320 / 412
videostroboscopy, which is considered to be meaningful espe-
cially because a large part of vibratory characteristics of VFA
were reflected in the time parameters (eg, open quotient).
Second, with the use of HSDI, the rate of successful image
evaluation increased by 1.5 folds at a rough estimate because
videostroboscopic study was successful only in 67.3% because
of desynchronization in the present study. The relatively high
rate of desynchronization in VFA may be explained by their
poor acoustic profile. Patel et al
8
reported that HSDI could be
used to augment videostroboscopy for assessment of
moderate-to-severe dysphonia, especially in patients with jitter
exceeding 0.87%, shimmer exceeding 4.4%, and a signal-to-
noise ratio of less than 15.4 dB. In the present study, 21.7%
of the VFA group fitted these criteria.
Third, with the application of multiple analysis methods, the
present study documented the characteristics of vocal fold vi-
brations of VFA more extensively and multidirectionally than
previous reports, in which vocal fold vibrations were either
qualitatively evaluated or quantitatively evaluated with only
limited parameters.
5,6
Although HSDI has disadvantages in comparison with video-
stroboscopy such as a relatively long time required for analysis
(approximately 30 minutes per HSDI at present), a high cost,
and the lack of instantaneity (with videostroboscopy, the result
of modulation in F
0
, sound pressure level, or register can be
observed directly and instantaneously),
7,8,16
HSDI is
considered to be a good supplementary tool in the assessment
of VFA.
Amplitude and integral glottal width
The amplitude mean of VFA was comparable with that of
vocally healthy subjects, in this study, which was a consistent
result with the previous study.
5
Although not statistically signif-
icant, the amplitude of VFA was larger in males and smaller in
females than vocally healthy subjects. Various factors can affect
amplitude such as amplitude increases as intensity or subglottal
pressure increase, or as pitch or stiffness decrease.
6,9,16,17
In
female VFA, poor pulmonary function is reported to be
frequently associated,
2
which may lead to decreased subglottal
pressure and decreased amplitude. In male VFA, a greater
glottal flow and lower tension of the thyroarytenoid muscle
owing to the muscular atrophy can increase the amplitude.
6,18,19
On the other hand, N
GL
-integral glottal width demonstrated a
significant difference between the control and VFA groups. The
N
GL
-integral glottal width may be a sensitive parameter than
the amplitude
per se
because it has the characteristics of both
amplitude and open quotient (
Table 6
).
13
Open quotient and speed index
Significant intergroup differences were observed in O
q
SLK
and
O
q
MLK
but not in O
q
GAW
. This is probably because O
q
GAW
was
not a parameter to reflect the size of glottal gap (O
q
GAW
be-
comes one whether a glottal gap is small or large). Interestingly,
the results of O
q
SLK
were comparable with those of O
q
MLK
,
although O
q
MLK
that assesses the overall glottal area should
reflect the pathophysiology of the disease better than O
q
SLK
.
Perhaps, the midglottal level may represent the vibratory dy-
namics of overall glottis well enough in VFA, and the informa-
tion of the glottal ends included in O
q
MLK
may have been less
important. Correlation analysis revealed that high O
q
SLK
and
O
q
MLK
were associated with poor aerodynamic and acoustic
conditions (
Table 4
). These results seem to stand to reason
because weak glottal closure reflected in high open quotient
should lead to high glottal flow with high air turbulence.
Speed index of VFAwas smaller than that of vocally healthy
subjects. Small speed index in the VFA group may originate
from the decreased restorative force of the laterally displaced
vocal fold toward the medial direction resulting from the disar-
rangement of collagen fibers or decreased elastin fibers in the
lamina propria, the decreased mass or tension of the vocal
fold owing to the muscular atrophy.
1,18,19
Contrary to open
quotient, SI
GAW
was more sensitive than SI
SLK
or SI
MLK
,
probably because SI
GAW
reflects the general vibratory
dynamics than SI
SLK
or SI
MLK
(
Table 2
). Speed index had
similar relationships with acoustic and aerodynamic parameters
to open quotient (
Table 4
). These results accord with the find-
ings in the literature, reporting that smaller speed index leads
to poorer aerodynamic or acoustic results.
20,21
TABLE 6.
Correlation Coefficients (r) Among High-Speed Digital Image Parameters
Parameters
N
L
- Minimal GA SI
GAW
O
q
MLK
O
q
SLK
N
L
-Amplitude Mean N
G
-O-LPD
MLK
GA difference index
0.90
*
0.46
*
0.65
*
0.54
*
0.34
y
0.15
N
L
-minimal GA
1
0.33
y
0.69
*
0.47
*
0.24
0.05
SI
GAW
—
1
0.47
*
0.56
*
0.04
0.39
*
O
q
GAW
—
— 0.69
*
0.42
*
0.10
0.10
N
G
-O-LPD
LTG
—
— 0.14
0.19
0.03
0.50
*
O
q
MLK
—
— 1
0.65
*
0.13
0.19
N
GL
-integral glottal width
—
— — 0.51
*
0.51
*
0.28
N
L
-MW magnitude mean
—
— — —
0.59
*
0.03
Abbreviations:
GA, glottal area; N
L
-, normalized by vocal fold length; SI, speed index; GAW, in glottal area waveform; O
q
, open quotient; N
G
-, normalized by
glottal cycle; O-LPD, opening longitudinal phase difference; LTG, in laryngotopography; MLK, in multi-line kymography; N
GL
-, normalized by glottal cycle and
vocal fold length; MW, mucosal wave; SLK, in single-line kymography.
*
P
< 0.001.
y
P
< 0.01.
Journal of Voice, Vol.
-
, No.
-
, 2015
95