SciELO - Scientific Electronic Library Online

vol.10 número4 índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Medicina Oral, Patología Oral y Cirugía Bucal (Ed. impresa)

versión impresa ISSN 1698-4447

Med. oral patol. oral cir. bucal (Ed.impr.) v.10 n.4 Valencia ago.-oct. 2005


Method of help for the diagnosis of the temporomandibular joint internal derangements. 
Discriminant Analysis applied to the temporomandibular derangements
Método de ayuda para el diagnóstico de los trastornos de la articulación temporomandibular.
Análisis discriminante aplicado a los Trastornos Temporomandibulares


Jorge Pesquera Velasco (1), Guillermo Casares García (2), Nieves Jiménez Pasamontes (3), Francisco Antonio García Gómez (4)

(1) Doctor en Medicina y Cirugía. Profesor Asociado de la Facultad de Ciencias de la Salud, Departamento de Ciencias de la Salud III, 
Universidad Rey Juan Carlos, Médico estomatólogo del área I del IMSALUD
(2) Doctor en Medicina y Cirugía. Médico estomatólogo del Servicio de Estomatología del Hospital General Universitario "Gregorio Marañón"
(3) Profesora Asociada de la Facultad de Ciencias de la Salud, Departamento de Ciencias de la Salud III, Universidad Rey Juan Carlos I
(4) Profesor Titular de la Facultad de Ciencias de la Salud, Departamento de Ciencias de la Salud III, Universidad Rey Juan Carlos, Madrid

Dr. Jorge Pesquera Velasco
Universidad Rey Juan Carlos, Facultad de Ciencias de la Salud.
Departamento de Ciencias de la Salud III, Estomatología.
Avda Atenas s/n, 28922, Alcorcón, Madrid.
Teléfono: 914888904

Received: 13-06-2004 Accepted: 12-12-2004

Pesquera-Velasco J, Casares-García G, Jiménez-Pasamontes N, García-Gómez FA. Method of help for the diagnosis of the temporomandibular joint internal derangements. Discriminant Analysis applied to the temporomandibular derangements. Med Oral Patol Oral Cir Bucal 2005;10:294-300.
© Medicina Oral S. L. C.I.F. B 96689336 -ISSN 1698-4447


Objective: The purpose of the study is to find an objective method of help for the clinician in the diagnosis of the pathology of the temporomandibular joint, different of the image methods habitually utilized until this moment.
Material and method: This study is based initially on the data obtained of a sample of 1164 patients with symptoms and/or signs of pathology of the temporomandibular joint. Nine different and excluding diagnostic groups settled down, according to the classification of the American Academy of Orofacial Pain (AAOP), in collaboration with the International Headache Society (IHS). We realized magnetic resonances to the patients and were selected those that adjust to the clinical criterion and of diagnosis for the image, and could only in a diagnostic group. Finally 449 patients were selected, 390 women and 59 men.
The results obtained (expressed in percentage of well classified cases) by means of the proposed method were: Arthrosis 98,9%, Anterior Disk Displacement with Reduction (ADDR) 87,5%, Anterior Disk Displacement without Reduction (ADD) 100%, Capsulitis 100%, Disk Immobile (DIN) 97.9%, Hypermobility Condylar (HC)95,8%, Lateral Displacement Without Reduction (LD) 100%, Pathology Muscular (PM)100%, Disk Hipomobile (DHM) 86,4%.
Conclusion: The proposed method reaches a fine percentage of successes in the diagnosis of these processes good enough, through its effectiveness as for its cost and should be considered an alternative in the diagnosis of temporomandibular derangements.

Key words: Diagnosis, temporomandibular joint, discriminant analysis, Magnetic resonance.



Objetivo: El propósito del estudio es encontrar un método objetivo de ayuda para el clínico en el diagnóstico de la patología de la articulación temporomandibular, diferente de los métodos de imagen habitualmente utilizados hasta este momento.
Material y método: Para ello se diseño un estudio basado en los datos obtenidos inicialmente de una muestra de 1164 pacientes con síntomas y/o signos de patología de la articulación temporomandibular. Se establecieron nueve grupos diagnósticos distintos y excluyentes, según la clasificación de la Academia Americana de Dolor Orofacial (AAOP), en colaboración con la de la Sociedad Internacional de Cefalea (IHS). Se realizaron resonancias magnéticas a los pacientes y se seleccionaron aquellos que cumplían el criterio clínico y de diagnóstico por la imagen de poder ser clasificados sólo en un grupo diagnóstico, 449 pacientes resultaron entonces seleccionados, 390 mujeres y 59 hombres.
Resultados: Los resultados obtenidos (expresados en porcentaje de casos bien clasificados) mediante el método propuesto fueron: Artrosis 98,9 %, Desplazamiento Anterior con Reducción del disco (DACR) 87,5 %, Desplazamiento sin Reducción del disco (DASR) 100 %, Capsulitis 100%, Disco Inmóvil (DIN) 97.9 %, Hipermovilidad Condilar 95,8%, Desplazamiento Lateral Sin Reducción DLSR 100 %, Patología Muscular 100 %, Disco Hipomóvil 86,4%.
El método propuesto alcanza un porcentaje de éxitos en el diagnóstico de estos procesos suficientemente bueno como para ser considerado, tanto por su eficacia como por su costo, una alternativa a considerar.

Palabras clave: Diagnóstico, articulación temporomandibular, análisis discriminante, Resonancia Magnética.



The diagnosis of the temporomandibular joint internal derangements has been ab initio a difficult task.

First caused by the difficulty that the professionals had with the nomenclature of the different diseases that affect the TMJ, "Alteration Temporomandibular" "Costen’s syndrome" "Pain-dysfunction syndrome", "Temporomandibular joint dysfunction syndrome" "Acute Arthritis and chronicle arthritis of the TMJ", "Temporomandibular disorders ", or "Pain-dysfunction miofascial syndrome"; as in same these pathologies can be different, equal or complementary. This situation leads to mistakes and confusions. Most of the studies suffer of lack of a precise diagnosis, not establishing diagnostic categories, only speaking of patient with pain-dysfunction. Today, the controversy has not disappeared all completely regarding to the denomination of the processes and to their correct classification (1).

Second caused by the absence of objective tests, (serologic, image diagnosis and others, so much serologic as of diagnosis for the image or of another type) or for the low quality of the existing, to corroborate each one of the realized diagnoses.

The first studies were done with the x-rays transcranial oblique lateral projection, that gave a very faulty image and only of the bone structures. Later on techniques as the tomography, the artrography, the double contrasts artrography, and the computerized tomography were used. From the use in 1977, of the magnetic resonance (MR), this situation has changed drastically. At the moment we can know the different disk positions (2), muscular atrophy (3), etc.

The MR has its advantages, its low danger because not x-rays are used and the perfect visualization of the tissue. Among MR disadvantages we can mention, the necessity of big and complex installations (these problems are being solved with technical advances. These disadvantages limit the systematic use of the magnetic resonance (MR) like habitual diagnostic test in these processes. The non existence of these installations in many place, the high cost of each exploration, as well as, the impossibility of carrying out it in some patients, E.g.: claustrophobia (although equipment of open MR exist), patient that use prosthesis ferromagnetic, patients with pacemakers or some other neuro-electric stimulator, certain percentage of false positive... inconveniences that force us to always investigate on their possible existence before the realization of the MR.

We attempt to present a method based on the health history of each patient and their clinical exploration and also using mathematical analysis methods and the power of calculation of a personal computer, out coming the realization of precise diagnoses in the case of the derangements of the Temporomandibular joint (TMJ). Also its cost has been very low.


For the realization of this study it has been used the clinical data picked up initially of 1164 patients of both sexes, with a range between 6 and 75 years, the average presentation age is 31 years old (1020 women with a range between 6 and 68 years, the average presentation age 31, 8 years old and 144 men between 7 and 75 years and the average presentation age 25, 5 years old). These patients were assisted in the Clinic of TMJ, of the Service of Stomatology of the University General Hospital "Gregorio Marañon" from Madrid (Spain). Data were obtained from the application form of CLINICAL HISTORY that is habitually used for the diagnostic of the patients when we receiving a new patient. This application form contains information of anamnestic type, and also the information obtained of the patient’s exploration. The anamnesis information was picked up from the questions answered by the patient and the exploratory discoveries were found following the exploration protocol used in the mentioned Clinic.

Over this initial sample, later on we are carried out MR to 756 patients, of which 449 patients were selected to have an unique diagnosis starting from the images of resonance (390 women of ages between 6 and 68 years, with mean age of 35,3 and 59 men of ages between 7 and 42, with mean age of 22,7).

Finally we did a DIAGNOSIS, by means of the data obtained by the exploration and from the obtained data of the MAGNETIC RESONANCE.

The image technique was done by means of an apparatus Gyroscan T-5 Upgrade, using the corporal antenna as transmitter and an antenna of circular surface of 8,5 cm. of diameter, placed parallelly to the patient’s face. The protocol to obtain the image it consisted of four phases:

1. Obtaining of two plane locators: A coronal, to appreciate the orientation of the condyle about the plane corresponding parasagital and another axial locator to get the angular capital oblique with respect to the plane biauricle.

2. A dynamic study in the obtained plane with use of sequences FFE (gradient echo) with a TR=200 ms, TE=15ms, angle alpha=50º, two excitements, FOV (field of visualization=170 mm, a matrix of 256 x 256 and 2 mm of section thickness. The dynamic study consists of four sequences done in progressive opening of the mouth, without positioner, leaving from the maximum intercuspation to the maximum opening. The total duration of the study is of 1, 30 x 4, for each joint.

3. When a possibility of a the sagittal plane of a lateral or medial displacement existed, some courts in the plane coronal were done, by means of sequences spin echo with a TR=400, TE=20, NEX=4, a matrix of 256 x 256, being carried out six courts for study in position of closed mouth.

4. When one suspects of a possible traumatic, tumour or inflammatory pathology, the study was completed with a sequence coronal spin echo empowered in T2 and proton density (TR=2000, TE=20-90, 1 NEX, 12 courts, thickness of court 5 mm, FOV=25 mm), only in position of closed mouth.

The Final diagnosis was divided in nine categories, using the discoveries of the Magnetic Resonance. The classification of The American Academy of Orofacial Pain was used (AAOP), in collaboration with the International Headache Society (HIS) (4). The code Numbers are the ones established in that classification (5):

1. Arthrosis (11.7.5): When bony degenerative signs exist.

2. Anterior Disk Displacement with Reduction (ADDR) ( When the disk is advanced in closed mouth, and it was placed in good position in situation of open mouth.

3. Anterior Disk Displacement without Reduction (ADD) ( When the disk appeared advanced in the four opening sequences.

4. Capsulitis ( When clinic of pain spontaneous preauricular exists and to the exploration and discoveries don’t exist in the image of resonance.

5. Immobile disk (IND) ( When the disk remained in the same position in the four courts, well in normal position or in early position although the condyle maintains its mobility (3).

6. Hypermobility Condylar (11.7.3): When in the image of maximum opening the condyle surpasses a vertical line traced perpendicularly to a horizontal line that goes by the most inferior point in the temporary prominence.

7. Anterior disk displacement without reduction with medial or lateral component (ADDL) ( valued in the sagittal cuts by means of the sign of the empty fossa (6) and two of the studies in courts crown them.

8. Muscular pathology (11.8): When signs of muscular degeneration are appreciated (fatty infiltration), such changes can be the result of a chronic inflammation or myositis (7, 8), without being appreciated pathology signs in the joint.

9. Hipomovil disk (11.7.7): When the mobility of the disk was not the normal, but still existed.

For the storage of these data a form was made with the database ACCESS® (Microsoft Access Version 2.0, Microsoft Corporation, Redmond, Washington).

For the development of an alternative mathematical model to the diagnosis of the pathology of the TMJ, the Statistical Discriminant Analysis Step by Step (stepwise) (9) was used. The calculations have been carried out by means of the program 7 M of the statistical package BMDPA® (10).

To evaluate the efficiency of the method a confusion table was built. This is a table of crossed frequencies that reflects the results of applying this procedure to the observed cases (11-13).


After the application of the Discriminant analysis to all the variables of our clinical history, the program selected 59 of the 111 variables that it considered more important to establish the diagnosis, making a mathematical matrix later on. Once done this matrix and applying the obtained values to each one of the patients of the sample, according to the method of the Discriminant analysis, the results obtained are:

1. The results obtained in the sample T (the total number of the patients). The percentage of good diagnoses according to this method oscillates among 86, 4% group of the DISK HIPOMOVIL (19 well classified, 3 bad classified) and 100% ADD (26 well classified, 0 bad classified), CAPSULITIS (4 well classified, 0 bad classified), DASRL (4 well classified, 0 bad classified), MUSCULAR PATHOLOGY (6 well classified, 0 bad classified) with a media value of 97, 3% for all the groups. Being of 98,9% for the group of ARTHROSIS (179 well classified, 2 bad classified), 87,5% for the ADDR (14 well classified, 2 bad classified), 95,8% for the HYPERMOBILITY CONDYLAR (23 well classified, 1 bad classified) and 97,9% for the DIN (47 well classified, 1 bad classified).

2. When the Jack-Knife technique is applied (9) to the same sample (this technique decreases the possible optimism of the Discriminant analysis) they decreases the percentages of well classified patients to 94, 6 % for all groups: ARTHROSIS (177 well classified, 4 bad classified) 97, 8%. ADDR (14 well classified, 2 bad classified) 87, 5%. ADD (24 well classified, 2 bad classified) 92, 3%. CAPSULITIS (3 well classified, 1 bad classified) 75%. DIN (46 well classified, 2 bad classified) 95, 8%. HYPERMOBILITY CONDYLAR (22 well classified, 2 bad classified) 91, 7%. DASRL (4 well classified, 0 bad classified) 100%. MUSCULAR PATHOLOGY (6 well classified, 0 bad classified) 100%. DISK HIPOMOVIL (17 well classified, 5 bad classified) 77, 3%.

3. Applying the "cross validation", the same method is applied but alone to a part of patients that the program selects at random (sample V); it decreases the percentage of well classified patients to 93, 2 %. ARTHROSIS (62 well classified, 1 bad classified) 100%. ADDR (5 well classified, 2 bad classified) 71, 4%. ADD (7 well classified, 1 bad classified) 87, 5%. CAPSULITIS (3 well classified, 0 bad classified) 100%. DIN (20 well classified, 1 bad classified) 95, 2%. HYPERMOBILITY CONDYLAR (9 well classified, 3 bad classified) 75%. DASRL (2 well classified, 1 bad classified) 66, 7%. MUSCULAR PATHOLOGY (1 well classified, 0 bad classified) 100%. DISK HIPOMOVIL (1 well classified, 0 bad classified) 100%.


Doubts exist regarding the establishment of the diagnosis being only based on the images offered by the MR. There are many studies in which these anomalies in the disk position in normal-asymptomatic patient can be seen (14), but in our case the patients presented all symptoms.

The MR is a reliable method for the study of the disk, existing an agreement between its results and the obtained intraoperative with a sensibility between 0,86 and 0,98, a specificity between 0,87 and 1,00 and a predictive negative value between 0,78 and 0,89 as for the correct demonstration of the disk position (15-17). It also exists to a high agreement intra and interobservers (95% and 91% respectively) (18, 19).

We have not found any other study that uses this same technique on the studied topic, although it is a broadly used statistical technique as much both medicine and dentistry, but applied up to now in other topics ej: quantitative definition of the sounds in the TMJ (20), studies on the pain in the TMJ (21), differences of psychological profiles among different affected populations of problems in the TMJ (22), quantitative evaluation of the state of the disk (23).

We will only be able to treat the derangements of the TMJ efficiently if we establish the correct diagnosis, therefore our therapeutic success will depend on our diagnosis, but unfortunately the diagnosis of the temporomandibular derangements in most cases is very confused. The key elements for the establishment of the diagnosis classically have been based on the anamnesis, in the clinical exploration and in the diagnostic anaesthetics blockade (24). Also the image techniques (25) and the arthroscopy (26) have been used. Nevertheless, the studies demonstrate that the agreement in diagnosis based in different methods jointly come to independent results proximate lucky results (27).

In our study we used the clinical discoveries as much as the anamnestic discoveries, which can be good to differentiate the joint derangements of the muscular illness. After that, the exploration should be enough to establish a precise diagnosis (28,29), this academically seems easy and simple, but the reality, is that in many cases the patients have or they can have processes with similar symptom that don’t even have anything to do with the derangements of the TMJ, examples of these could be: from the Eagle’s syndrome , secondary trismus to traumatism for injection, other traumatisms, infections in the region, pulpitis, or endodontics treatments, until the migraines, tetanus, Parkinson disease or lesions of the central nervous system, scleroderma, fibromyalgia, tumours or metastasis(30). The method of the Discriminant analysis provides us of another more tool to carry out the precise diagnosis in our patients, differentiating the processes that affect the TMJ, as much as those that cause similar symptoms without any relation.


The presented method has reveals as a simple, easy and reliable method for the realization of the diagnosis of some of the most frequent pathologies in the joint temporomandibular, with some percentages of good diagnoses superior to 90% in many cases, that can be used in any patient, without contraindications of any type, with a small cost that allows its repetition as many times as needed allowing a long term pursuit, not requesting any professional additional knowledge. Point out that the method only has effectiveness when the patient suffers a single pathological process in the TMJ.

The study presents a very effective method for the diagnosis of the pathology of the TMJ, based on the clinical history and a mathematical model.


1. Steenks, MH. Inclusion, Exclusion or Diagnosis?. Journal of Orofacial Pain 2004;18:81.        [ Links ]

2. Isberg, A. Disfunción de la Articulación Temporomandibular, una guía práctica. Sao Paulo: Artes Médicas; 2003        [ Links ]

3. Benito C, Casares G, Benito C. TMJ static disk: correlation between clinical findings and pseudodynamic magnetic resonance images. Cranio 1998;16:242-51.        [ Links ]

4. McNeill C.Temporomandibular Disorders. Guidelines for classification, assessment and management. 2ª ed. American Academy of Orofacial Pain. Carol Strean, Illinois: Quintesence; 1993.        [ Links ]

5. Okeson JP. Orofacial pain: guidelines for assessment, diagnosis, and management, 3ª Ed., Chicago: Quintessence; 1996.        [ Links ]

6. Kaplan PA, Tu HK, Williams SM, Lydiatt DD. The normal TMJ. MR and artrographic correlation. Radiology 1987;165:177-8.        [ Links ]

7. Katzberg RW, Besette RW, Tallents RM. Normal and abnormal TMJ: MRI with surface coil. Radiology 1986;158:183-9.        [ Links ]

8. Schellas KP, Wilkes CM. TMJ: MRI of internal derangement and postopera-tive changes. AJNR 1987;8:1093-101.        [ Links ]

9. Spicer C.C.; Hywel Jones, J.; Lennard-Jones, J.E. Discriminant and Bayes analysis in the differential diagnosis of Crohn’s disease and proctocolitis. Methods Inform Med 1973;12:118-22.        [ Links ]

10. Dixon, W.J. BMDP stadistical software 1981. Berkeley: University of California; 1981.        [ Links ]

11. Juez Martel P, Díez Vegas F.: Probabilidad y estadística en medicina, Madrid: Díaz de Santos, S.A.; 1997.        [ Links ]

12. Salvador Figueras, M: "Análisis Discriminante",, Estadística        [ Links ]

13. Huberty, C.J. Applied Discriminant Analysis. Wiley. Interscience 1994.        [ Links ]

14. Katzberg RW, Westesson PL, Tallents RH, Drake CM: Anatomic disorders of the temporomandibular joint disc in asymptomatic subjects. J Oral Maxillofac Surg.1996 Feb;54:147-53.        [ Links ]

15. Westesson PL. Reliability and validity of imaging diagnosis of temporomandibular joint disorder. Adv Dent Res 1993;7:137-51.        [ Links ]

16. Bell KA, Miller KD, Jones JP. Cine magnetic resonance imaging of the temporomandibular joint. Cranio 1992;10:313-7.        [ Links ]

17. Tasaki MM, Westesson PL. Temporomandibular joint: diagnostic accuracy with sagittal and coronal MR imaging. Radiology 1993;186:723-9.        [ Links ]

18. Tasaki MM, Westesson PL, Raubertas RF. Observer variation in interpretation of magnetic resonance images of the temporomandibular joint. Oral Surg Oral Med Oral Pathol 1993;76:231-4.        [ Links ]

19. Nebbe B, Brooks SL, Hatcher D, Hollender LG, Prasad NG, Major PW. Interobserver reliability in quantitative MRI assessment of temporomandibular joint disk status. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;86 :746-50.        [ Links ]

20. Leader JK, Robert Boston J, Rudy TE, Greco CM, Zaki HS, Henteleff HB. Quantitative description of temporomandibular joint sounds: defining clicking, popping, egg shell crackling and footsteps on gravel. J Oral Rehabil. 2001;28:466-78.        [ Links ]

21. Mongini F, Italiano M, Raviola F, Mossolov A. The McGill Pain Questionnaire in patients with TMJ pain and with facial pain as a somatoform disorder. Cranio. 2000;18:249-56.        [ Links ]

22. Suvinen TI, Reade PC, Sunden B, Gerschman JA, Koukounas E. Temporomandibular disorders: Part II. A comparison of psychologic profiles in Australian and Finnish patients. J Orofac Pain. 1997;11:147-57.        [ Links ]

23. Nebbe B, Major PW, Prasad NG, Hatcher D. Quantitative assessment of temporomandibular joint disk status. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85:598-607.        [ Links ]

24. Okeson J.P.: Tratamiento de la oclusión y afecciones temporomandibulares, Madrid: Harcourt Brace de España S.A.; 1999.        [ Links ]

25. Martinez Blanco M, Bagán JV, Fons A, Poveda Roda R. Osteoarthrosis of the temporomandibular joint. A clinical and radiological study of 16 patients. Med Oral. 2004;9:106-10.        [ Links ]

26. Alvarez J, Barbier L, Carmelo Martín J, Romo L, Andikoetxea B, Santamaria J. Temporomandibular arthroscopy: a retrospective clinical study of 61 cases. Med Oral 2001;6:383-90        [ Links ]

27. Emshoff R, Rudisch A. Validity of clinical diagnostic criteria for temporomandibular disorders. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001;91:50-5        [ Links ]

28. Lobbezoo-Scholte AM, De Leeuw JR, Steenks MH, Bosman F, Buchner R, Olthoff LW. Diagnostic subgroups of craniomandibular disorders. Part I: Self-report data and clinical findings. J Orofac Pain. 1995;9:24-36.        [ Links ]

29. Lobbezoo-Scholte AM, Lobbezoo F, Steenks MH, De Leeuw JR, Bosman F. Diagnostic subgroups of craniomandibular disorders. Part II: Symptom profiles. : J Orofac Pain. 1995;9:37-43.        [ Links ]

30. Min-Suk H, Byung-Mo A, Sam-Sun L, Sonn-Chul Ch. Use of advanced imaging modalities for the differential diagnosis of pathoses mimicking temporomandibular disorders. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2003;96:630-8.        [ Links ]

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons