0008-3194/88/077-080/$2.00/0/)JCCA
Correlation of thermography with spinal dysfunction: preliminary results
Peter RP Diakow BSc, DC, FCCS(C)* Sandra Ouellet DC** Selina Lee DC** Edward J Blackmore BSc, DC
The results of a blinded pilot study researching the correlation of thermographic abnormalities and spinal segmental dysfunction are presented. The highest agreement resulted between focal increases in thermal emission andfixation (64.7%). Other parameters studied were tenderness and textural skin changes which agreed with thermographic findings 63.7% and 59.9%, respectively. Correlation coefficients were calculated for each subject and reasons for the wide range (r =.14 to.77) were discussed. Discussion of this research design and suggestions for future study were also presented. (JCCA 1988; 32(2):

77-80)

K E Y W 0 R D S: thermography, fixation, tenderness, skin rolling, chiropractic.
Les resultats d'une etude d'essai a l'insu sur la correlation des anomalies thermographiques et la dysfonction segmentaire de la colonne vertebrale sont ici pre'sente'es. La plus grande conformite a ete le resultat entre l'augmentation focale dans l'emission thermale et la fixation (64,7%). Des etudes supplementaires ont demontre une sensibilite et un changement de texture de la peau en accord avec les re'sultats thermographiques, 63,7% et 59,9% respectivement. Les coefficients de correlation ont ete calcules pour chaque patient et la raison de leur grande difference (r =.14 a.77) a ete discutee. On a aussi soumis des suggestions pour defutures etudes apres discussion de cetteformule de recherche. (JCCA 1988; 32(2): 77-80) M O T S C L t S: thermographie, fixation, sensibilite, prelevement d'un greffon, chiropraxie.

Introduction

Thermography is basically photography of heat emissions. In the health care field, thermography is playing an ever more important role in the diagnosis of disease" 2'3. Although modem thermography is less than 30 years old, its applications range from breast cancer screening, to carotid artery stenosis detection, to evaluation of neuromusculoskeletal problems. Research into this last field is increasing significantly, not only fueled by the spiralling costs encountered in the treatment of musculoskeletal problems, but also due to the attractiveness of thermography's objective, non-invasive and low-risk characteristics. The chiropractic profession also has an historical involvement in the detection of bodily heat emissions for diagnostic purposes. The neurocalometer was a heat-sensing device promoted by BJ Palmer in 19244. It was the forerunner of a series of instruments which used thermocouples placed in contact with the skin to sense surface heat. Later, Adelman and Kimmel used an instrument called the analyte which evaluated the light reflection from hemoglobin in surface capillaries5. More recently, both infrared sensing equipment and cholesteric ester solutions (liquid crystals) have been used to study spinal segmental dysfunction6'7. Although widely recommended as a means of providing objective evidence of segmental dysfunction by chiropractic thermographers, there is a paucity of controlled investigations correlating thermographic abnormalities with spinal dysfunction8'9. Further, there is some discrepancy in description of shape and location of the thermographic findings caused by spinal dysfunction. McFadden describes crescent-shaped areas of increased thermal emission lateral to the spine which he correlated with facet syndromel. Wexler, however, describes circular, focal areas of hyperthermia lateral to the spine"

Complex combinations of the proceeding two descriptions, as well as centrally located focal areas of hyperthermia, are proposed by Chapman to correspond to different categories of spinal dysfunction8. Reports from Europe differ quite extensively from these descriptions. In fact, Durianova and Engel, during three separate investigations, describe hypothermic zones appearing paraspinally which they ascribe to segmental spinal dysfunction'2"13"14. Engel also reported warming of hypothermic areas in response to spinal manipulation'2. This would seem to be in direct contrast to the North American experience. This study was designed to investigate the relationship between thermographic abnormalities overlying or adjacent to the spine and palpatory indicators of spinal segmental dysfunction. The palpatory indicators chosen were fixation, tenderness to palpation and textural changes noted upon skin rolling. Fixation, or restricted segmental mobility was detected by passive joint challenge techniques. Restrictions in joint mobility have become well accepted criteria for the presence of spinal segmental dysfunction. Tenderness over the spinous process also has a long history of association with spinal dysfunction 6. As well, textural changes noted upon skin rolling have been used as indicators of spinal dysfunction by osteopaths, chiropractors and medical doctors for years, and have received renewed interest lately as objective evidence of low-back paint7 18, For the purposes of this study then, segmental dysfunction was defined by the three indicators of fixation, tenderness, and textural skin changes. Thus, the presence of all three indicators at the same level increased the certainty that segmental dysfunction existed at that level.

Materials and methods

The thermography was conducted in a temperature controlled, draughtless room, using a Flexitherm Mark II liquid crystal thermography unit. Standard protocol as described elsewhere, was followed'9. Thermograms of the complete thoracic, lumbar
In partial fulfillment to maintain FCCS(C) certification.
** Private practice of chiropractic.

X JCCA 1988

The Journal of the CCA / Volume 32 No. 2 / June 1988
Correlation of thermography
and sacral areas were taken with markers set at the spine of the scapula, the inferior angle of the scapula, the T12-LI junction and the iliac crest. Thermograms were permanently recorded on Fujichrome 50 ASA slide film with a Pentax Program Plus camera and 1:mm. lens. All thermograms were read by one of the authors (PRP Diakow), an experienced thermographer certified by the American Board of Clinical Thermology. Thermograms were considered positive at a particular spinal level if a centrally or paraspinally located focal area was found to be distinctly hyperthermic or hypothermic. A temperature difference of 1C between a localized area and its surrounding area was considered significant. Palpation was performed by another member of the research team who was blinded to the thermographic findings. All subjects except #9 were palpated by the same examiner. The sample consisted of 10 volunteers all with a history of spinal segmental dysfunction but not in acute pain at the time of examination. Five were males and five were females, with ages ranging from 23 to 37 years.

Table I Agreement between thermography and joint dysfunction

Subject No.

Overall
78.9% 63.2% 68.4% 73.6% 47.4% 47.4% 47.4% 63.2% 26.3% 36.8%

Thermography fixation

84.2% 68.4% 73.6% 73.7% 63.2% 57.9% 68.4% 73.6% 26.3% 57.9%

Thermography tenderness

78.9% 63.2% 73.6% 73.6% 63.2% 52.6% 68.4% 68.4% 42.1% 52.6%
Thermography skin rolling

84.2% 63.2%

68.4% 73.7% 52.6% 47.4%

78.9% 36.8% 47.3% 59.9%

Averages 55.3%
Table II Pearson r for each subject

Subject

r.14.28.76.68.52.59.77.49.29.56

Results

Table I displays the agreement between thermography and the indicators of segmental dysfunction. The average agreement across all subjects for the combined palpatory examination is 55.3%. If the hypothermic findings are also considered, the average for the combined palpatory examination agrees with thermography 56.8%. Although the highest correlations occur between thermography and fixation (64.7%), -the differences are small considering the size of this preliminary sample. All thermographic abnormalities found were hyperthermic foci except for 7 distinctly hypothermic areas, all found between TI and T3. These hypothermic findings correlated with palpatory findings in 3 cases. One subject exhibited a central hypothermic focus with a focal hyperthermic area just lateral to it, which correlated with fixation and tenderness. Subject 9 was palpated by a different examiner, and one will note that the agreements are markedly below those of other subjects. In such cases, it is difficult to assess the effect of differences between the palpators and the variability in agreement that is also seen among other subjects. To better delineate the association between thermography and spinal segmental dysfunction, the Pearson Product-Moment Correlation coefficient 'r' was calculated. This statistic is used to identify the degree of linear relationship that exists between two variables. To calculate 'r', we converted the exact temperatures on the thermograms to a 7 point ordinal scale where 0 was assigned to 29.4C, 1 was assigned to 30.3C and so on until 6 represented 34.40C. All thermographic temperatures over each segmental and adjacent area were then converted to this scale and listed as the xi variable. The yi variable was calculated by giving each of the indicators of spinal dysfunction a value of 1. Thus, if there were no tenderness, textural changes or fixation found at T3, then YT3 would have a value of 0. If any one of these indicators were
.020.078.58.46.27.35.59.24.084.31
found, YT3 would have a value of 1, and so on. This scale implies that there is a greater chance that segmental dysfunction exists at a level when there are more indicators present at the same time. The number of levels measured was 19 (12 thoracic, 5 lumbar and 2 sacroiliac joints). The r and r2 values are listed in Table II. r2 varies from.078 to.59. In two of the subjects (subject 3 and 7) segmental dysfunction accounted for over half of the variation in segmental surface temperature.

Discussion

Although the results of this pilot study showed moderate agreement, the correlations calculated were poor except in two subjects: The original study design did not include the calculation of a correlation statistic. In retrospect, the authors believe that minor changes in experimental conditions would better suit data collection for the calculation of Pearson's r. First, room temperature was maintained at 22C for all subjects. Although each subject was given 15 minutes to equilibrate, the normal variation in their average surface temperatures gave some subjects "background" readings of 32.0C or even 32.8C. Focal hyperthermic areas were easily identified visually, however when temperature from these hotter individuals were converted to the 0 to 6 scale for analysis, the effect was that the range of
Correlation ofthermography
possible scores was restricted. Statistically, this lowers the resulting correlation20. In fact, the two subjects with the highest correlations also had lower background temperatures recorded. With this in mind, future studies should cool the subjects individually to obtain a predetermined "background" temperature. Instead of maintaining a room temperature to suit each individual which would be time consuming and cumbersome, the individuals may be artificially cooled with a water mist and blower technique. These techniques have been used widely to highlight thermal abnormalities and remove anomalies without changing the essential interpretation or reliability of the thermogram2l. In so doing, a greater range of spinal temperatures will be produced, thereby providing the greatest opportunity to study the actual association between thermographic findings and palpation of spinal dysfunction. The choice of palpation for spinal dysfunction to correlate with thermography was a difficult one. Unfortunately, there does not exist a "gold standard" for the detection of joint dysfunction; in fact, the reliability and validity of palpating for spinal dysfunction are not known. Nonetheless, palpation for tenderness, joint restriction and skin rolling are central to chiropractic diagnostics22'23'24. This problem may be eliminated by the use of three palpators (as opposed to a "gold standard"), who would provide a consensus of opinion regarding positive or negative findings at each level25, and still utilize our three aspects of joint dysfunction. These three qualities define what we are looking for, and may in fact represent different aspects of segmental dysfunction. It is possible that each of these attributes might correspond with temporal differences or differences in degree of severity. Tenderness or textural changes each may appear earlier or later in the complex process of dysfunction. Similarly, thermographic changes may also not appear until later in the process, thus resulting in poor correlation with earlier indicators. These relationships warrant further study. Many of the thermographic abnormalities listed at levels not found to be dysfunctional are, in fact, adjacent to dysfunctional segments. There are two possible explanations. First, with only 4 markers to estimate the spinal level on thermograms, this may represent the resolution limits of defining spinal level. Although marking each segment would be impractical, two or three more markers, may increase the reliability of assigning thermographic findings to one spinal segment or another. Alternately, it has been well established that within the sensory fields of an individual, adjacent nerve roots overlap extensively on the trunk26. Therefore, abberent neurological reflexes originating in one segment could affect thermal abnormalities in the suprajacent or subjacent levels. Perhaps this represents the 'neurological resolution' of thermography. In any case, a large sample with attention to these details may help to clarify this relationship. An important outcome of this preliminary investigation is the relatively low proportion of hypothermic areas or areas of abnormally decreased thermal emission. Contrary to the EuroThe Journal of the CCA / Volume 32 No. 2 / June 1988

pean literature, most of the spinal and paraspinal thermal abnormalities noted were increases in thermal emission. This discrepancy likely results from subject selection. Our subjects were all relatively healthy volunteers without disability. The studies of Durianova, and Engel utilized symptomatic subjects with chronic, painful syndromes'2"13"4. Alternately, McFadden and others have concentrated on acute cases6' 0. Could there be a difference in the thermal properties of acute vs. chronic segmental dysfunction? In the present study, subjects 2, 3, and 8 all mentioned their "cold" spots as being chronic problems. Further, it has been established that recent nerve root or peripheral nerve injuries show increased thermal emission along the related dermatome, whereas chronic nerve injuries are imaged as 'cold' dermatomes27. Perhaps the underlying mechanisms producing first hyperthermia, then hypothermia in nerve injuries could also operate to produce similar time-related thermal effects for spinal fixations. Further investigation should be designed to elucidate this aspect of the thermal-fixation relationship. In conclusion, there is convincing evidence that thermography could provide objective evidence of segmental dysfunction. There are questions pertaining to the exact relationship, sensitivity, and the relationship of thermographic findings to chronicity which should be addressed in future studies.

Acknowledgements

The authors would like to thank Diagnostic Thermographic Services, Inc. for the use of their facility and equipment, the Niagara District Chiropractic Society for volunteers and CMCC for its financial support.

References

1 Henahan J. Thermography finds multitude of applications. JAMA 1982; 247(24): 3296-3303. 2 Uricchio JV. Electronic thermography. J Florida Med Assoc 1983; 70(10): 889-894d. 3 Carter LM. The clinical role of thermography. J Med Eng Technol 1978; 2: 125-9. 4 Gibbons RW. The evolution of chiropractic: Medical and social protest in America. In: Haldeman S, ed. Modem developments in the principles and practice of chiropractic. Appleton-Century Croft. NY, 1980: 3-24. 5 Triano JJ. The use of instrumentation and laboratory examination procedures by the chiropractor. in: Modern developments in the principles and practice of chiropractic. Ed. Scott Haldeman. Appleton-Century-Croft. NY, 1980: 231-267. 6 Dudley WN. Preliminary findings in thermography of the back. ACAJ Chiropr. 1978; 15: 583. 7 Kelso AF, Grant RG. Use of thermograms to support assessment of somatic dysfunction or effects of osteopathic manipulative treatment: preliminary report. JAOA 1982; 82(3): 59-65. 8 Chapman GE. Neuromusculoskeletal thermography. CTA Publishers, Chula Vista, Ca, 1984: 72-128. 9 Dudley WN. Discovery and correction of dermatomes using thermography. ACAJ Chiropr. 15 (April), 1981: 31-33. 10 McFadden JW. Thermography used to diagnose the facet syndrome. J Neuro Ortho Surg 1983; 4(4): 354-357.

11 Wexler CE. Lumbar, thoracic and cervical thermography. J Neuro Ortho Surg 1979; 1(1): 37-41. 12 Engel JM. Quantitative thermographie in der diagnostik und therapiekontrolle der manuellen medizin. Manuelle Medizin 1981; 20: 36-43. 13 Engel JM. Thermographische objektivierung der segmentalen neuropathophyiologie. Manuelle Medizin 1984; 22: 30-40. 14 Durianova J. Die thermografishe objektivierung der wirkung der reflextherapie bei schmerzhaften zustanden der HWS. Manuelle Medizen 1986; 24: 60-64. 15 Denslow JS: An analysis of the variability of spinal reflex thresholds. J Neurophysiol 1944; 7: 207-215. 16 Janse J, Houser RH, Wells BF. Chiropractic principles and technic. National College of Chiropractic, Chicago, 1947: 273. 17 Haldeman S. Discussion. In: The Neurobiologic Mechanisms in Manipulative Therapy, Ed. Korr IM., NY, Plenum Press, 1978:133. 18 Gunn CC, Milbrandt WE. Early and subtle signs in low-back sprain. Spine 1978; 3(3): 267-281. 19 Feldman F, Nicholoff EL. Normal thermographic standards for the cervical spine and upper extremities. Skeletal Radiol. 1984; 12: 235-249. 20 Christensen LB, Stoup CM. Introduction to statistics for the social
and behavioral sciences. Brooks/Cole, Belmont, Ca, 1986: 132-134. Weinstein G. The diagnosis of trigger points by thermography. Postgrad Med 1986; Mar spec: 96-98. West HG. Physical and spinal examination procedures utilized in the practice of chiropractic. In: Haldeman S (ed): Modem developments in the principles and practice of chiropractic. New York: Appleton-Century-Croft, 1980: 269-296. Johnson WL. The role of static and motion palpation in structural diagnosis. In: Goldstein M (ed): The research status of manipulative therapy. NINCDS, Monograph 15, US DHEW, NIH, Bethesda, MD, 1975: 249-253. Kirkaldy-Willis WH, Hill RJ. A more precise diagnosis for low-back pain. Spine 1979; 4(2): 102-107. Walter SD. Measuring the reliability of clinical data: the case for using three observers. Rev Epidem et Sante Publ, 1984; 32: 206-211. Chusid JG. Correlative neuroanatomy and functional neurology. 17 ed. Los Altos, Ca: Lange Medical Pub, 1979: 210. Green J, Noran WH, Coyle MC, Gildemeister RG. Infrared electronic thermography: a non-invasive diagnostic neuroimaging tool. Contemporary Ortho 1985; 11(1): 39-47.

INTERFERENTIAL THERAPY

MULTIPLEX STIMULATOR LTD.
now offers you more features than ever before at very affordable prices.

MULTIPLEX PT-1

portable model for home visits

MULTIPLEX CL-1

clinic model with suction electrodes
To receive more information about all our products and our dealer nearest you, simply fill in your name and address and mail to: