to the inherent variability of the individual's chewing movements.
This study was designed to measure functional mandibular movements for a number
of individuals using a standardized protocol and recording equipment and to repeat
these measurements at a fixed time later. It would then be possible to test the re-
producibility of jaw movement within a single recording session or between recording
sessions, to determine which measures of jaw movement gave the greatest repro-
ducibility, and which variables in the recording procedure contributed most to the
differences in the recorded values.

Method
Twenty subjects, 10 male and 10 female, participated in this study and had an age
range between 19—63 years with a mean age for men of 35-2 years and for women
of 36-9 years. All subjects had unrestricted fluent jaw movements, no evidence of
CMDs, an Angle Class I incisor relationship and a minimum of 26 natural teeth.
No dental extractions or restorative procedures had been carried out during the
previous 12 months.
The experimental arrangement has been described in detail (Howell et al, 1992).
Briefly a Sirognathograph* was used to monitor the position of the lower incisor teeth
in three dimensions. Muscle activity was measured bilaterally form three muscles with
either surface disc or intradermal hook electrodes. The subjects were either given a
free choice of chewing side or directed to chew on the left side or right side. They
chewed either half a peanut or half a stick of chewing gum. The jaw movements and
EMG activity were continuously monitored during these tasks with a PDP-11 computer!
sampling the signals at a rate of 1 KHz. The data was archived and analysed off line

* Siemens AG, Bensheim, Germany.

t General Robotics (Australia) Pty Ltd., Sydney, Australia.
 EMG and jaw tracking II 35

from the main experiment. Each subject repeated each of the six chewing tasks (two
Foods and three directed chewing Actions). A second recording session was made
14 days later.
A computer program identified the individual chewing cycles. Ten measurements
of each of these strokes were made. Five parameters described the timing of jaw
movements, two were measures of the speed of movement and three were the dis-
placements in the A-P, lateral and vertical planes. They were:
(i) total time of the chew (Time total);
(ii) time of the opening movement (Time open);
(iii) time of the closing movement (Time close);
(iv) time of the intercuspal pause (Time dwell);
(v) time of the movement phase of chew (Time move);
(vi) maximum velocity during opening (Max.Op Vel.);
(vii) maximum velocity during closing (Max.Cl. Vel.);
(viii) range of antero-posterior movement (AP range);
(ix) range of horizontal movement (Lat. range);
(x) range of vertical movement (Vert, range).
Each chewing stroke was carefully examined using the numeric data output, the
time-displacement plots and the plots of the individual strokes in the frontal, lateral
and horizontal planes. Unwanted strokes were identified as those not displaying fluent
unrestricted movement, but rather an irregular path during jaw opening or closuring.
Similarly identified were those strokes where the subject did eat on the directed side
during unilateral chewing; or where there was an unusually long, or nonexistent tooth
contact (or dwell) time, uncharacteristic of that subject; or where there was limited
or excessive movement indicative of speech or a clearing stroke. These unwanted
strokes were noted and rejected from the data pool. The first stroke of each sequence
was routinely rejected as being atypical. This provided a sample of approximately
10 chewing strokes for each subject for each task (Action) and each food (Food)
eaten for each repeated recording (Run) for each repeated recording session (Session).
Means, standard deviations and variances were calculated for each subject, for
each parameter for Action, Food, Run and Session. The mean values were entered
into an analysis of variance (ANOVA) to test the null hypothesis that there was no
difference in the recorded jaw movements either as a result of the subject, the
food being chewed, the manner in which it was chewed, or between recordings made
within the same recording session or between recording sessions.
An indicator of the variation with the measured data for any parameter is the
variance. If the variation of data within a sample is to be studied and its variability to
be ascribed to a factor associated with the protocol design, then the variance values
should be entered into an ANOVA. However, variance does not follow a normal, but
rather a chi-square distribution. The variance data must then be transformed using
the following transformation (Gebsky, Personal communication, 1991): —

(variance — degrees of freedom)

Modified variance =
sqrt (2* degrees of freedom)
The variance data was transformed to the above equation and entered into an

" Val Gebski, Director, Statistical Laboratory, Macquarie University, Sydney 2109, Australia.
36 P.G.T. Howell et al.

ANOVA. The initial hypothesis to be tested was that for the group as a whole, the
(modified) variance did not depend upon the Subject, Session, Run, Action or Food
used in the test. Similarly to be tested was that for an individual, there was no dif-
ference in the variance as a result of the Session, Run, Action or Food upon which
the data was calculated. The number of occasions on which a level of significance
(P>0-01) was obtained for this latter hypothesis was aggregated for each parameter.
The occurrence of a significant level being reached for the higher order effect of the
product of Food and Action was also examined.

Results

Mean value data

From the total data pool of chewing strokes by the 20 individuals in this study, 4447
chewing strokes were identified as satisfying the prescribed criteria and were analysed.
The mean and standard deviations were calculated for the 20 subjects for each of the
10 parameters used here to assess jaw movement and the results are shown in Table
1. Additionally this table shows the range of means calculated for each individual.
There is wide variation in the data indicated both by the extensive range of individual
values, and by the large standard deviation for the sample as a whole.
The results from entering the mean values into an ANOVA are shown in Table 2.
For each of the 10 parameters there is a significant (F>0-01) contribution to the
variation in the data due to the individuals of the sample group (Subject). The ANOVA
also showed that while Run and Session showed infrequent instances of significant
differences (3 out of 10 in each case). Action and Food gave numerous occurrences
of significance (7 and 8 out of the 10 parameters respectively). With Action these
significant differences were found with the timing and speeds of movement, while
for Food they were spread across the whole range of parameters.

Modified variance data — (all subjects together)

The results of the modified variance data entered into the ANOVA is shown in Table 3.
Again the main effect can be seen to be due to the individual (Subject) which was
significant (P>0-01) for all of the 10 measured parameters.

Table 1. The mean and standard deviations for the ten parameters from the 20 subjects of this study.
The range for the individuals is also displayed

Parameter Mean SD Range

Time total (ms) 920 200 568-1641

Time open (ms) 340 89 141-613
Time close (ms) 373 76 234-680
Time dwell (ms) 206 82 30-571
Time move (ms) 714 146 418-1196
Max. Op. Vel. (mms"') 83 24 35-150
Max. Cl. Vel. (mms~') 79 19 41-153
AP. range (mm) 3-3 1'2 0-7-8-1
Lat. range (mm) 5'3 19 M-10-7
Vert, range (mm) 13-7 2-4 7-3-21-7
 EMG and jaw tracking II 37

Table 2. This shows the occurrences of significance when the mean value data is entered into an
ANOVA for the variables Subject, Session, Run, Action and Food. Where — = not significant and
+ = significant (P>0-01)

Parameter Subject Run Session Action Food

Time total -1- -1- -I-

Time open -1-
Time close -h -I-
Time dwell
Time move -1- -I-
Max. Op. Vel. -f-
Max. Cl. Vel -1- -I-
AP. range
Lat. range
Vert, range

Table 3. This shows the occurrences of significance when the modified variance data for the 20
individuals is grouped together and entered into an ANOVA for the variables Subject, Session, Run,
Action and Food. Where - = not significant and + = significant (P>0-01)

Parameter Subject Run Session Action Food

Time total
Time open
Time close
Time dwell
Time move
Max. Op. Vel. -I-
Max. Cl. Vel 4-
AP. range -I-
Lat. range -I-
Vert, range

Food was significant for six of the 10 parameters, three movement timings, both
velocities and the vertical range of movement. Action showed significance for both
speeds of movement plus lateral and vertical range of movement. Session and Run
only show a single instance (for Session) of significance in the measured parameters
(AP range).

Modified variance data — {individuals treated separately)

When the individuals are treated separately to investigate the effect upon the (modified)
variance by Session, Run, Action, Food through an ANOVA, the results shown in Table
4 are obtained.
In total there were 41 instances of significance (P>0-01) from a potential 800
occurrences for the 20 subjects, for the 10 parameters examined for the effects of
Session, Run, Action, Food (Table 4). Some subjects were found to be less consistent
than others. The number of occasions of significance was found to be generally low,
with a maximum of 7 (7/40) for the least reproducible subject. Three quarters of
38 P.G.T. Howell et al.

c
re

S-5
to o
^ A

Sg <

if 3 «

•3
i
o

C3
T3 o
^
o
re

> Sri 00

c c
o re

<

00
o .9

O 4-
r- C c
.2 o O X5 . 2 o
o -o .2 o -a
•r= ? ^ 5 '5 o Jg § "S o

C3
O O o
o •a
p
o
X! O
I H
E
H
 EMG and jaw tracking II 39

u
o H
D
O 00
X
o oo
05

o <
X

+ • o
+
o
4- o
+ Q

4-

4-

4- 4- U
oo
4- 4- 4- en

4- en i

4- 4- 4- 00
00
4- 4- 4-

4- 4- ^
N
4- 4-
c c c
. Vel . Session

Sess ion

Sess ion

Sess ion
Vel. Sess ion

o 5 o -a o 5 _
T3 •n •a
Acti

Acti
Fooi

Fooi

Fooi
Acti

Run

Run

Acti

Acti
Run

Run

Run
Fooi

Tote
Foo

o
4) ec
a. 60 c
C re
0 U IS rt u
u u
X >< . . -^^
re re
<
40 P.G.T. Howell et al.

the subjects studied had fewer than four (4/40) instances of significances, and half
the subjects had either only a single or no significant differences at all.
Food produced the greatest number of significant differences (18) for the 10
measured parameters and was nearly twice the number found for Session and Action
(11 and 10 respectively), while Run yielded only two instances (Table 5). There was
no significant differences noted as a result of the higher order effect due to the Action
*Food interaction.
The vertical and lateral ranges (Vert, range and Lat. range) of the 10 parameters
used to describe jaw movement had nearly twice the number of significant differences
(9) than the next group of seven parameters (AP range. Max. Cl. Vel, Time total.
Time move. Time dwell. Time close, with 5, 4 and 3). While the remaining two
parameters (Time open and Max. Op. Vel) had a single or no significant differences
respectively.

Discussion
The experimental method has been shown to be satisfactory in its ability to record
accurately jaw movement and EMG activity in real time (Howell et al, 1992) An
identical protocol was used for each subject with the recordings being carried out on
the same day of the week and at the same time of day, to standardise the recording
procedure. As well a calm, temperature-controlled environment existed. Few of the
previous investigations have studied large groups of subjects; exceptions being Howell
(1987) and Proschel & Hoffmann (1988); this study examined 20 subjects with a
standard occlusal scheme.
Considerable care was taken to reproduce the orientation of the headframe of the
Sirognathograph® between recording sessions. However, it is possible that some error
due to its misahgnment may have occurred. Michler et al. (1987) and Maruyama et al.
(1984) tested the Sirognathograph® and found that a 10° rotation of the magnet caused
a 1mm displacement. Additional errors due to head movement during the recording
are always possible. This was reduced by providing that the subject's head was sup-
ported throughout the recordings and furthermore they were asked to look straight

Table 5. Summary of the results from entering the modified variances data for an individual
into an ANOVA. Each occasion that a significance (f >0-01) was obtained was noted and the
total number of instances aggregated for each parameter due to Session, Run, Action, Food

Parameter Run Action Session Food Total

Max. Op. Vel. 0 0 0 0 0

Time open 1 0 0 0 1
Time close 0 0 1 2 3
Time dwell 0 0 2 1 3
Time move 0 1 1 1 3
Time total 1 2 0 1 4
Max. Cl. Vel. 4
T-H

0 1 2
AP. range 0 1 2 2 5
Lat. range 0 4 1 4 9
Vert, range 0 1 3 5 9
Total 2 10 11 18 41
Max = 800
 EMG and jaw tracking II 41

ahead at a point immediately to their front. Some variation in the mean value data
could be due to this and would account for significant differences due to Session and
to a lesser extent Run (Table 2).
The two foods used as test materials have very different consistencies but are
readily available and have been widely used previously. The mean values found in
this study for the parameters selected to describe functional mandibular movements
during mastication fall within the range of values reported elsewhere (Shepherd,
1960; Schweitzer, 1961; Woelfel et al, 1962; Beyron, 1964; Ahlgren, 1966 & 1967;
M0ller, 1966; Rudd et al, 1969; Gillings et al, 1973; Bates et al, 1976; Hannam
et al, 1977; Hannam and Wood, 1977; Jemt et al, 1979; Lucas et al, 1986; Neill and
Howell, 1986; Plesh et al, 1986; Howell, 1987; Michler et al, 1987; Proschel, 1987;
Proschel and Hoffman, 1988; Feine et al, 1988; Horio and Kawamura, 1989).
The mean value data confirms the inherent variability of an individual's movements
during mastication (Tables 1 & 2). A contribution to this variability comes from the
task (Action) that is being performed and the consistency of the Food being chewed.
However differences due to the repetition of recording of these movements either
within a single session (Run) or between recordings (Session) taking place with a
wide time difference are minimal. These data indicate that it is necessary to look at
the variability of an individual rather than a number of individuals grouped together.
The variability of the individual was examined through a transformation of the
variance. The ANOVA of the modified variance confirms the individuality of chewing,
it also shows that the Food being eaten has a more major contribution to the vari-
ability than had been previously thought (Tables 3 and 5). The Action, Session or Run
made little effect upon the variance. None of the subjects studied exhibited any sign
or symptom of CMDs, nor had they had any alteration to their stomatognathic system
between recordings. It is not surprising that there was good inter- and intra-session
recording reproducibility. Even the most variable of the subjects had only seven out
of 40 potential instances of significant difference, while five (25%) achieved worse
than four in 40 occurrences. Conversely 15 (75%) showed good reproducibility.
The three measures of the dimensions of the chewing envelope (AP range, Lat.
range and Vert, range) were the least consistent of the 10 parameters selected to
describe jaw movement (Table 5). The timing of the movements was in general more
consistent. The implication being that jaw movement occurs with a great degree of
rhythmicity, but is not possible to repeatedly position the jaw with great accuracy.

Conclusions
All people are variable, but some are more variable than others (modified from
Animal Farm by George Orwell).
All subjects showed differences in the dimensions and timing of their chewing
movements under controlled conditions and performing standard directed tasks. Within
an individual there was a wide degree of variability for all 10 parameters selected to
describe jaw movement. This variability was examined by transforming the variance
data. 75% of the subjects showed that while they might be inconsistent and have a
wide range in their jaw movements these were similar under similar circumstances
ie they showed little inter- and intra-recording variability. Changing the Food being
eaten made the greatest difference to the variability of jaw movement. The measures
of the dimensions of the chewing envelope showed the greatest variability whilst the
timing of the movements was less so.
42 P.G.T. Howell et al.

Acknowledgments
We should like to thank Dr Val Gebsky, Department of Statistics, Macquarie
University, for his help in the statistical analysis of the data and Mr Bruno Nicoletti,
Technical Officer, Department of Prosthetic Dentistry, for his assistance during re-
cordings and his electronics support.

References
AHLGREN, J. (1966) Mechanism of mastication: A quantitative cinematographic and electromyographic
study of masticatory movements in children with special reference to occlusion of teeth. Acta
Odontologica Scandinavica, 24, Supplement 44.
AHLGREN, J. (1967) The pattern of chewing and malocclusion of teeth. A clinical study. Acta
Odontologica Scandinavica, 25, 3.
BATES, J.F., STAFFORD, G.D. & HARRISON, A. (1975a) Masticatory function — a review of the
literature. I The form of the masticatory cycle. Journal of Oral Rehabilitation, 2, 281.
BATES, J.F., STAFFORD, G.D. & HARRISON, A. (1975b) Masticatory function — a review of the
literature. II Speed of movement of the mandible, rate of chewing and forces developed in
chewing. Journal of Oral Rehabilitation, 2, 349.
BATES, J.F., STAFFORD, G.D. & HARRISON, A. (1976) Masticatory function — a review of the
literature. Ill Masticatory performance and efficiency. Journal of Oral Rehabilitation, 3, 57.
BEYRON, H.L. (1964) Occlusal relations and mastication in Australian aborigines. Acta Odontologica
Scandinavica, 22, 597.
FEINE, J.S., HUTCHINS, M . O . & LUND, J.P. (1988) An evaluation of the criteria used to diagnose
mandibular dysfunction with the mandibular kinesiograph. Journal of Prosthetic Dentistry, 60, 374.
GEORGE, J.P. (1983) Using the kinesiograph to measure mandibular movement during speech: a pilot
study. Journal of Prosthetic Dentistry, 49, 263.
GIBBS, C.H., MESSERMAN, D.T., RESWICK, J.D. & DERDA, H.J. (1971) Functional movements of the
mandible. Journal of Prosthetic Dentistry, 26, 604.
GILLINGS, B.R.D., GRAHAM, C.H. & DUCKMANTON, N . A . (1973) Jaw movement in young adults
during chewing. Journal of Prosthetic Dentistry, 29, 616.
HANNAM, A.G., SCOTT, J.D. & D E COU, R.E. (1977) A computer based system for simultaneous
measurement of muscle activity and jaw movement during mastication in man. Archives of Oral
Biology, 22, 17.
HANNAM, A.G. & WOOD, W . W . (1977) Muscle activity and jaw displacement during unilateral
chewing in Man. Journal of Dental Research, 56, (Sp Issue 6) 230.
HILDEBRAND, G . Y . (1931) Studies in the masticatory movements of the lower jaw. Scandanavian
Archives of Physiology, Supplement 61.
HORIO, T . & KAWAMURA, Y . (1989) Effects of food texture on chewing pattern in the human subjects.
Journal of Oral Rehabilitation, 16, 177.
HOWELL, P.G.T. (1987) Sexual dimorphism in mastication and speech? Or do men and women eat
and talk differently? Australian Prosthodontic Journal, 1, 9.
HOWELL, P.G.T., JOHNSON, C.W.L., ELLIS, S.D., WATSON, I.B. & KLINEBERG, I. (1992) The recording
and analysis of EMG and jaw tracking. I. the recording procedure. Journal of Oral Rehabilitation,
19, 595.
JANKLESON, B . , HOFFMAN, G.M. & HEDRON, J.A. (1953) The physiology of the stomatognathic
system. Journal of the American Dental Association, 46, 375.
JEMT, T . , KARLSSON, S. & HEDEGARD, B . (1979) Mandibular movements of young adults recorded by
intraorally placed light emitting diodes. Journal of Prosthetic Dentistry, 42, 669.
KYDD, W . L . & SANDLER A. (1961) A study of the posterior mandibular movements from the inter-
cuspal occlusal position. Journal of Dental Research, 40, 419.
LUCAS, P.W., LUKE, D.A., VOON, F.C.T., CHEW, C.L. & Ow, R.K.K. (1986) Food breakdown
patterns produced by the human subject possessing artificial and natural teeth. Journal of Oral
Rehabilitation, 13, 205.
LUND, J.P. & WIDMER, C . G . (1989) An evaluation of surface electromyography in the diagnosis,
documentation and treatment of dental patients. Journal of Craniomandibuiar Disorders Facial
and Oral Pain, 3, 125.
MARUYAMA, T . , KUWABARA, T . , NAKAMURA, Y . , AKANISHI, M . , MIYAUCHI, S. & KURODA, T . (1984) A
 EMG and jaw tracking II 43

new mandibular movement recording and analysing system composed of Sirognathograph and a
personal computer and its clinical application. Journal of the Osaka University Dental School,
24, 97.
MICHLER, L . , BAKKE, M . & M0LLER, E . (1987) Graphic assessment of natural mandibular movements.
Journal of Craniomandibuiar Disorders, 1, 97.
M0LLER, E. (1966) The chewing apparatus. Acta Physiologica Scandinavica, 60, Supplement 280.
NEILL, D.J. & HOWELL, P.G.T. (1986) Computerised kinesiography in the study of mastication in
dentate subjects. Journal of Prosthetic Dentistry, 55, 629.
PLESH, O . , BISHOP, B . & MCCALL, W . (1986) The effect of gum hardness on chewing patterns.
Experimental Neurology, 92, 502.
PROSCHEL, P. (1987) An extensive classification of chewing patterns in the frontal plane. Journal of
Craniomandibuiar Practice, 5, 55.
PROSCHEL, P. & HOFFMANN, M . (1988) Frontal chewing patterns of the incisor point and their
dependence on resistance of food and type of occlusion. Journal of Prosthetic Dentistry, 59, 617.
RUDD, K . , MORROW, R . & JENDRESEN, M . (1969) Fluorescent photo-anthropometry. A method for
analysing mandibular motion. Journal of Prosthetic Dentistry, 21, 495.
SCHWEITZER, J.M. (1961) Masticatory function in man. Journal of Prosthetic Dentistry, 11, 625.
SHEPHERD, R.W. (1960) A further report on mandibular movement. Australian Dental Journal, 5, 337.
WOELFEL, J.B., HICKEY, J.C. & ALLISON, M . C . (1962) Effect of posterior tooth form on jaw and
denture movement. Journal of Prosthetic Dentistry, 12, 922.