UNIVERSIDADE ESTADUAL PAULISTA
“JÚLIO DE MESQUITA FILHO”
Instituto de Ciência e Tecnologia
Campus de São José dos Campos
ORIGINAL ARTICLE DOI: https://doi.org/10.4322/bds.2023.e3649
1
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
Análise de textura por ressonância magnética da articulação temporomanbibular com alterações no disco articular em
individuos com cefaleia migranea
Karolina Aparecida Castilho FARDIM
1
, Thais Mara Aparecida Monfredini RIBEIRO
1
,
Elaine Cristina de Carvalho Beda Correa de ARAÚJO
1
, Celso Massahiro OGAWA
2
, Andre Luiz Ferreira COSTA
2
,
Sérgio Lucio Pereira de Castro LOPES
1
1 - Departamento de Diagnóstico e Cirurgia, Faculdade de Odontologia de São José dos Campos, Universidade Estadual Paulista
(UNESP), São José dos Campos, SP, Brazil.
2 - Programa de Pós-Graduação em Odontologia, Universidade Cruzeiro do Sul (UNICSUL), São Paulo, Brazil.
How to cite: Fardim KAC, Ribeiro TMAM, Araújo ECCBC, Ogawa CM, Costa ALF, Lopes SLPC. Magnetic resonance imaging texture
analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache. Braz Dent Sci.
2023;26(1):e3649. https://doi.org/10.4322/bds.2023.e3649
ABSTRACT
Objective: the aim of this study was to analyse the performance of the technique of texture analysis (TA) with
magnetic resonance imaging (MRI) scans of temporomandibular joints (TMJs) as a tool for identication of
possible changes in individuals with migraine headache (MH) by relating the ndings to the presence of internal
derangements. Material and Methods: thirty MRI scans of the TMJ were selected for study, of which 15 were
from individuals without MH or any other type of headache (control group) and 15 from those diagnosed with
migraine. T2-weighted MRI scans of the articular joints taken in closed-mouth position were used for TA. The
co-occurrence matrix was used to calculate the texture parameters. Fisher’s exact test was used to compare the
groups for gender, disc function and disc position, whereas Mann-Whitney’s test was used for other parameters. The
relationship of TA with disc position and function was assessed by using logistic regression adjusted for side and
group. Results: the results indicated that the MRI texture analysis of articular discs in individuals with migraine
headache has the potential to determine the behaviour of disc derangements, in which high values of contrast,
low values of entropy and their correlation can correspond to displacements and tendency for non-reduction
of the disc in these individuals. Conclusion: the TA of articular discs in individuals with MH has the potential
to determine the behaviour of disc derangements based on high values of contrast and low values of entropy
KEYWORDS
Texture Analysis; Migraine disorders; Temporomandibular disorders; Radiomics; Temporomandibular Joint Disc.
RESUMO
Objetivo: o objetivo deste estudo foi analisar o desempenho da técnica de análise de textura (AT) em exames de
ressonância magnética (RM) das articulações temporomandibulares (ATM) como ferramenta para identicação
de possíveis alterações em indivíduos com cefaléia migrânea (CM) relacionando os achados com a presença de
desarranjos internos. Material e Métodos: trinta exames de RM das ATM foram selecionados para estudo, sendo
15 de indivíduos sem cefaleia migrânea ou qualquer outro tipo de cefaléia (grupo controle) e 15 diagnosticados
com CM. As imagens de RM ponderadas em T2 das articulações realizadas na posição de boca fechada foram
usadas para AT. A matriz de co-ocorrência foi usada para calcular os parâmetros de textura. O teste exato de
Fisher foi usado para comparar os grupos quanto ao sexo, função do disco e posição do disco, enquanto o teste de
Mann-Whitney foi usado para os demais parâmetros. A relação da AT com a posição e função do disco foi avaliada
por meio de regressão logística ajustada para lado e grupo. Resultados: a AT por RM dos discos articulares em
2
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
INTRODUCTION
The temporomandibular joint (TMJ) is a
synovial articulation consisting of the condyle
(mandibular head), tubercle (articular eminence)
and mandibular fossa. These anatomical structures
make the interaction between temporal bone at
the skull base and the mandible, thus justifying
why TMJ is also called craniomandibular joint [1].
Temporomandibular dysfunctions (TMDs)
are considered changes in the masticatory muscles
and TMJ, or both [2]. There is a significantly
relevant number of patients with TMD-related
disorders affecting directly their quality of life [3,4].
Moreover, these disorders are considered one of
the most common chronic conditions of non-
dental orofacial pain seen by dental surgeons and
other healthcare professionals [5]. Patients with
TMD can present several signs and symptoms,
among them the most frequent are the following:
pain in the TMJ region, headache, pain in the
masticatory muscles, ear pain, facial pain,
functional limitation, cervical pain, difcult mouth
opening, pain during mastication, ear buzzing,
mandibular pain, among others [6,7].
Headaches are very common chronic mani-
festations in our society as they can incapacitate
the affected individuals by not allowing them
to perform their daily activities, thus being
considered a public health problem. They are
classied into primary and secondary types, in
which the former has an idiopathic origin as
clinical and laboratory examinations cannot
explain its aetiology. The so-called migraine
headache (MH), popularly known as migraine,
is an example of primary headache. As for the
secondary headache, pain occurs in response to
an injury, such as the tension headaches [8].
The tendency of an individual to have MH
is associated with genetic bases. In this sense, a
retrospective study [9] showed that MH is the
most common disease related to TMD. There
already exist studies correlating the presence of
internal defragment to MH, but none of them
analysed objectively the structures of the TMJ
components, such as the articular disc [10-12].
The imaging study of TMJ soft tissues, as is
the case of the articular disc, can be performed
with the gold-standard method, which consists of
using magnetic resonance imaging (MRI) scans.
MRI can determine the precise localisation of the
articular joint [13], thus allowing the study of its
position and shape so that possible changes can be
identied, such as presence of disc displacement
and internal derangements, which are dened
as changes in the normal relationship between
disc and bone components of the TMJ [14].
Additionally, MRI is a non-invasive technique
as it is based on image acquisition by means of
radiofrequency waves rather than of ionising
radiation, thus having no deleterious effect on
living organisms.
With the advance in the computing techniques,
a method to quantify efciently complex structures
in images on a non-invasive basis has been
developed by measuring the distribution of grey
levels in the region of interest (ROI) delimited in
the image. This technique is called texture analysis
(TA). Prior studies used TA to characterise lesions
in the several regions of the body and distinguish
them from normal tissues, since pathological
tissues present a greater heterogeneity in the
parameters of texture [15-17].
The objective of this study was to use the TA
in MRI scans as a tool to identify possible changes
in the articular disc of the TMJ in individuals with
MH due to internal derangements by comparing
the values obtained to those in the control group
of individuals without MH.
MATERIAL AND METHODS
All the steps of the study were carried out
in the Radiology Clinics of the Department of
Diagnosis and Surgery of the Paulista State
University Dental School (UNESP) in São José dos
indivíduos com cefaleia migrânea tem o potencial de determinar o comportamento dos desarranjos discais, em
que altos valores de contraste, baixos valores de entropia e sua correlação podem corresponder a deslocamentos
e tendência a não redução do disco nesses indivíduos. Conclusão: a análise de textura dos discos articulares
em indivíduos com CM tem potencial para determinar o comportamento dos desarranjos do disco com base em
altos valores de contraste e baixos valores de entropia.
PALAVRAS-CHAVE
Transtornos da Cefaleia; Análise de Textura; Transtornos da Articulação Temporomandibular; Radiômica; Disco
da Articulação Temporomandibular.
3
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
Campos, SP, Brazil. The study was approved by
the local research ethics committee according to
protocol number CAAE 32339720.8.0000.0077.
Fifteen MRI scans from individuals without
MH or any other headache (Control Group)
and 15 from individuals diagnosed with MH
(Migraine Group) were selected for study. All
the MRI scans are from the personal archive of
the main researcher, which had been used in a
previous study on MRI for analysis of MH [10].
The diagnosis of MH was made at the time of
image acquisition by a neurologist, who followed
the criteria of the International Classication of
Headache Disorders criteria [18].
All the scans were acquired by using a MRI
scanner 1.5 T (Sigma, General Electric, Milwaukee,
WI, USA) with an 8-channel bilateral surface coils,
in which parasagittal and lateral-medial images
were obtained perpendicularly to the long axis
of the mandibular head with mouth closed and
maximum mouth opening. The same protocol was
used for all image acquisitions as follows:
• T1: repetition time of 850 ms, echo time of
8.5 ms, section thickness of 3.0 mm, FOV of
150 x 150 mm and matrix of 512 x 512 pixels;
• T2: fat saturation: repetition time of 1500 ms,
echo time of 100.2 ms, section thickness of
3.0 mm, FOV of 150 x 150 mm and matrix
of 512 x 512 pixels;
Evaluation of the MRI scans was performed
in two steps, in which presence of internal
derangements (i.e. disc displacement and
changes in the disc function) was determined
rst and then the TA of the discs was calculated.
The MRI scans were evaluated on a 19-inch
monitor under reduced illumination by a
previously trained examiner using the Merge
eFilm Workstation software, version 1.5 (Merge
Healthcare, Chicago, IL, USA).
The disc position was evaluated on images
acquired with mouth closed and the disc function
on images acquired with maximum mouth
opening, according to the principles set by
Tasaki et al. [19].
MRI TEXTURE ANALYSIS OF THE AR-
TICULAR DISCS
MRI texture analysis was performed by
using MaZda software version 3.20 (Institute of
Electronics, Technical University of Lodz, Poland),
which consists of a specic package for this purpose.
The texture of articular discs was determined by
using parasagittal sections of the T2-weighted
images. Three consecutive sections corresponding
to lateral, central and medial views were chosen.
The reason for performing TA of T2-weighted
images is that the normal articular disc presents
hyposignal, whereas those with hypersignal
indicate presence of liquid. In this way, a change
in the articular disc generating a hypersignal in
T2-weighted images reects a region or tissue
containing liquid, such as oedema and hydropic
degeneration, which would involve intracellular
accumulation of water (cell hyper-hydration)
resulting from an imbalance in the osmotic
gradient control in the cytoplasmic membrane
and mechanisms of absorption and elimination of
water and intracellular electrolytes [20].
In both experimental and control groups,
each parasagittal section of the T2-weighted
image acquired with closed mouth was exported
in bitmap format (.bmp) before using the tool
“draw polygon” of the MaZda software version
3.20 to delimitate the entire articular disc in
order to determine the region of interest (ROI)
for TA (Figure 1).
TA is based on the so-called co-occurrence
matrix (COM) [21], which provides information
on the spatial relationship between the pixels of
the image within the ROI as determined by the
operator. The MaZda software can make variations
in the coordinates of the spatial relationship among
elements of this matrix in order to determine the
frequency of different information on the pixel
values analysed. Therefore, this software allows
selecting specic parameters whose meanings are
used to interpret the behaviour of the tissue, thus
providing values for TA.
Next, matrix parameters regarding the
directions of the horizontal (horzl), vertical
(vertl), 40-degrees (45 dgr) and 135-degrees
(135 dgr) pixels were selected so that pixel values
could be analysed in different spatial positions.
The values for each section were tabulated
and a mean value of the three sections (i.e.
lateral, central and medial) for each parameter
was obtained by segmenting the disc before
performing TA. This mean value corresponded to
the value to be analysed so that values covering
the whole extension of the disc could be obtained.
4
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
All the texture analyses were performed
by a previously trained evaluator in the in the
Radiology Clinics of the Department of Diagnosis
and Surgery of the School of Dentistry of UNESP.
The examiner was calibrated as follows: the
examiner performed the analyses of the position
and function of the discs by using 20 MRI scans
at intervals of 7 days. The resulting data were
submitted to Kappa test and a concordance
coefcient above 90% conrmed the examiner’s
calibration. The same procedure was performed
for TA, with the resulting data being submitted
to ICC test to be considered calibrated (90%).
Statistical analyses
Fisher’s exact test was used to compare the
groups regarding gender and disc’s function and
position, whereas Mann-Whitney test was used to
compare the other parameters. The comparison
between the groups was performed by side.
The relationship of TA with disc’s position and
function was assessed by using logistic regression
analysis adjusted for side and group. All statistical
analyses were performed by using the software R,
version 3.6.0 © (The R Foundation for Statistical
Computing).
Figure 1 – Example of ROI delimitation of the articular disc in a T2-weighted image of the TMJ and calculation of texture parameters by using
the MaZda software.
5
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
RESULTS
Thirty MRI scans of the TMJ were selected
for study, of which 15 were from individuals
without migraine (control group) and 15 from
those diagnosed with migraine (migraine group).
Control group had 11 (73.3%) females and the
migraine group had 13 (86.7%) ones.
There was no difference between the groups
regarding gender (
P-
value = 0.989; Fisher’s exact
test). There was also no signicant difference
between the groups regarding age (
P-
value =
0.651; Mann-Whitney test), with control and
migraine groups having the same mean age of
42.7 years old.
Twelve different distances were evaluated by
TA, with each distance having 11 parameters of
texture, totalising 132 parameters. Considering
the 30 MRI scans, there was a very great number
of texture parameters being evaluated. Therefore,
the following parameters were considered the
most important: AngScMom (angular second
moment), Contrast, Correlat (correlation),
InvDfMom (inverse difference moment),
SumOfSqs (sum of squares), SumEntrp (sum
of entropy) and Entropy. However, even if all
directions were evaluated, there would be a
very excessive number of variables given the
number of MRI scans in our sample. In this way,
Spearman’s correlation was calculated for all
parameters and dimensions, in which parameters
with correlation ≥ 0.75 or ≤ - 0.75 in relation to
other parameter were excluded, thus remaining
33 texture parameters in total.
Tables I-II present the descriptive measure-
ments and the comparison between the groups
on the right side. In Table I, one can observe
the median, minimum and maximum values,
including
P
-value in the comparison between the
groups on the right side, whereas in Table II one
can observe mean value and standard deviation
of the texture parameters by group
In Table I, one can observe that the migraine
group presented more disc displacements (
P
-value
= 0.009; Fisher’s exact test) and more changes
in disc function (non-reduced discs) (
P
-value =
0.007; Fisher’s exact test). With regard to TA,
the migraine group showed higher values for
the parameters S10Contrast and S22Contrast as
well as lower values for parameters S10Correlat
and S11Correlat.
Figure 2 shows the texture parameters which
presented statistically significant differences
between control and migraine groups on the
right side.
Figure 2 – Boxplots of the texture parameters which presented statistically significant difference between the groups on the right side.
6
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
Table I – Comparison between groups on the right side by using Fisher’s exact test or Mann Whitney’s test
Right Side
Parameter
Control (N=15) Migraine (N=15)
P-value
Median [Min; Max] Median [Min; Max]
Disc Position 0.009
Displaced 2 (13.3%) 10 (66.7%)
Normal 13 (86.7%) 5 (33.3%)
Disc Function 0.007
Non-reduced 1 (6.67%) 9 (60.0%)
Reduced 14 (93.3%) 6 (40.0%)
S10AngScMom 0.02 [0.01; 0.03] 0.03 [0.01; 0.10] 0.604
S10Contrast 38297872 [4.00; 84148936] 17234043 [0.53; 51808511] 0.019
S10Correlat 0.82 [0.63; 0.91] 0.94 [0.82; 0.98] <0.001
S10SumOfSqs 20511883 [1889588; 88044364] 21913762 [1432874; 71523314] 0.820
S10SumEntrp 12912529 [1436964;14652918] 13511286 [0.99; 16572731] 0.330
S01Contrast 29791667 [15625; 72708333] 22395833 [0.94; 73854167] 0.171
S01Correlat 0.87 [0.72 ;0.91] 0.81 [0.64; 0.93] 0.071
S01SumOfSqs 20970459 [208294; 88057454] 21399468 [151301;71484104] 0.820
S01SumEntrp 13104067 [1273463;14494688] 13139168 [0.98; 15742717] 0.494
S11Contrast 45666667 [4.00; 83444444] 18688889 [3.10; 90666667] 0.310
S11Correlat 0.72 [0.41; 0.85] 0.83 [0.59; 0.94] 0.002
S11SumOfSqs 23431451 [1782034; 90758025] 39192469 [1136284; 71902469] 0.917
S11Entropy 18378762 [1667239; 20110383] 17148857 [1592995; 21917217] 0.885
S1m1Contrast 30777778 [5.20; 96111111] 31688889 [0.83; 89777778] 0.852
S1m1Correlat 0.70 [0.31; 0.83] 0.72 [0.26; 0.82] 0.950
S1m1SumOfSqs 20729969 [2132034; 86888889] 22217284 [66275; 73521914] 0.694
S20Contrast 20037037 [1345679; 72222222] 18345679 [1437037; 63209877] 0.787
S20SumOfSqs 21879134 [1324036; 88148148] 22961439 [689148; 72130316] 0.633
S02Contrast 23505882 [4.60; 97176471] 41764706 [119; 92788235] 0.950
S02SumOfSqs 22071315 [1380955; 92804498] 19391349 [4776692; 72549827] 0.373
S22Contrast 17383562 [2209589; 35424658] 49109589 [1160274; 88219178] 0.001
S22SumOfSqs 22325389 [2064271; 95241603] 16858322 [2383374;72957262] 0.330
S2m2Contrast 16479452 [1490411; 62465753] 18219178 [22.0; 78835616] 0.663
S2m2SumOfSqs 19790814 [2032356; 68221055] 38733768 [11315538; 99713361] 0.093
S30Contrast 33264706 [10529412; 90441176] 32647059 [8.50; 85441176] 0.724
S30SumOfSqs 21988754 [12547362; 93122837] 35467939 [6642247; 98234753] 0.983
S03Contrast 17824324 [1177027; 78648649] 28783784 [10175676; 99351351] 0.101
S03SumEntrp 11126686 [136782; 14039459] 11764806 [0.82;14836123] 0.852
S33Contrast 23431034 [14017241; 53655172] 15931034 [4.50; 86948276] 0.229
S33SumOfSqs 20416394 [1602341; 69693074] 36524078 [4785003; 99912827] 0.237
S33SumEntrp 11951044 [0.82;13617913] 12976173 [0.78; 14739491] 0.110
S3m3Contrast 19482759 [26.0; 99137931] 24310345 [277; 95275862] 0.917
S3m3SumOfSqs 22726219 [1978478; 91964923] 25079518 [119919; 68456822] 0.663
7
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
Tables III-IV present the descriptive measure-
ments and the comparison results regarding the
left side. In Table III, one can observe the median,
minimum and maximum values, including
P
-value
in the comparison between the groups on the left
side, whereas in Table IV one can observe the
mean value and standard deviation of the texture
parameters by group.
In Table III, one can observe that the migraine
group presented more disc displacements (
P
-value
= 0.002; Fisher’s exact test) and more changes in
disc function (non-reduced discs) (
P
-value = 0.014;
Fisher’s exact test). With regard to TA, the migraine
group showed higher values for the parameters
S10Correlat, S10SumOfSqs, S01SumOfSqs,
S11Correlat, S02SumOfSqs, S30SumOfSqs and
S33Contrast as well as lower values for parameters
S01Correlat and S1m1Correlat.
Figure 3 shows the texture parameters which
presented statistically significant differences
between control and migraine groups on the left
side.
Table II – Mean and standard deviation of TA by group on the right side
Right Side
Parameter
Control (N=15) Migraine (N=15)
Mean (S.D.) Mean (S.D.)
S10AngScMom 0.02 (0.01) 0.03 (0.03)
S10Contrast 41104965 (26184884) 20439716 (17541999)
S10Correlat 0.81 (0.08) 0.92 (0.05)
S10SumOfSqs 33093818 (27563737) 32041355 (24999555)
S10SumEntrp 12462408 (3199419) 12401667 (5101838)
S01Contrast 29381514 (19830559) 22921703 (22523574)
S01Correlat 0.85 (0.06) 0.80 (0.08)
S01SumOfSqs 31391401 (27722097) 30805948 (25541016)
S01SumEntrp 10932826 (5026899) 11341142 (5607201)
S11Contrast 38031114 (31115973) 24497778 (21753120)
S11Correlat 0.68 (0.12) 0.81 (0.09)
S11SumOfSqs 35405483 (28279072) 35592885 (26475587)
S11Entropy 15343282 (7106420) 14870546 (7300256)
S1m1Contrast 38055557 (32444607) 36441482 (25002615)
S1m1Correlat 0.68 (0.12) 0.66 (0.16)
S1m1SumOfSqs 35598426 (26650193) 32915157 (23534763)
S20Contrast 24668066 (22584783) 22653416 (19304385)
S20SumOfSqs 35624008 (25997660) 32428396 (22565161)
S02Contrast 45674510 (36826003) 44759302 (30046573)
S02SumOfSqs 36406723 (27091603) 30146047 (23401706)
S22Contrast 17124475 (9586326) 47125571 (26058604)
S22SumOfSqs 35999944 (27651068) 29317819 (25151435)
S2m2Contrast 24704110 (17803459) 29171965 (24763984)
S2m2SumOfSqs 26364179 (21962415) 43401414 (27077348)
S30Contrast 37634314 (24973893) 36562748 (28643558)
S30SumOfSqs 35774917 (25005521) 38199236 (26858117)
S03Contrast 24928739 (21587368) 40134865 (30893439)
S03SumEntrp 9276817 (5316520) 9014256 (6181636)
S33Contrast 26351724 (11642693) 28031059 (30413793)
S33SumOfSqs 31207655 (23581784) 44895214 (30620045)
S33SumEntrp 9666962 (4866084) 11129134 (4721075)
S3m3Contrast 31873105 (27893633) 33185421 (28173496)
S3m3SumOfSqs 35730319 (28691724) 32781770 (23112979)
8
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
Table III – Comparison between groups on the left side by using Fisher’s exact test or Mann Whitney’s test
Left Side
Parameter
Control (N=15) Migraine (N=15)
P-value
Median [Min; Max] Median [Min; Max]
Disc Position 0.002
Displaced 1 (6.67%) 10 (66.7%)
Normal 14 (93.3%) 5 (33.3%)
Disc Function 0.014
Non-reduced 1 (6.67%) 8 (53.3%)
Reduced 14 (93.3%) 7 (46.7%)
S10AngScMom 0.01 [0.01; 0.02] 0.01 [0.01; 0.08] 0.384
S10Contrast 51276596 [2306383; 84148936] 38510638 [0.93; 62234043] 0.071
S10Correlat 0.82 [0.63; 0.91] 0.93 [0.70; 0.97] <0.001
S10SumOfSqs 15056926 [2569794; 95330325] 40064396 [3164472; 77787772] 0.040
S10SumEntrp 14090097 [1339538;16960223] 15088759 [155121;16127587] 0.373
S01Contrast 31354167 [29375; 87916667] 26052083 [1653125;79895833] 0.576
S01Correlat 0.90 [0.75; 0.96] 0.81 [0.59; 0.89] <0.001
S01SumOfSqs 18615207 [1392749; 63152751] 41903402 [13801649; 75208333] 0.003
S01SumEntrp 14114043 [1317346;16434899] 14846802 [1588665;16008947] 0.633
S11Contrast 16266667 [9.60; 93444444] 19444444 [21.7; 97222222] 0.885
S11Correlat 0.69 [0.57; 0.83] 0.81 [0.47; 0.93] 0.024
S11SumOfSqs 18923951 [1482358; 93934414] 30792469 [53475; 77865556] 0.152
S11Entropy 19516522 [1887702; 21448948] 20182137 [2056646; 21139066] 0.576
S1m1Contrast 15111111 [17.8; 97777778] 26155556 [28.2; 50811111] 0.820
S1m1Correlat 0.75 [0.50; 0.90] 0.68 [0.46; 0.82] 0.029
S1m1SumOfSqs 17829506 [133275; 99612191] 33312222 [478475; 74592469] 0.101
S20Contrast 24481481 [70.0; 73358025] 17074074 [11481481;50518519] 0.648
S20SumOfSqs 17993484 [1108642; 91769852] 32682061 [4261656; 78595679] 0.085
S02Contrast 18858824 [37.2; 98235294] 60470588 [54.0; 97294118] 0.191
S02SumOfSqs 18021488 [10088249; 59681696] 36974256 [5093564; 74844291] 0.024
S22Contrast 28082192 [2209589; 80589041] 43630137 [4630137; 90136986] 0.101
S22SumOfSqs 23234378 [1921322; 99566523] 38844248 [4343465;78515669] 0.254
S2m2Contrast 23547945 [2930137; 42465753] 27808219 [10027397; 88890411] 0.191
S2m2SumOfSqs 16751783 [4232933; 98084444] 36503847 [2686808; 70033824] 0.290
S30Contrast 41558824 [19.5; 77191176] 33514706 [18.0; 87382353] 0.950
S30SumOfSqs 17217939 [1133045; 89633164] 38455666 [13374946; 78198962] 0.012
S03Contrast 22824324 [1377027; 59324324] 20054054 [10294595; 93972973] 0.576
S03SumEntrp 13000026 [1346483; 15843146] 12934547 [134533; 14577216] 0.263
S33Contrast 35586207 [86.5; 94741379] 65534483 [12262069; 96862069] 0.007
S33SumOfSqs 22688689 [1001843; 93783294] 31782625 [3950379; 77980826] 0.221
S33SumEntrp 12343825 [1213839; 14882474] 12521779 [1378634; 15011721] 0.885
S3m3Contrast 40672414 [3112069; 87431034] 21965517 [10518966; 85655172] 0.373
S3m3SumOfSqs 17316513 [141875; 92875372] 27691439 [2692063; 73344233] 0.494
9
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
For analysis of the relationship of the disc
position and function with texture parameters,
logistic regression models were used for 60 results
(two sides
per
patient) in which the comparison
was adjusted for side and group. Tables V-VI
present the descriptive measurements of the
texture parameters by disc position, including
P
-value of the comparison between the positions.
It should be observed that this
P
-value is adjusted
for side and group only, meaning that the
inuence of the other two factors on the disc
position is not considered.
In Table V, one can observe that individuals
with normal disc position had higher values
of entropy for parameters S11Entropy and
S03SumEntropy and lower value of contrast for
S02Contrast.
Figure 4 shows the texture parameters
which present statistically signicant differences
between the disco positions.
Tables VII-VIII present the descriptive mea-
surements of texture parameters by disc function
and the comparison between the disc functions.
No statistically signicant differences were found
between disc functions in relation to the texture
parameters.
Table IV – Mean and standard deviation of TA by group on the left side
Left Side
Parameter
controlee (N=15) Migraine (N=15)
Mean (S.D.) Mean (S.D.)
S10AngScMom 0.01 (0.00) 0.02 (0.02)
S10Contrast 44301986 (30346039) 28641844 (21966789)
S10Correlat 0.80 (0.08) 0.91 (0.07)
S10SumOfSqs 23528580 (25054501) 39057451 (23520290)
S10SumEntrp 12838915 (4691008) 12867007 (5123403)
S01Contrast 35026681 (29889545) 26651180 (21880585)
S01Correlat 0.89 (0.05) 0.77 (0.09)
S01SumOfSqs 23689571 (15755646) 43688475 (18671290)
S01SumEntrp 12954637 (4798048) 13721724 (3585310)
S11Contrast 33607410 (33125708) 29422964 (27066075)
S11Correlat 0.69 (0.09) 0.78 (0.13)
S11SumOfSqs 28517725 (23414819) 37532854 (24253372)
S11Entropy 16395932 (7490512) 18373084 (4852539)
S1m1Contrast 36976298 (37124643) 24531857 (15028628)
S1m1Correlat 0.74 (0.12) 0.64 (0.11)
S1m1SumOfSqs 26295186 (26423731) 36012836 (21369717)
S20Contrast 24763544 (18766159) 22433745 (10642390)
S20SumOfSqs 26052119 (23656425) 37283380 (22684207)
S02Contrast 37362358 (35096837) 53840004 (30263461)
S02SumOfSqs 23928806 (14270107) 39535013 (21095887)
S22Contrast 32985936 (23918170) 47646575 (25096768)
S22SumOfSqs 34095393 (29117517) 39962861 (22251140)
S2m2Contrast 22687945 (12863432) 37212603 (27577898)
S2m2SumOfSqs 26913427 (24194430) 34259099 (21979002)
S30Contrast 37177259 (23566443) 38783336 (23032607)
S30SumOfSqs 24807160 (23814403) 42007872 (17727533)
S03Contrast 25613423 (18287854) 36822252 (31554449)
S03SumEntrp 12762818 (3341505) 10593494 (5141356)
S33Contrast 33760236 (24282762) 60477241 (26216645)
S33SumOfSqs 29657138 (23455764) 38063010 (23163898)
S33SumEntrp 11087894 (4154230) 10555907 (4902189)
S3m3Contrast 39260345 (25030948) 33853218 (26607776)
S3m3SumOfSqs 25950167 (23550052) 31729412 (24013232)
10
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
Figure 3 – Boxplots of the texture parameters which presented statistically significant difference between the groups on the left side.
11
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
Figure 4 – Boxplots of the texture parameters which presented statistically significant difference between disc positions.
Table V – Comparison between the disc positions by using logistic regression analysis adjusted for side and group
Position of the Disc
Parameter
Displaced (N=23) Normal (N=37)
P-value*
Median [Min; Max] Median [Min; Max]
S10AngScMom 0.02 [0.01; 0.10] 0.02 [0.01; 0.08] 0.842
S10Contrast 20319149 [0.53; 62234043] 32021277 [0.93; 84148936] 0.804
S10Correlat 0.92 [0.70; 0.98] 0.86 [0.63; 0.98] 0.527
S10SumOfSqs 40064396 [1432874; 88044364] 19521277 [1889588; 95330325] 0.345
S10SumEntrp 14090097 [0.99; 16572731] 13965937 [1339538; 16960223] 0.320
S01Contrast 22395833 [0.94; 79895833] 28958333 [6.50; 87916667] 0.503
S01Correlat 0.80 [0.59; 0.89] 0.87 [0.64; 0.96] 0.281
S01SumOfSqs 38998671 [151301; 88057454] 20970459 [208294; 75208333] 0.286
S01SumEntrp 14262555 [0.98; 16008947] 13824142 [1273463; 16434899] 0.857
S11Contrast 14444444 [3.10; 82444444] 28444444 [7.10; 97222222] 0.100
S11Correlat 0.77 [0.47; 0.94] 0.74 [0.41; 0.92] 0.508
S11SumOfSqs 39192469 [1136284; 90758025] 22869722 [53475; 93934414] 0.191
S11Entropy 18288017 [1592995; 21917217] 19467541 [1667239; 21448948] 0.025
S1m1Contrast 27977778 [0.83; 76888889] 19744444 [5.40; 97777778] 0.715
S1m1Correlat 0.68 [0.44; 0.82] 0.73 [0.26; 0.90] 0.362
S1m1SumOfSqs 31994321 [66275;,86888889] 20729969 [133275; 99612191] 0.777
S20Contrast 17012346 [1345679;,63209877] 21802469 [70.0; 73358025] 0.293
S20SumOfSqs 31495542 [689148; 88148148] 21879134 [1108642; 91769852] 0.746
S02Contrast 65764706 [10129412; 97294118] 22317647 [4.60; 98235294] 0.041
S02SumOfSqs 36974256 [6843218; 92804498] 20334394 [1380955; 74844291] 0.072
S22Contrast 42054795 [4630137; 88219178] 20849315 [1160274; 90136986] 0.650
S22SumOfSqs 28808032 [2383374; 95241603] 23234378 [1921322; 99566523] 0.489
S2m2Contrast 18219178 [10078082; 88890411] 23479452 [22.0;.62465753] 0.339
S2m2SumOfSqs 36503847 [2781291; 99713361] 19790814 [2032356; 98084444] 0.686
S30Contrast 34088235 [11426471; 87382353] 33264706 [8.50; 90441176] 0.054
S30SumOfSqs 36507299 [6642247; 98234753] 19791306 [1133045; 89633164] 0.220
S03Contrast 16768919 [1177027; 99351351] 23513514 [1377027; 93918919] 0.463
S03SumEntrp 11764806 [0.82; 14758456] 12873781 [136782; 15843146] 0.032
S33Contrast 50862069 [4.50; 96862069] 25689655 [19.5; 94741379] 0.139
S33SumOfSqs 31782625 [3950379;99912827] 23331971 [1001843; 93783294] 0.847
S33SumEntrp 12012297 [0.78; 15011721] 12343825 [0.82; 14882474] 0.737
S3m3Contrast 19482759 [26.0; 86551724] 32293103 [1912069; 99137931] 0.186
S3m3SumOfSqs 27691439 [2692063; 91964923] 21422934 [119919; 92875372] 0.477
*Adjusted for side and group.
12
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
DISCUSSION
Our results indicated that individuals with MH
had a higher incidence of changes in the position of
articular discs, that is, displaced ones with mouth
closed (
P
= 0.009 and
P
= 0.002 for the right
and left sides, respectively) in comparison with
controls (without MH). In addition, the majority
of the discs showed no reduction with maximum
mouth opening in these same individuals in view
of the changes in disc function (
P
= 0.007 and
P
= 0.14 for the right and left sides, respectively).
This finding did not surprise us, since it is in
accordance with other studies [9,10] indicating
that MH can be an indicator of temporomandibular
dysfunctions related to both masticatory muscles
and internal derangements of the articular disc.
The issue would involve the assessment of texture
parameters of the discs and correlate them to the
presence of alterations in patients with MH based
on the null hypothesis that articular discs in these
individuals would present some differences in the
texture values. This would allow us to correlate
these parameters to disc displacement and/or
Table VI – Mean and standard deviation of TA by disc position (
Source
: own authorship)
Position of the Disc
Parameter
Displaced (N=23) Normal (N=37)
Mean (S.D.) Mean (S.D.)
S10AngScMom 0.03 (0.02) 0.02 (0.01)
S10Contrast 27486587 (21679002) 37436113 (27558505)
S10Correlat 0.89 (0.07) 0.84 (0.09)
S10SumOfSqs 37633606 (26677601) 28385003 (24089053)
S10SumEntrp 12005890 (5537771) 13038230 (3712523)
S01Contrast 23669408 (20608351) 31495130 (25152009)
S01Correlat 0.79 (0.09) 0.85 (0.07)
S01SumOfSqs 39010060 (25254416) 28281069 (20874603)
S01SumEntrp 12515034 (4799391) 12065112 (4901604)
S11Contrast 22028021 (21518303) 37209311 (30663151)
S11Correlat 0.77 (0.11) 0.72 (0.12)
S11SumOfSqs 40350438 (26592918) 30477680 (23968210)
S11Entropy 14408895 (7974506) 17387515 (5657410)
S1m1Contrast 30397105 (20217078) 36241743 (32610271)
S1m1Correlat 0.64 (0.12) 0.70 (0.14)
S1m1SumOfSqs 34885873 (25298199) 31349973 (23871936)
S20Contrast 20483575 (15837016) 25585387 (19045593)
S20SumOfSqs 35297462 (25577973) 31323701 (22434287)
S02Contrast 57069616 (27339067) 38160579 (34247000)
S02SumOfSqs 39614148 (24147140) 28084417 (20046357)
S22Contrast 45757475 (24536921) 30292336 (23776524)
S22SumOfSqs 36798750 (26578718) 33628892 (25580648)
S2m2Contrast 34798273 (27397168) 24494299 (16529478)
S2m2SumOfSqs 38637101 (25726855) 29065364 (22957490)
S30Contrast 44153453 (22705109) 33427985 (24992735)
S30SumOfSqs 42574963 (25676031) 30611179 (21803316)
S03Contrast 34104348 (31481543) 30488897 (23106472)
S03SumEntrp 8342337 (6223982) 11698298 (4003925)
S33Contrast 47025263 (31103351) 31019537 (23057666)
S33SumOfSqs 40799595 (29025107) 32944719 (22922210)
S33SumEntrp 10575789 (4859843) 10631225 (4480825)
S3m3Contrast 29030298 (25652487) 37969851 (26642413)
S3m3SumOfSqs 34400481 (25551685) 29774701 (24112013)
13
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
changes in the disc function (i.e. non-reduction),
thus working as an auxiliary means to predict
these ndings.
No study on texture analysis of articular
discs of TMJ in individuals with MH was found in
the literature. We found a study [22] evaluating
the lateral pterygoid muscle in TMJs of patients
with TMD by using TA, which was shown to be
a very promising technique for identication of
parameters indicating changes in these muscles
of individuals who have TMD.
Firstly, we analysed the values of texture
parameters which were statistically different
between control between controls and individuals
with MH before evaluating the disc derangements.
Among all the seven texture parameters
addressed in the present study, the results
showed that there were statistically signicant
differences (
P
< 0.05) between individuals
with MH and controls regarding the parameters
Contrast for both right and left sides, with the
former presenting higher values of contrast.
Table VII – Comparison between the disc positions by using logistic regression analysis adjusted for side and group
Function of the Disc
Parameter
Non-Reduced (N=19) Reduced (N=41)
P-value*
Median [Min; Max] Median [Min; Max]
S10AngScMom 0.02 [0.01; 0.10] 0.02 [0.01; 0.08] 0.711
S10Contrast 20319149 [0.53; 62234043] 32021277 [0.93; 84148936] 0.836
S10Correlat 0.92 [0.70; 0.98] 0.86 [0.63; 0.98] 0.239
S10SumOfSqs 40397097 [1432874; 88044364] 19768758 [1889588; 95330325] 0.433
S10SumEntrp 14090097 [0.99;.16572731] 13965937 [1339538;1 6960223] 0.177
S01Contrast 22395833 [0.94;.79895833] 28958333 [6.50; 87916667] 0.707
S01Correlat 0.81 [0.66; 0.89] 0.86 [0.59; 0.96] 0.596
S01SumOfSqs 30378228 [151301; 88057454] 23521457 [208294; 75208333] 0.956
S01SumEntrp 14262555 [0.98; 16008947] 13824142 [1273463; 16434899] 0.815
S11Contrast 14444444 [3.10; 82444444] 25566667 [4.00; 97222222] 0.319
S11Correlat 0.78 [0.59; 0.94] 0.73 [0.41; 0.92] 0.708
S11SumOfSqs 39192469 [1136284; 90758025] 23431451 [53475;.93934414] 0.389
S11Entropy 18288017 [1592995; 21917217] 19467541 [1667239; 21448948] 0.171
S1m1Contrast 26333333 [0.83; 76888889] 21122222 [5.40; 97777778] 0.441
S1m1Correlat 0.69 [0.44; 0.82] 0.72 [0.26; 0.90] 0.732
S1m1SumOfSqs 25708025 [478475; 86888889] 23618858 [66275; 99612191] 0.845
S20Contrast 17074074 [1345679; 63209877] 20691358 [70.0; 73358025] 0.374
S20SumOfSqs 22961439 [689148; 88148148] 24086725 [1108642; 91769852] 0.530
S02Contrast 60470588 [10129412; 97294118] 23505882 [4.60; 98235294] 0.463
S02SumOfSqs 37964187 [10332318; 92804498] 21896194 [1380955; 74844291] 0.088
S22Contrast 40958904 [4630137; 88219178] 21890411 [1160274; 90136986] 0.711
S22SumOfSqs 25537249 [2383374; 95241603] 25789454 [1921322; 99566523] 0.902
S2m2Contrast 23547945 [10078082; 88890411] 21945205 [22.0; 68273973] 0.225
S2m2SumOfSqs 30633515 [2781291; 99713361] 21334772 [2032356; 98084444] 0.710
S30Contrast 34088235 [11426471; 87382353] 33264706 [8.50; 90441176] 0.340
S30SumOfSqs 36507299 [11501298; 98234753] 21988754 [1133045;.89633164] 0.166
S03Contrast 24581081 [1177027; 99351351] 22824324 [1377027; 93972973] 0.521
S03SumEntrp 11764806 [0.82;14758456] 12667773 [136782; 15843146] 0.053
S33Contrast 50862069 [4.50; 96862069] 25689655 [19.5; 94741379] 0.235
S33SumOfSqs 29330187 [3950379; 99912827] 24234468 [1001843; 93783294] 0.949
S33SumEntrp 12012297 [0.78; 15011721] 12343825 [0.82;14882474] 0.502
S3m3Contrast 21965517 [26.0; 86551724] 29603448 [1912069; 99137931] 0.204
S3m3SumOfSqs 27691439 [6872176; 91964923] 21422934 [119919; 92875372] 0.636
*Adjusted for side and group.
14
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
A high value of contrast can be interpreted as
being the presence of differentiated regions on the
image of articular disc, that is, low-signal areas
(hyposignal) interspersed among high-signal
ones (hypersignal) without a grey-scale gradation
which might reduce the contrast (intermediate
signals of radiofrequency). Considering that
T2-weighted MRI scans were used for calculation
of texture parameters of the articulacy discs, we
can infer that high-contrast images as observed
by using the TA may be due to the presence of
structural degenerative processes in the articular
discs of the individuals with MH. This would lead
to deterioration of the normal brocartilaginous
content, with possible presence of internal
hydropic regions, resulting in hypersignal images
as liquids have high signals on T2-weighted MRI
scans. Interestingly, no region with altered signals
was visually found in the image analyses of the
articular discs, that is, subjectively. This makes
our ndings more relevant as they corroborate
the hypothesis that the TA can be a mathematic
Table VIII – Mean and standard deviation of TA by disc function
Function of the Disc
Parameter
Non-Reduced (N=19) Reduced (N=41)
Mean (S.D.) Mean (S.D.)
S10AngScMom 0.03 (0.03) 0.02 (0.01)
S10Contrast 26963046 (22915912) 36708044 (26645399)
S10Correlat 0.89 (0.08) 0.85 (0.09)
S10SumOfSqs 37728092 (27579161) 29243520 (24049436)
S10SumEntrp 11569397 (5990785) 13139791 (3563493)
S01Contrast 24136377 (22003432) 30515244 (24363674)
S01Correlat 0.80 (0.07) 0.84 (0.09)
S01SumOfSqs 35801832 (25305987) 30814539 (22079911)
S01SumEntrp 12237826 (5229157) 12237469 (4696687)
S11Contrast 23678363 (22422002) 34963417 (30249391)
S11Correlat 0.79 (0.10) 0.72 (0.12)
S11SumOfSqs 39498146 (25996887) 31835840 (24846656)
S11Entropy 15062162 (7392286) 16794185 (6431425)
S1m1Contrast 28583045 (21623480) 36512196 (31035225)
S1m1Correlat 0.65 (0.13) 0.69 (0.13)
S1m1SumOfSqs 34786664 (25584359) 31740913 (23911341)
S20Contrast 20456075 (15915583) 25100393 (18778918)
S20SumOfSqs 32115929 (26271904) 33185753 (22526148)
S02Contrast 51935046 (27053007) 42384798 (35135353)
S02SumOfSqs 40377609 (24407669) 28855470 (20453156)
S22Contrast 43547080 (23324898) 32825459 (25349461)
S22SumOfSqs 33997994 (27452211) 35236058 (25325930)
S2m2Contrast 36705191 (27582132) 24615871 (17509827)
S2m2SumOfSqs 36333751 (26785819) 31066598 (23216076)
S30Contrast 41412539 (20005099) 35744552 (26381443)
S30SumOfSqs 44117101 (26532841) 31063728 (21658449)
S03Contrast 34236913 (30655525) 30780191 (24565835)
S03SumEntrp 8171181 (6504172) 11450203 (4147384)
S33Contrast 47093013 (30801741) 32549675 (24613418)
S33SumOfSqs 40727816 (29798015) 33744311 (23296334)
S33SumEntrp 10299174 (5295333) 10754004 (4286682)
S3m3Contrast 28224608 (24331018) 37471068 (27109970)
S3m3SumOfSqs 33954923 (23416866) 30432475 (25281251)
15
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
method allowing the identification of subtle
alterations in the structures and which are not
evident for a radiologist by means of a simple
inspection. It is also important to highlight that,
according to the TA, the parameter Contrast does
not consist purely of an image histogram, but of
statistical calculations for comparing pixel values
of adjacent distances, thus allowing the structure
to be evaluated as a whole.
A statistical and objective approach for
identication of a subtle alteration can be very
useful to make a decision regarding the treatment
planning and its successful outcome.
A second parameter addressed here was the
parameter Correlat (correlation), which was found
to have a statistically signicant difference (
P
<
0.05) between both groups (controls x individuals
with MH). It was observed that individuals with
MH had lower values for both right and left sides
of the TMJ compared to controls.
In the TA, the parameter Correlat indicates
the presence of a grey-scale homogeneity in the
image, which is related to the spatial organisation
pattern of the tissues [20]. High values of
correlation indicate that images are homogeneous,
with aligned tissues presenting regular pixel
values and uniform distribution patterns without
discrepancies. In our study, individuals with MH
had lower values of correlation, meaning that the
images of their articular disc had a heterogeneous
resolution.
Based on the principle that articular discs
of TMJ present anterior, intermediate, posterior
bands in their normal anatomical aspects
in T2-weighted MRI scans with hyposignal
intensity, and that this occurs mainly by the
linear orientation of the collagen bres (mainly
type I collagen) composing the disc, a lack of
homogeneity expressed by the low value of
correlation could indicate a disorganization of
spatial pattern of these bres, in addition to areas
of hydropic degeneration, as cited before.
The values of the parameters Contrast and
Correlat obtained from TA provide information on
the composition and structure of the tissue being
analysed, respectively, and which are inversely
proportional to their values. These ndings can
be considered as a more specic form of image
analysis of the articular discs in MRI scans.
It was found that the parameter SumOfSqs
(sum of squares), on the left side, also showed
a statistically significant difference between
the groups, with individuals with MH having
higher values. According to the TA, the
parameter SumOfSqs is related to the variation
in the background shades, that is, it is used for
differentiating the different anatomical structures
within the segmented ROI [21,23]. Because in our
study only the articular disc was segmented, this
nding is not signicant enough as it occurred
only on one side. This parameter could indicate
that images close to the disc, such as the bilinear
zone, whose delimitation with the posterior band
is difcult to determine, occasionally could have
been included in the ROI. However, in this case,
the TA process would exclude these regions
from the analyses of other parameters due to the
divergent high values found.
Once the behaviours of the texture
parameters of the discs have been analysed in
the MRI scans, including their meaning, we can
discuss the ndings on their relationships with
the position and function of the articular discs,
which is the main objective of this study.
With regard to the position of the articular
discs (i.e. normal or displaced), our results
indicated that there were statistically signicant
differences between the parameters Entrp
(entropy) (
P
= 0.025), SumEntrp (sum of
entropy) (
P
= 0.032) and Contrast (contrast)
(P = 0.041) for those individuals presenting
normal disc position. Also, these individuals have
higher values of Entrp and SumEntrp as well as
lower values of Contrast compared to individuals
with displaced disc.
In general, the parameters Entrp and
SumEntrp have a very similar behaviour in
the texture analysis, in which the disarray of
pixels in the image is analysed rst and then the
structure of the ROI as a whole [21,24]. A high
value of entropy indicates more grey levels in the
image, which corresponds to a low contrast [25].
Conversely, when the values of entropy are low,
we observed less grey levels and more contrast in
the image. These ndings are consistent to those
previously observed for contrast.
If one considers that normal disc position
(high values of Entrp and SumEntrp), and
therefore low contrast, was mostly observed in
individuals without MH (controls) compared to
those with MH, and therefore high contrast, low
values of Entrp and SumEntrp should actually
indicate the presence of disc displacement
16
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
as these parameters behave conversely, as
demonstrated by the results for disc position.
These results signal that high values of
contrast and low values of entropy (and sum
of entropy) can indicate disc displacement in
individuals with MH. This was one of our main
objectives in the present study.
With regard to the disc function (i.e. reduced
or non-reduced), our results indicated that
there were no statistically signicant differences
between the texture parameters and the presence
or not of functional reduction of the articular disc.
However, by analysing the tendency of each
one of the parameters in relation to the disc
function, it was observed that SumEntrp was the
only parameter showing a statistical difference
between reduced and non-reduced functions
(S03) (
P
= 0.053), with the former presenting
higher mean values (12667773) compared to the
latter (11764806).
This nding is consistent with what has been
addressed regarding the position of the articular
discs, suggesting that high values of SumEntrp
tend to be related to a reduction function
antagonistically, whereas low values would be
related to a non-reduced function, which was
observed in the majority of the individuals with
MH. It should be emphasised again that high
values of entropy (and sum of entropy) are
related to many grey levels, and thus, to a low
contrast, whereas low values of entropy are
related to high contrast. Once again our results
are consistent, since individuals with MH and
non-reduced disc (high contrast) should present
low entropy, which corroborates the nding on
the sum of entropy for reduced disc function.
Nevertheless, some limitations of the present
study should be considered, such as the sample size
(15 individuals in each group), although previous
studies on MRI texture analysis of migraine
headache used samples similar to ours and upon
which we based our sampling calculation.
We consider that our results are very
promising, as they indicate that the MRI texture
analysis of articular discs in individuals with MH
has the potential to determine the behaviour of
the discs in terms of displacement and reduction
mainly on the basis of the values of contrast and
entropy. Moreover, as this technique is based
on contrast and correlation, it can be used as
an auxiliary means for identication of subtle
alterations in the composition and structure of
articular discs not identified visually in these
individuals.
CONCLUSION
The TA of articular discs in individuals with
migraine headache has the potential to determine
the behaviour of disc derangements based on
high values of contrast and low values of entropy,
showing that the correlation of these texture
parameters can correspond to displacements
and a tendency for non-reduction of the discs in
these individuals.
Author’s Contributions
KACF, CMO, ALFC, TMAMR, ECCBA:
Software, Formal analysis, Validation,
Investigation, Data Curation, Writing, Review &
Editing. SLPCF: Conceptualization, Methodology,
Writing, Original Draft Preparation, Supervision,
Project administration.
Conict of Interest
No conicts of interest declared concerning
the publication of this article.
Funding
This study was supported by the São Paulo
Research Foun- dation FAPESP (grant number
2019/22315-0).
Regulatory Statement
This study was conducted in accordance
with all the provisions of the local human
subjects oversight committee guidelines and
policies of: the Institutional Review Board of
UNESP. The approval code for this study is:
32339720.8.0000.0077. Informed consent:
Written informed consent was obtained from
each participant, after informed about the study.
REFERENCES
1. Costa AL, Yasuda CL, Appenzeller S, Lopes SL, Cendes F.
Comparison of conventional MRI and 3D reconstruction model
for evaluation of temporomandibular joint. Surg Radiol Anat.
2008;30(8):663-7. http://dx.doi.org/10.1007/s00276-008-
0400-z. PMid:18704257.
17
Braz Dent Sci 2023 Jan/Mar;26 (1): e3649
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the temporomandibular joint for changes in the articular disc in individuals with migraine headache
Fardim KAC et al.
Magnetic resonance imaging texture analysis of the
temporomandibular joint for changes in the articular disc in
individuals with migraine headache
2. Isberg A. Temporomandibular joint dysfunction : a
practitioner’s guide. London: CRC Press; 2001. http://dx.doi.
org/10.1201/9780203633151.
3. Oliveira SSI, Gonçalves SLM, Weig KM, Magalhães TR Fo,
Martinez OER, Kalil MTAC,etal. Temporomandibular disorders:
guidelines and sef-care for patients during Covid-19 pandemic.
Braz Dent Sci. 2020;23(2):1-8. http://dx.doi.org/10.14295/
bds.2020.v23i2.2255.
4. Ton LAB, Mota IG, Paula JS, Martins APVB. Prevalece of
temporomandibular disorder and its association with stress and
anxiety among university students. Braz Dent Sci. 2020;23(1):1-9.
http://dx.doi.org/10.14295/bds.2020.v23i1.1810.
5. Contreras EFR, Fernandes G, Ongaro PCJ, Campi LB, Gonçalves
DAG. Systemic diseases and other painful conditions in patients
with temporomandibular disorders and migraine. Braz Oral Res.
2018;32:e77. http://dx.doi.org/10.1590/1807-3107bor-2018.
vol32.0077. PMid:30043839.
6. Costa AL, D’Abreu A, Cendes F. Temporomandibular joint internal
derangement: association with headache, joint effusion, bruxism,
and joint pain. J Contemp Dent Pract. 2008;9(6):9-16. http://
dx.doi.org/10.5005/jcdp-9-6-9. PMid:18784854.
7. Anderson GC, John MT, Ohrbach R, Nixdorf DR, Schiffman
EL, Truelove ES, et al. Influence of headache frequency on
clinical signs and symptoms of TMD in subjects with temple
headache and TMD pain. Pain. 2011;152(4):765-71. http://dx.doi.
org/10.1016/j.pain.2010.11.007. PMid:21196079.
8. Franco AL, Gonçalves DA, Castanharo SM, Speciali JG, Bigal ME,
Camparis CM. Migraine is the most prevalent primary headache
in individuals with temporomandibular disorders. J Orofac Pain.
2010;24(3):287-92. PMid:20664830.
9. Paolo C, D’Urso A, Papi P, Sabato F, Rosella D, Pompa G,etal.
Temporomandibular disorders and headache: a retrospective
analysis of 1198 patients. Pain Res Manag. 2017;2017:3203027.
http://dx.doi.org/10.1155/2017/3203027. PMid:28420942.
10. Lopes SLPC, Costa ALF, Gamba TO, Flores IL, Cruz AD, Min LL.
Lateral pterygoid muscle volume and migraine in patients with
temporomandibular disorders. Imaging Sci Dent. 2015;45(1):1-5.
http://dx.doi.org/10.5624/isd.2015.45.1.1. PMid:25793177.
11. Gonçalves DAG, Bigal ME, Jales LCF, Camparis CM, Speciali
JG. Headache and symptoms of temporomandibular disorder:
an epidemiological study. Headache. 2010;50(2):231-41. http://
dx.doi.org/10.1111/j.1526-4610.2009.01511.x. PMid:19751369.
12. Ballegaard V, Thede-Schmidt-Hansen P, Svensson P, Jensen R.
Are headache and temporomandibular disorders related? A
blinded study. Cephalalgia. 2008;28(8):832-41. http://dx.doi.
org/10.1111/j.1468-2982.2008.01597.x. PMid:18498400.
13. Larheim TA. Role of magnetic resonance imaging in the clinical
diagnosis of the temporomandibular joint. Cells Tissues Organs.
2005;180(1):6-21. http://dx.doi.org/10.1159/000086194.
PMid:16088129.
14. Sano T. Recent developments in understanding temporomandibular
joint disorders. Part 1: bone marrow abnormalities of the
mandibular condyle. Dentomaxillofac Radiol. 2000;29(1):7-10.
http://dx.doi.org/10.1038/sj.dmfr.4600492. PMid:10654030.
15. Lubner MG, Smith AD, Sandrasegaran K, Sahani DV, Pickhardt PJ.
CT texture analysis: definitions, applications, biologic correlates,
and challenges. Radiographics. 2017;37(5):1483-503. http://
dx.doi.org/10.1148/rg.2017170056. PMid:28898189.
16. Raja JV, Khan M, Ramachandra VK, Al-Kadi O. Texture analysis
of CT images in the characterization of oral cancers involving
buccal mucosa. Dentomaxillofac Radiol. 2012;41(6):475-80.
http://dx.doi.org/10.1259/dmfr/83345935. PMid:22241875.
17. Albuquerque M, Anjos LG, Andrade HMT, Oliveira MS,
Castellano G, Rezende TJR,etal. MRI texture analysis reveals
deep gray nuclei damage in amyotrophic lateral sclerosis. J
Neuroimaging. 2016;26(2):201-6. http://dx.doi.org/10.1111/
jon.12262. PMid:26011757.
18. Headache Classification Subcommittee of the International
Headache Society. The International Classification of Headache
Disorders: 2nd edition. Cephalalgia. 2004;24(Suppl 1):9-160.
PMid:14979299.
19. Tasaki MM, Westesson PL, Isberg AM, Ren YF, Tallents RH.
Classification and prevalence of temporomandibular joint disk
displacement in patients and symptom-free volunteers. Am J
Orthod Dentofacial Orthop. 1996;109(3):249-62. http://dx.doi.
org/10.1016/S0889-5406(96)70148-8. PMid:8607470.
20. Pinto GNS, Grossmann E, Iwaki L Fo, Groppo FC, Poluha RL,
Muntean SA, et al. Correlation between joint effusion and
morphology of the articular disc within the temporomandibular
joint as viewed in the sagittal plane in patients with chronic
disc displacement with reduction: a retrospective analytical
study from magnetic resonance imaging. Cranio. 2021;39(2):119-
24. http://dx.doi.org/10.1080/08869634.2019.1582166.
PMid:30786838.
21. Oliveira MS, Fernandes PT, Avelar WM, Santos SL, Castellano G, Li LM.
Texture analysis of computed tomography images of acute ischemic
stroke patients. Braz J Med Biol Res. 2009;42(11):1076-9. http://
dx.doi.org/10.1590/S0100-879X2009005000034. PMid:19820884.
22. Fan WP, Liu MQ, Zou Y, Wang X, Wang PH, Chen ZY. MRI
histogram texture analysis of lateral pterygoid muscle in patients
with temporomandibular disorders. Zhonghua Kou Qiang Yi Xue
Za Zhi. 2020;55(12):963-8. PMid:33280361.
23. Ganeshan B, Miles KA, Young RC, Chatwin CR. Texture
analysis in non-contrast enhanced CT: impact of malignancy
on texture in apparently disease-free areas of the liver. Eur
J Radiol. 2009;70(1):101-10. http://dx.doi.org/10.1016/j.
ejrad.2007.12.005. PMid:18242909.
24. Haralick RM, Shanmugam K, Dinstein IH. Textural
features for image classification. IEEE Trans Syst Man
Cybern. 1973;SMC-3(6):610-21. http://dx.doi.org/10.1109/
TSMC.1973.4309314.
25. Gonçalves BC, Araújo EC, Nussi AD, Bechara N, Sarmento D,
Oliveira MS,et al. Texture analysis of cone-beam computed
tomography images assists the detection of furcal lesion. J
Periodontol. 2020;91(9):1159-66. http://dx.doi.org/10.1002/
JPER.19-0477. PMid:32003465.
Date submitted: 2022 Oct 04
Accept submission: 2022 Dec 16
Sérgio Lucio Pereira de Castro Lopes
(Corresponding address)
Universidade Estadual Paulista, Instituto de Ciência e Tecnologia,
Departamento de Diagnóstico e Cirurgia, Av. Eng. Francisco José Longo, 777 -
Jardim São Dimas, CEP 12245-000, São José dos Campos, SP, Brasil.
E-mail: sergio.lopes@unesp.br
