Medical infrared thermal imaging, also known as thermal imaging or thermography, is a valuable noninvasive diagnostic imaging technique increasingly used in clinical practice. Medical infrared thermal imaging involves the recording of cutaneous thermal patterns generated by the emission of heat; these patterns are recorded in the form of a color map that is then interpreted.1–5 Medical infrared thermal imaging relies on sympathetic nervous system control of skin blood flow to reflect sympathetic nervous system response to dysfunction. There is no direct conduction of heat from deeper structures to the skin for heating or cooling.1,2,6–13 In this regard, MITI may be used as an adjunctive imaging technique to aid in the interpretation of physical examination findings, as a screening test to help guide therapeutic management, and as a means to assess long-term response to treatment.1–4,12,14–21 In humans, MITI has been successfully used for evaluation of burn patients22 and assessment of breast cancer,22,23 vascular disorders,7,10,13,24 scrotal varicocele,25 pneumothorax,26 radiculopathies,27 intervertebral disk disease,28,29 Chiari malformation and syringomyelia,30,31 pulpal blood flow to teeth,32 and joint disease.17,20,33–38 In veterinary medicine, it has been used in lameness evaluations to detect ligamentous, osseous, muscular, articular, and neurologic injuries in horses, cats, dogs, and raccoons.3,4,18,19,39–43
Hyperthyroidism is a multisystemic disease defined as an excessive production and secretion of thyroid hormones. In humans, the most common cause for hyperthyroidism is Graves’ disease, which is associated with an autoimmune disease process.44,45 Toxic nodular goiter affects geriatric humans; in this condition, one or more hyperfunctioning adenomatous nodules are present in the thyroid gland. Toxic nodular goiter closely resembles hyperthyroidism in cats46; however, the true etiopathogenesis of feline hyperthyroidism remains unknown. Thyroid carcinoma has been documented in 1% to 3% of hyperthyroid cats47–51 and is the primary cause of hyperthyroidism in dogs. Excessive circulating concentrations of thyroid hormones can increase metabolic rate and sensitivity to catecholamines, both of which cause cardiovascular and metabolic abnormalities.47,52–55 These abnormalities can perpetuate multisystemic organ dysfunction.
Hyperthyroidism develops in cats > 6 years old, with a mean age of onset of 12.8 years.52,56 There is no known breed or sex predilection, unlike in humans, among whom females are 4.5 times as likely as males to develop hyperthyroidism.57–59 In cats, the most common clinical sign is polyphagia with concurrent weight loss.52,56 Approximately 10% of cats will have signs of depression, weakness, progressive weight loss, and anorexia, and often these signs are a consequence of concurrent nonthyroidal illness. In humans, this nonthyroidal illness is described as apathetic hyperthyroidism.56 Clinical signs and physical examination findings of cats with hyperthyroidism may include weight loss, polyphagia, hyperactivity, tachycardia, polyuria, polydipsia, cardiac murmurs, vomiting, diarrhea, increased fecal volume, anorexia, polypnea, dyspnea, muscle weakness or tremors, and congestive heart failure. Poor coat quality (matting and shedding) have been described as well. Palpably enlarged thyroid gland lobes are evident in approximately 90% of hyperthyroid cats,47,49,52,60 although normally sized thyroid glands have also been reported.61
The clinical signs of hyperthyroidism in cats are considered nonspecific, and a variety of disease processes may possibly result in any combination of these signs. For most cats, establishing a diagnosis of hyperthyroidism is achieved by detection of clinical signs consistent with the disease, the presence of a palpably enlarged thyroid gland, and high serum thyroxine concentration (> 4.0 mg/dL).47,52 Other diagnostic techniques that may be used to evaluate a cat for hyperthyroidism include the triiodothyronine suppression test, thyrotropin-releasing hormone stimulation test, nuclear scintigraphy, thoracic radiography, thoracic or cervical CT, and cervical ultrasonography.47,52,61 The assessment of basal serum free thyroxine concentration by equilibrium dialysis, in combination with serum total thyroxine measurement, may aid in the diagnosis of hyperthyroidism in cats with mild disease, for which previous total thyroxine concentrations had remained within the reference interval.47 Nuclear scintigraphy with technetium Tc 99m compounds and the use of a gamma camera is very effective in identifying hyperfunctioning thyroid glands as well as locating possible ectopic thyroid tissue47,50,61; however, nuclear scintigraphy is expensive, requires patient quarantine, and necessitates regulatory controls to limit radiation exposure to both patient and operator.
Medical infrared thermal imaging may have potential as a noninvasive screening test for hyperthyroidism in cats. Studies62,63 involving MITI in humans with thyroid gland disease have revealed that this imaging technique is less expensive than scintigraphy (infrared cameras cost $8,000 to $18,000), noninvasive, and safe. Because MITI does not emit radiation, the procedure is not under legal regulation and does not require isolation of patients. The purpose of the study reported here was to determine the usefulness of MITI as a screening tool for hyperthyroidism in cats, evaluate the need for hair clipping over the ventral aspect of the neck to achieve optimal images, and determine whether there is a change in thermal patterns at 1 and 3 months after radioactive sodium iodide I 131 treatment.
Materials and Methods
INCLUSION CRITERIA FOR HYPERTHYROID CATS
All cats examined at Long Island Veterinary Specialists between October 2010 and February 2012 that had clinical signs consistent with hyperthyroidism and had a serum total thyroxine concentration > 4.0 mg/dL (reference range, 0.8 to 4.0 μg/dL) were included in the study. Written consent from the owner was acquired for each cat with hyperthyroidism in the study. Cats were excluded from the study when the physical examination revealed any comorbid disease associated with the ventral cervical area (eg, trauma or dermatitis) that may have affected image interpretation.
CONTROL CATS
Healthy cats owned by employees of Long Island Veterinary Specialists that had no clinical signs compatible with hyperthyroidism and had serum thyroxine concentrations within the reference range provided by the laboratory performing the tests were selected for inclusion in the study. Written consent from the owner was acquired for each control cat's study participation. Cats were excluded from the study if physical examination revealed any disease associated with the ventral cervical area.
PHYSICAL EXAMINATION AND SERUM THYROXINE CONCENTRATION TESTING
All protocols for this study were approved by the department heads of surgery and internal medicine at Long Island Veterinary Specialists. Each cat underwent physical examination by a board-certified internist (JWT) at least 24 hours prior to admission for MITI and serum thyroxine concentration measurement. A blood sample (2 to 3 mL) was collected from the medial saphenous vein for biochemical testing. Control cats underwent MITI and serum thyroxine concentration measurement once; cats with hyperthyroidism underwent MITI and serum thyroxine concentration measurement 3 times (before and at 1 and 3 months after radioiodide treatment).
MITI
Prior to MITI, each cat had exercise restricted and was housed in a cage in a temperature-controlled (24°C), draft-free room for 30 minutes. The MITI was performed in another room (also maintained at 24°C). The MITI was performed in a room maintained at room temperature. A stand-mounted infrared camera with a focal plane array amorphous silicon microbolometera with an emissivity setting of 1 was used. For real-time data analysis, the camera was connected to a laptop computer. To minimize thermal artifacts from manual contact, the cat was handled by the forelimbs and head by trained technicians who wore latex gloves at all times during handling. To minimize background artifact potentially created by temperature differences between exterior walls, each cat was positioned in front of a uniform interior wall with the camera placed 1 m in front of the cat. To expose the ventral aspect of the neck for MITI, the forelimbs were held and the head was extended dorsally (Figure 1). Image acquisition was instantaneous, similar to that achieved with a digital camera.
Schematic drawings illustrating positioning of a cat undergoing MITI. A—The cat is placed in a sitting position with the head extended dorsally and the forelimbs held (not illustrated). B—The infrared camera is placed 3 feet in front of the cat to obtain an image of the ventral aspect of the neck.
Citation: American Journal of Veterinary Research 76, 1; 10.2460/ajvr.76.1.53
For each cat, the first set of infrared images was obtained with the coat intact. After the initial images of the unshaved cat were obtained, the hair of the ventral aspect of the neck was clipped with a No. 40 clipper blade in a manner consistent with surgical preparation (Figure 2). The imaging protocol used to obtain the initial images was again applied for each shaved cat 60 minutes after clipping, on the basis of previous research findings.15
Representative images of the ventral aspect of the neck of 2 cats obtained by MITI. A—Image obtained from a cat without hyperthyroidism before the ventral aspect of the neck was shaved. B—Image obtained from the same cat in panel A after the ventral aspect of the neck was shaved. C—Image obtained from a cat with hyperthyroidism before the ventral aspect of the neck was shaved. D—Image obtained from the same cat in panel C after the ventral aspect of the neck was shaved. Dorsal is toward the top of each image.
Citation: American Journal of Veterinary Research 76, 1; 10.2460/ajvr.76.1.53
A software programb was used to save, analyze, and review the image data. Images were converted from black and white to a preset 8°C temperature scale and a 16-shade color map. White and red were selected to represent warmer temperatures and blue and black were selected to represent cooler temperatures. The program calculated the mean, maximum, and minimum temperatures of each image. Imaging processing softwareb was used for pattern recognition and to analyze and evaluate thermographic patterns.
After the initial MITI data (obtained before and after hair clipping) for control and hyperthyroid cats were evaluated, the hyperthyroid cats underwent radioactive sodium iodide I 131 treatment. As part of posttreatment evaluations, MITI of cats with hyperthyroidism was performed in a similar manner (ie, before and after clipping of the hair on the ventral aspect of the neck) at 1 and 3 months after treatment. Control cats underwent MITI only once.
RADIOACTIVE SODIUM IODIDE I 131 TREATMENT
Cats with hyperthyroidism received radioiodide treatment in accordance with New York State regulations. The dose of radioactive iodine used to treat each cat was based on serum thyroxine concentrations. Each cat was brought to a lead-lined cat boarding room and given an SC injection (on the dorsum) of the appropriate radioiodide treatment. The cat was housed in an isolated feline boarding facility within the lead-lined room maintained at 24°C. Each cat remained in the hospital for a minimum of 5 days and was discharged only when the radiation level was < 0.5 mR/h. Each cat was discharged from the hospital with appropriate instructions for the owners regarding handling and discarding of litter in accordance with New York State regulations.
At 1 month after the initial examination, imaging, and radioactive sodium iodide I 131 treatment, treated cats were reevaluated. For each cat, the reevaluation included a complete history from the owners to evaluate clinical signs, a complete physical examination performed by a board-certified internist (JWT), and assessment of serum thyroxine concentration. Medical infrared thermal imaging was repeated (Figure 3). Treated cats were similarly reevaluated at 3 months after radioiodide treatment.
Representative images of the unshaved ventral aspect of the neck of 2 cats obtained by MITI. A—Image obtained from a cat without hyperthyroidism. B—Image obtained from a cat with hyperthyroidism prior to radioactive sodium iodide I 131 treatment. C—Image obtained from the same cat in panel B 1 month after radioiodide treatment. D—Image obtained from the same cat in panel B 3 months after radioiodide treatment.
Citation: American Journal of Veterinary Research 76, 1; 10.2460/ajvr.76.1.53
PRETREATMENT IMAGE ANALYSIS
Ten sets of experiments were performed by applying 4 color transforms to the original images for both shaved and unshaved cats; 1 set of experiments corresponded to 1 set of images. After the color transforms, there were 5 sets of images obtained before shaving of the ventral aspect of the neck and 5 sets of images obtained after shaving for each cat. An experiment consisted of a fixed set of image features with a data normalization method. Then, a leave-one-out cross-validation experiment was performed with the entire set of images by means of the k-nearest neighbors’ algorithm for pattern classification. Percentage of correct classifications of images as being from a control or hyperthyroid cat for the leave-one-out pattern classification was the success rate. The temperature data were mapped to colors in the original images for visualization. To assure the pattern classification was unbiased, this visualization-based mapping was transformed into other mathematical color spaces to explore which mapping provided the best classification results. For this study, 4 color normalization methods (color normalization with luminance, grayscale color normalization, RGB color normalization, and RGB color normalization with luminance)64 along with the original RGB images were used to compose the 5 sets of experiments. To range-normalize the feature data before classification, 2 methods were used: softmax (or sigmoidal) normalization and standard normal density. The experiments were performed with computer vision and image processing software,b with each set having 2,046 permutations, and by means of the k-nearest neighbors’ algorithm for pattern classification, histographic, spectral, and texture features with a pixel distance of 6. The number of cats was not related to the number of experiments.
1- AND 3-MONTH RECHECK IMAGE ANALYSIS
Twenty sets of experiments were performed, with 10 sets of images obtained before shaving of the ventral aspect of the neck and 10 sets of images obtained after shaving for each cat. These experiment sets were the same as pretreatment images. Of those 10 sets of images for each cat, 5 sets of images comprised the 1-month recheck images and the other 5 sets comprised the 3-month recheck images. Additionally, 4 color normalization methods were used along with the original RGB images, as performed for pretreatment data. For the original data, 2 data normalizations were used: softmax normalization and standard normal density. Twenty sets of experiments were performed with the computer vision and image processing software,b and by means of the k-nearest neighbors’ algorithm for pattern classification, histographic, spectral, and texture features with a pixel distance of 6. Of these 16 sets, those with color transformations had 2,046 permutations; 4 sets, on the original data which did not include the mean feature, had 1,023 permutations.
Results
CLINICAL DATA
Seventeen cats with hyperthyroidism (consecutively examined) met the inclusion criteria. Their ages ranged from 9 to 17 years (mean, 12.6 years; median, 13 years). The cats’ weights ranged from 2.27 to 8.30 kg (mean, 5.67 kg; median, 4.69 kg). Before radioactive sodium iodide I 131 treatment, serum thyroxine concentrations in these cats ranged from 4.2 to 24 μg/dL (mean, 13.65 μg/dL; median, 12.3 μg/dL).
Twenty-eight cats were considered as controls. Sixteen cats were excluded because they were deemed too young to age match the clinically affected cats. One cat was excluded because of a high serum thyroxine concentration. Twelve cats confirmed to be healthy were included in the control group. These cats ranged from 6 to 18 years of age (mean, 10.75 years; median, 10 years). The cats’ weights ranged from 3.18 to 9.5 kg (mean, 4.9 kg; median, 4.5 kg). The serum thyroxine concentrations of the control cats ranged from 1.1 to 2.4 μg/dL (mean, 1.73 μg/dL; median, 1.8 μg/dL). Grayscale color normalization method analysis of images obtained prior to shaving of the ventral aspect of the cats’ necks provided an 80.5% accuracy in differentiating cats with high serum thyroxine concentrations (> 4.0 μg/dL) from the control population. Grayscale color normalization method analysis of images obtained after shaving of the ventral aspect of the cats’ necks provided an 87.5% accuracy in differentiating cats with high serum thyroxine concentrations (> 4.0 μg/dL) from the control population.
Cats with high serum thyroxine concentration (> 4.0 μg/dL) that received radioactive sodium iodide I 131 treatment and returned for re-evaluation 1 month later were all considered responsive to treatment, with improvement of clinical signs and reduction in serum thyroxine concentration. Fourteen cats (11 domestic shorthair, 2 domestic longhair, and 1 Siamese) underwent MITI 1 month after treatment. The 3 cats that were not reevaluated at 1 month after treatment included a domestic shorthair that was euthanized at its regular veterinarian's office because of metastatic lung disease, a domestic longhair that died of secondary heart disease at home, and a Himalayan that was euthanized at our facility because of renal failure. Of the 14 reevaluated cats, 10 had serum thyroxine concentrations within the reference interval (0.8 to 4.0 μg/dL), and 4 had serum thyroxine concentrations < 0.8 μg/dL.
Grayscale color normalization method analysis of images obtained before shaving of the ventral aspect of the treated cats’ necks at the 1-month recheck evaluation provided a 92.86% success rate in detecting cats with normal serum thyroxine concentration (< 4.0 μg/dL). Grayscale color normalization method analysis of images obtained after shaving of the ventral aspect of the treated cats’ necks at the 1-month recheck evaluation provided an 85.71% success rate in detecting cats with normal serum thyroxine concentration (< 4.0 μg/dL).
At the 3-month recheck evaluation, 13 treated cats (11 domestic shorthair and 2 domestic longhair) were considered euthyroid with resolution of clinical signs and a return of serum thyroxine concentration within the reference interval. For these cats, the hyperthyroidism was considered successfully managed. Of those 13 cats, 11 had serum thyroxine concentrations within a range of 0.8 to 2.7 μg/dL and 2 had serum thyroxine concentration < 0.8 μg/dL (0.08 and 0.7 μg/dL). One cat was not returned for reevaluation.
The RGB color normalization with luminance method analysis of images obtained before shaving of the ventral aspect of the treated cats’ necks at the 3-month recheck evaluation had a 100.00% success rate in detecting cats with normal serum thyroxine concentration (< 4.0 μg/dL). Grayscale color normalization of the same images had a 85.71% success rate. The RGB color normalization with luminance method analysis of images obtained after shaving of the ventral aspect of the treated cats’ necks at the 3-month recheck evaluation had a 100.00% success rate for detecting cats with a normal serum thyroxine concentration (< 4.0 μg/dL). Grayscale color normalization of the same images had a 92.86% success rate.
Discussion
In the present study, the computer recognition pattern analysis of the MITIs was successful in differentiating cats without hyperthyroidism from cats with hyperthyroidism. Results from images obtained from the shaved ventral aspect of the cats’ necks were better than those obtained prior to shaving, and grayscale color normalization was the best method of image analysis, with a success rate of 87.5%. The results of the present study are similar to results of human studies62,65 in which thermographic criteria were developed for euthyroidism, hypothyroidism, and hyperthyroidism. In 1 human study,65 thermal images of the thyroid gland of pregnant women were similar to the thermal images of euthyroid glands in other individuals, indicating the possible use of thermography as an objective method of diagnosis of thyroid gland disease in pregnant women.
Detection of hyperthyroid cats can be difficult because serum thyroid hormone concentrations are in a constant state of fluctuation and cats with developing hyperfunctioning thyroid glands may have serum thyroxine concentration that is considered normal or equivocal at time of initial blood testing. As with any disease process, early detection and early initiation of treatment can affect the pathogenesis or outcome of the disease. Hyperthyroidism may either mask the presence of chronic kidney disease66 or contribute to the progression of chronic renal damage.52 It may also contribute to the development of cardiomyopathy (thyrotoxic cardiomyopathy)67 and result in systemic hypertension68 and the known constellation of hyperthyroidism-related clinical signs. It has been reported that treatment of the underlying disease of hyperthyroidism can reverse the changes in cardiac muscle66; however, if untreated, cardiac remodeling may become permanent, resulting in long-term (often progressive) clinical signs associated with cardiomyopathy. The secondary effect of systemic hypertension has also been reported to be reversible with treatment of the primary underlying disease process (hyperthyroidism).68
This suggests that early diagnosis and early treatment are beneficial. Medical infrared thermal imaging has also been successful in identifying cutaneous temperature changes associated with disease before onset of clinical signs.69 In the present study, several cats with slightly high serum thyroxine concentration had abnormal thermograms; therefore, MITI can be used to screen cats with borderline high serum thyroxine concentration for possible hyperthyroidism.
Given the difficulty in shaving unsedated hyperthyroid cats, which can be fractious, and the success rate of 80.5% for MITI in cats with shaved neck regions, the need to shave the ventral aspect of a cat's neck to achieve greater imaging success is questionable. The use of MITI as a screening test for hyperthyroidism in cats could be useful in areas with high cat populations, such as a feline shelter or cattery, or in rescue situations where cats are difficult to handle and costs of serologic evaluation are high. A quick, noninvasive test could separate clinically normal cats from high-risk cats and potentially decrease the time to diagnosis and treatment initiation. Medical infrared thermal imaging offers an ability to screen a large population of cats with less restraint of individuals and simple and fast image acquisition. Image generation is not complicated, and training of veterinary support staff is feasible; however, clinical pattern analysis software would be required.
In the present study, MITI analysis was also successful in identifying radioiodide-treated cats that had serum thyroxine concentration within the reference interval at the 1- and 3-month recheck evaluations. The 1-month MITI reevaluation analysis had a success rate of 92.86% in cats that were not shaved and a success rate of 85.71% in cats that were shaved. The results also supported the notion that shaving the ventral aspect of a cat's neck for MITI may not be necessary. The 3-month MITI reevaluation analysis had even better results, with a 100% success rate for cats that were or were not shaved. Decreased handling of the cats that were not shaved may be responsible for the improved analysis success rates because of less stress and a diminished effect on the autonomic nervous system. These results have suggested that shaving the ventral aspect of a cat's neck is not essential for successful imaging and that MITI can be used to screen cats in shelters and catteries to identify those for which serum thyroid hormone concentrations should be assessed. It has been reported that < 5% of cats do not respond adequately to a single radioiodide treatment and thus require a second treatment.70–72 In the present study, no cat required a second treatment and among the 14 treated cats that were evaluated, serum thyroid concentrations had decreased by the 1-month recheck and remained within the reference interval at the 3-month recheck.
Limitations of the present study were related to numbers of cats within the experimental and control groups. Additionally, it is unknown whether findings would be similar in cats treated orally with medication; this was outside the scope of this investigation. Because no cats required a second or third radioactive sodium iodide I 131 treatment, it is unknown whether cats that do not fully respond to radioiodide treatment would have results of MITI analysis that correlated to serum thyroxine concentration.
Medical infrared thermal imaging could be a suitable noninvasive screening test for identification of cats likely to have overproductive thyroid glands. Results of the present study have indicated that MITI can detect cats with hyperthyroidism and monitor response to radioactive sodium iodide I 131 treatment. The method cannot be considered a specific test for diagnosis of hyperthyroidism because of other potential disease processes in the ventral cervical area. Other means of screening (ie, assessment of serum total and free thyroxine concentrations, triiodothyronine suppression testing, thyrotropin-releasing hormone stimulation tests, and scintigraphy) can be used to successfully evaluate a cat suspected of having hyperthyroidism.
ABBREVIATION
| MITI | Medical infrared thermal imaging |
Footnotes
Med 2000 IRIS, Meditherm Inc, Beaufort, NC.
CVIPtools, Computer Vision and Image Processing Laboratory, Department of Electrical and Computer Engineering, School of Engineering, Southern Illinois University, Edwardsville, Ill.
References
1. Leandro P. Meditherm manual of clinical thermology. Beaufort, NC: Meditherm, 2004; 3–6.
2. Uematsu S, Edwin DH, Jankel WR, et al. Quantification of thermal asymmetry. Part 1: normal values and reproducibility. J Neurosurg 1988; 69: 552–555.
3. Levet T, Martens A, Devisscher L, et al. Distal limb cast sores in horses: risk factors and early detection using thermography. Equine Vet J 2009; 41: 18–23.
4. Turner TA. Diagnostic thermography. Vet Clin North Am Equine Pract 2001; 17: 95–113.
5. Vainionpää M, Salla K, Restitutti F, et al. Thermographic imaging of superficial temperature in dogs sedated with medetomidine and butorphanol with and without MK-467 (L-659'066). Vet Anaesth Analg 2013; 40: 142–148.
6. Chan EK, Pearce JA. Visualization of dynamic subcutaneous vasomotor response by computer-assisted thermography. IEEE Trans Biomed Eng 1990; 37: 786–795.
7. Geatti O, Gasparini D, Shapiro B. A comparison of scintigraphy, thermography, ultrasound and phlebography in grading of clinical varicocele. J Nucl Med 1991; 32: 2092–2097.
8. Friedman MS. The use of thermography in sympathetically maintained pain. Iowa Orthop J 1994; 14: 141–147.
9. Park ES, Park CI, Jung KI, et al. Comparison of sympathetic skin response and digital infrared thermographic imaging in peripheral neuropathy. Yonsei Med J 1994; 35: 429–437.
10. Seifalian AM, Stansby G, Jackson A, et al. Comparison of laser Doppler perfusion imaging, laser Doppler flowmetry, and thermographic imaging for assessment of blood flow in human skin. Eur J Vasc Surg 1994; 8: 65–69.
11. Thomas D, Collins S, Strauss S. Somatic sympathetic vasomotor changes documented by medical thermographic imaging during acupuncture analgesia. Clin Rheumatol 1992; 11: 55–59.
12. Uematsu S, Jankel WR, Edwin DH, et al. Quantification of thermal asymmetry. Part 2: application in low-back pain and sciatica. J Neurosurg 1988; 69: 556–561.
13. Lawson W, BenEliyahu D, Meinken L, et al. Infrared thermography in the detection and management of coronary artery disease. Am J Cardiol 1993; 72: 894–896.
14. Collins AJ, Ring F, Bacon PA, et al. Thermography and radiology complimentary methods for the study of inflammatory diseases. Clin Radiol 1976; 27: 237–243.
15. Infernuso T, Loughin CA, Marino DJ, et al. Thermal imaging of normal and cranial cruciate ligament-deficient stifles in dogs. Vet Surg 2010; 39: 410–417.
16. Loughin CA, Marino DJ. Evaluation of thermographic imaging of the limbs of healthy dogs. Am J Vet Res 2007; 68: 1064–1069.
17. Ring EF, Dieppe PA, Bacon PA. The thermographic assessment of inflammation and anti-inflammatory drugs in osteoarthritis. Br J Clin Pract 1981; 35: 263–264.
18. Vaden MF, Purohit RC, McCoy MD, et al. Thermography: a technique for subclinical diagnosis of osteoarthritis. Am J Vet Res 1980; 41: 1175–1179.
19. Vainionpää MH, Raekallio MR, Junnila JJ, et al. A comparison of thermographic imaging, physical examination and modified questionnaire as an instrument to assess painful conditions in cats. J Feline Med Surg 2013; 15: 124–131.
20. Varjú G, Pieper CF, Renner JB, et al. Assessment of hand osteoarthritis: correlation between thermographic and radiographic methods. Rheumatology (Oxford) 2004; 43: 915–919.
21. Ventä I, Hyrkäs T, Paakkari I, et al. Thermographic imaging of postoperative inflammation modified by anti-inflammatory pretreatment. J Oral Maxillofac Surg 2001; 59: 145–150.
22. Head JF, Elliott RL. Infrared imaging: making progress in fulfilling its medical promise. IEEE Eng Med Biol Mag 2002; 21: 80–85.
23. Gautherie M, Haehnel P, Walter JP, et al. Thermovascular changes associated with in situ and minimal breast cancers. Results of an ongoing prospective study after four years. J Reprod Med 1987; 32: 833–842.
24. Verheye S, De Meyer GR, Krams R, et al. Intravascular thermography: immediate functional and morphological vascular findings. Eur Heart J 2004; 25: 158–165.
25. Tucker AT. Infrared thermographic assessment of the human scrotum. Fertil Steril 2000; 74: 802–803.
26. Rich PB, Dulabon GR, Douillet CD, et al. Infrared thermography: a rapid, portable, and accurate technique to detect experimental pneumothorax. J Surg Res 2004; 120: 163–170.
27. So YT, Aminoff MJ, Olney RK. The role of thermography in the evaluation of lumbosacral radiculopathy. Neurology 1989; 39: 1154–1158.
28. Domán I, Illés T. Thermal analysis of the human intervertebral disc. J Biochem Biophys Methods 2004; 61: 207–214.
29. Pochaczevsky R, Wexler CE, Meyers PH, et al. Liquid crystal thermography of the spine and extremities. Its value in the diagnosis of spinal root syndromes. J Neurosurg 1982; 56: 386–395.
30. Fujimoto A, Matsumura A, Nakamura K, et al. Chiari malformation type I associated with familial spastic paraplegia: report of a surgically treated case. Childs Nerv Syst 2005; 21: 336–338.
31. Koyanagi I, Iwasaki Y, Isu T, et al. Thermographic findings of syringomyelia [in Japanese]. No Shinkei Geka 1988; 16: 1149–1154.
32. Kells BE, Kennedy JG, Biagioni PA, et al. Computerized infrared thermographic imaging and pulpal blood flow: part 2. Rewarming of healthy human teeth following a controlled cold stimulus. Int Endod J 2000; 33: 448–462.
33. Gratt BM, Sickles EA. Thermographic characterization of the asymptomatic temporomandibular joint. J Orofac Pain 1993; 7: 7–14.
34. Gratt BM, Sickles EA, Wexler CE. Thermographic characterization of osteoarthrosis of the temporomandibular joint. J Orofac Pain 1993; 7: 345–353.
35. Rajapakse C, Grennan DM, Jones C, et al. Thermography in the assessment of peripheral joint inflammation—a re-evaluation. Rheumatol Rehabil 1981; 20: 81–87.
36. Salisbury RS, Parr G, De Silva M, et al. Heat distribution over normal and abnormal joints: thermal pattern and quantification. Ann Rheum Dis 1983; 42: 494–499.
37. Shilo R, Engel J, Farin I, et al. Thermography as a diagnostic aid in tennis elbow. Handchirurgie 1976; 8: 101–103.
38. Thomas D, Siahamis G, Marion M, et al. Computerised infrared thermography and isotopic bone scanning in tennis elbow. Ann Rheum Dis 1992; 51: 103–107.
39. Purohit RC, McCoy MD. Thermography in the diagnosis of inflammatory processes in the horse. Am J Vet Res 1980; 41: 1167–1174.
40. Purohit RC, McCoy MD, Bergfeld WA III. Thermographic diagnosis of Horner's syndrome in the horse. Am J Vet Res 1980; 41: 1180–1182.
41. Tunley BV, Henson FM. Reliability and repeatability of thermographic examination and the normal thermographic image of the thoracolumbar region in the horse. Equine Vet J 2004; 36: 306–312.
42. Vainionpää M, Raekallio M, Tuhkalainen E, et al. Comparison of three thermal cameras with canine hip area thermographic images. J Vet Med Sci 2012; 74: 1539–1544.
43. Dunbar MR, MacCarthy KA. Use of infrared thermography to detect signs of rabies infection in raccoons (Procyon lotor). J Zoo Wildl Med 2006; 37: 518–523.
44. Joffe DJ. Feline hyperthyroidism: a case report and review of the literature. Can Vet J 1986; 27: 125–130.
45. Kidd A, Okita N, Row VV, et al. Immunologic aspects of Grave's and Hashimoto's diseases. Metabolism 1980; 29: 809–826.
46. Wakeling J, Elliott J, Syme H. Evaluation of predictors for the diagnosis of hyperthyroidism in cats. J Vet Intern Med 2011; 25: 1057–1065.
47. Feldman EC, Nelson RW. Feline hyperthyroidism (thyrotoxicosis). In: Feldman EC, Nelson RW, eds. Canine and feline endocrinology and reproduction. 3rd ed. St Louis: Saunders, 2004; 152–218.
48. Padgett S. Feline thyroid surgery. Vet Clin North Am Small Anim Pract 2002; 32: 851–859.
49. Flanders JA. Surgical options for the treatment of hyperthyroidism in the cat. J Feline Med Surg 1999; 1: 127–134.
50. Harvey AM, Hibbert A, Barrett EL, et al. Scintigraphic findings in 120 hyperthyroid cats. J Feline Med Surg 2009; 11: 96–106.
51. Hibbert A, Gruffydd-Jones T, Barrett EL, et al. Feline thyroid carcinoma: diagnosis and response to high-dose radioactive iodine treatment. J Feline Med Surg 2009; 11: 116–124.
52. Peterson ME. Feline hyperthyroidism. Vet Clin North Am Small Anim Pract 1984; 14: 809–826.
53. van Hoek I, Lefebvre HP, Peremans K, et al. Short- and long-term follow-up of glomerular and tubular renal markers of kidney function in hyperthyroid cats after treatment with radioiodine. Domest Anim Endocrinol 2009; 36: 45–56.
54. van Hoek I, Meyer E, Duchateau L, et al. Retinol-binding protein in serum and urine of hyperthyroid cats before and after treatment with radioiodine. J Vet Intern Med 2009; 23: 1031–1037.
55. Naan EC, Kirpensteijn J, Kooistra HS, et al. Results of thyroidectomy in 101 cats with hyperthyroidism. Vet Surg 2006; 35: 287–293.
56. Peterson ME, Kintzer PP, Cavanagh PG, et al. Feline hyperthyroidism: pretreatment clinical and laboratory evaluation of 131 cases. J Am Vet Med Assoc 1983; 183: 103–110.
57. Werner SC, Ingbar SH. The thyroid: a fundamental and clinical text. 10th ed. Philadelphia: Lippincott Williams and Wilkins, 2013; 354–355.
58. Vanderpump MP. The epidemiology of thyroid disease. Br Med Bull 2011; 99: 39–51.
59. Bjoro T, Holmen J, Kruger O, et al. Prevalence of thyroid disease, thyroid dysfunction and thyroid peroxidase antibodies in a large, unselected population. The Health Study of Nord-Trondelag (HUNT). Eur J Endocrinol 2000; 143: 639–647.
60. Norsworthy GD, Adams VJ, McElhaney MR, et al. Palpable thyroid and parathyroid nodules in asymptomatic cats. J Feline Med Surg 2002; 4: 145–151.
61. Peterson SL, Yoshioka MM. The use of technetium 99m pertechnetate for thyroid imaging in a case of feline hyperthyroidism. J Am Anim Hosp Assoc 1983; 19: 1015–1017.
62. Helmy A, Holdmann M, Rizkalla M. Application of thermography for non-invasive diagnosis of thyroid gland disease. IEEE Trans Biomed Eng 2008; 55: 1168–1175.
63. Jin C, He ZZ, Yang Y, et al. MRI-based three dimensional thermal physiological characterization of thyroid gland of human body. Med Eng Phys 2014; 36: 16–25.
64. Umbaugh SE, Solt P. Veterinary thermographic image analysis: data and temperature normalization. Report No. 4878–3. Edwardsville, Ill: Southern Illinois University Computer Vision and Image Processing Laboratory, 2008.
65. Shatsova EN, Popov VA. The possibilities of infrared thermography in the diagnosis of thyroid pathology [in Russian]. Probl Endokrinol (Mosk) 1989; 35: 34–37.
66. Graves TK, Olivier NB, Nachreiner RF, et al. Changes in renal function associated with treatment of hyperthyroidism in cats. Am J Vet Res 1994; 55: 1745–1749.
67. Bond BR, Fox PR, Peterson ME, et al. Echocardiographic findings in 103 cats with hyperthyroidism. J Am Vet Med Assoc 1988; 192: 1546–1549.
68. Kobayashi DL, Peterson ME, Graves TK, et al. Hypertension in cats with chronic renal failure or hyperthyroidism. J Vet Intern Med 1990; 4: 58–62.
69. Clark JA, Cena K. The potential of infra-red thermography in veterinary diagnosis. Vet Rec 1977; 100: 402–404.
70. Peterson ME, Randolph JF, Mooney CT. Endocrine diseases. In: Sherding RG, ed. The cat: diagnosis and clinical management. New York: Churchill Livingstone, 1994; 1404–1506.
71. Peterson ME. Hyperthyroidism. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine: diseases of the dog and cat. 5th ed. Philadelphia: WB Saunders Co, 2000; 1400–1419.
72. Mooney CT, Peterson ME. Feline hyperthyroidism. In: Mooney CT, Peterson ME, eds. Manual of endocrinology. 3rd ed. Gloucester, England: British Small Animal Veterinary Association, 2004; 95–111.
