Estimation of From

Body Sites Temperatures Tympanic Measurements Jacob Fraden, PhD*, Robert P.

Lackey, MSEE**

For many years, measurement of body temperature in routine medical practice was limited to oral, rectal and axillary sites. Recent introduction of infrared non-contact thermometers for the auditory canal requires the establishing of temperature relationships between the ear and more traditional thermometry sites. Since an auditory canal is exposed to the environment, the infrared readings from it are influenced by ambient temperature. A linear model of thermal gradients in the vicinity of an ear canal allows us to find simple formulas connecting temperatures taken from the ear with those from traditional core sites like bladder or pulmonary artery, in addition to rectal and oral. The formulas contain environment coupling coefficients. Their values have been found experimentally by measuring body temperatures from subjects in a walk-in environmental chamber and from multiple clinical studies. The derived coefficients are used in the Thermoscan™ PRO-1 Instant Thermometer to calculate core, oral and rectal equivalent temperatures.

Introduction Due to new technological developments during recent years, the infrared (IR) thermometers have gained a substantial interest in the medical community. Even more interest was generated when the PRO-1 Instant Thermometer was introduced in 1990 by Thermoscan@ Inc. (San Diego). Just a few years ago, many medical professionals did not consider ear canal thermometry as a viable alternative to oral and rectal readings. There remain some questions concerning the accuracy and limitations of these new instruments. Among them is the question of relationships between the ear temperature and those of the traditional thermometry body sites. *Vice President of Research, and **Vice President of R&D, Thermoscan® Inc. Correspondence to Dr. Fraden, V.P. Research, Thermoscan Inc., 6295 Ferris Sq., Suite G, San Diego, CA 92121

Indeed, an auditory canal is a near ideal cavity for body temperature measurement. It is not directly affected by

respiration, food, drink or smoking. To some degree it is affected by physical activities and emotional state. Anatomically, the auditory canal is a slightly curved tube about 3.5 cm long in an adult. It is exposed to the outside through the helix (auricle) and limited inside by the tympanic membrane (Fig. 1 A). The canal is well-insulated from the exterior. It is located in close proximity to the major brain arteries and veins, and ends only about 3.5 cm from the hypothalamus. It was recognized about 20 years ago that the tympanic membrane temperature is the most accurate approximation of core temperature. 1,2 Further, the interior of an auditory canal forms a cavity which has properties of a blackbody, an ideal emitter of electromagnetic radiation. An IR thermometer measures natural electromagnetic radiation which is emanated from the skin surfaces Often, such a thermometer is called &dquo;tympanic,&dquo; which is not 65

As in any other part of the body, the ear canal temperadepends on local metabolism, temperature of arterial blood supply and heat flow to the neighboring tissues. Blood supply of a distal portion of an auditory canal is provided by several branches of the external carotid and one branch of the internal carotid arteries. One end of an auditory canal is exposed to the outside atmosphere. The open portion of the canal is subjected to air movement around the patient’s head. Moving air causes heat loss from the skin surface. Temperature of air within the ear canal may be as low as 3°C below esophageal (core). ’I In addition, depending on the outside temperature, five to nine mW of thermal power is radiated to the environment from every square centimeter of skin surface near the ear canal opening. Therefore, we may speak about an ear canal-to-ambient temperature relationship. Several studies have been conducted on the subject. Nadel et al.’ have found that tympanic temperature changes by 0.04°C per 1 °C ambient. A much smaller number was reported by Brinnel et aI.3: 0.0025°C/°C for the cases when temperature was measured at the lower portion of the tympanic membrane with more intense blood supply from the internal carotid artery. This portion of the tympanic membrane is also the warmest portion of the ear canal. These studies clearly suggest that temperatures are quite non-uniform within the ear canal and any averaged readings taken by an infrared thermometer will be generally lower than those taken directly from the brain blood supply. An average temperature of the ear canal is also a function of local metabolic processes (which, in turn, may depend on such factors as age, emotional state, time of the day) and those of the neighboring tissues, namely of the skin around the external ear opening. As a general rule, the deepest portion of the ear canal is the most stable and more closely resembles temperatures of the tympanic membrane and that of the core. The larger the area of an auditory canal that is included in the thermometer’s field of view, the more influence of outside air temperature one may expect. It is appropriate to study the influence of ambient temperatures and, if possible, to introduce some type of a correction in temperature calculation to enhance accuracy of the infrared thermometers. Generally, core temperature (or its customary representatives, oral and rectal) is of interest to a medical provider. Knowing the ear canal temperature, however accurate it might be, is not yet generally acceptable for diagnostic and monitoring purposes. Therefore, a goal of the study was to find how the oral and rectal temperature readings are correlated with those taken from the auditory ture

FIG. 1 (A). Temperature measurement by infrared thermometer in the auditory canal. Infrared radiation (IR) collected from the field of view is channeled by the wave guide to the thermal radiation sensor.

quite correct. In practice, all IR thermometers have relatively wide angles of view (30° to 90°) and a limited depth of insertion into the auditory canal (Fig. 1 A). This results in taking average temperature from a relatively large area which includes the walls of the auditory canal and a portion of the tympanic membrane as it is depicted in Fig. 1B.

FIG.

66

1 (B).

Field of view

as seen

through

the

wave

guide.

canal by a PRO-1 infrared thermometer (Thermoscan) in order to estimate the oral and rectal &dquo;equivalents&dquo; from the ear

temperature

measurements.

Model Since an auditory canal has one end &dquo;immersed&dquo; into the body and the other end exposed to the outside, there may be a quite non-uniform temperature distribution along its surface. To better analyze this distribution, we have to construct a thermal model. It should include a core, a tympanic membrane, an interior of the canal and an outside temperature. Three assumptions have been made: 1. Temperatures of interest (like core or tympanic) are considered to be localized in specific points rather than being distributed over a surface. 2. Thermal resistances between specific points are linear. Such a resistance represents a combined effect of heat movement resulting from blood circulation, heat conduction, air convention and infrared radiation. 3. The core, mouth and rectum are thermally disconnected (thermal resistances between them are infinitely large), however they are under common physiological control. In the earlier stage of the work, we assumed that oral and rectal temperatures differ from core by constant offsets. The experiments have demonstrated that such an assumption was not correct. Later, it was found that all temperatures of body interior are functions of ambient conditions and that the thermal model should be built in a more general configuration. We are going to build that model for core temperature in the form of an equivalent circuit for heat flow between the core, the tympanic membrane, the ear canal, the outer ear and its surroundings. Five temperatures are of interest to the analysis: 1) core temperature (T ) is that of blood taken from a major artery (pulmonary or carotid) or in the vicinity of hypothalamus; 2) average temperature of the tympanic membrane (Tm), which is close but generally not exactly equal to that of core (T ); 3) the average temperature (Tb) of the auditory canal interior covered by the field of view of the infrared thermometer, including some portion of the tympanic membrane (see Fig. 1B). This is the temperature which is actually measured by the IR thermometer; 4) temperature (Te) of the outer ear which is directly exposed to the environment; 5) ambient temperature (Ta) which can be easily measured by an external thermometer. Since the above named temperatures are not equal to one another, thermal gradients exist between the specific points of the system. These gradients can be characterized by

FIG. 2.

Equivalent circuit of thermal gradients in the auditory canal.

Tb and T. are the only temperatures which can be directly measured. Thermal resistances F~and Rm are small while Rband R. are c

m

a

relatively high.

thermal resistances which determine heat flow between the points. The thermal equivalent circuit (Fig. 2) consists of serially connected hard temperature sources and thermal resistances. The word hard indicates that the source is

capable of generating a specific temperature regardless of any change in the values of the model elements. However, temperatures of the hard

sources

may vary because of

influences, such as changing environmental conditions, physiological control, etc. There are two hard external sources

in the model: ambient temperature

Ta

and

core

temperature T~. The tympanic membrane is located at the further end of auditory canal and is well-insulated from the environment. Yet, it is somewhat exposed to the outside and may lose small amounts of heat. Therefore, a thermal resistance

the

Roc between the point which we call core and the tympanic membrane should be placed in the circuit diagram in series with Tc. An auditory canal interacts with the environment even more than the tympanic membrane. This results in a thermal resistance R . The closer it is to the outside, the higher the temperature gradient and the higher the temperature drop. Helix temperature (near the auditory canal opening) is represented by T,. The thermal resistance from the auditory canal to its opening is designated as Rb. This resistance greatly depends on a capillary blood flow which, in turn, is a function of such variables as age, emotional state and environmental conditions. The last resistance in the model is R which determines loss of heat to the environment. It is a function of the ear shape, size and such factors as ear m

a

coverage

by cloths, hairs, etc.

Applying the first Kirchoff’s law to the equivalent circuit of Fig. 2, we can calculate some temperatures which 67

cannot be

derived

directly measured.

The core temperature can be

On the other hand, 6 from equation (3):

as:

&dquo; ~ RR +R~ +R

To = Ta T T + (Tb (Tb - Ta) =

e

a

+

m

11 +

The resistive ratio in the parenthesis represents

experimentally

f3 - T~Tcb .TaT~b c

(1 )>

determined

(~)

Similarly, the coupling coefficients for the oral and rectal sites may be derived from the following equations:

a

ronment-to-core

be

can

an

envi-

coupling coefficient:

To

Tb

b

a

Q

(5)

~r = Tr _ Tb Tb Ta

(6)

13 B0 0

=

R +R

(3~ =

R° a + Rm b

(~>z

r

a

The numerator is a thermal resistance from the ear canal to the core and the denominator is a thermal resistance from

the

canal to the environment. The coefficient 13 strongly depends on the ear anatomy and a capillary blood circulation. Subsequently, it could be dependent on factors which affect peripheral vascular resistance in the external ear. It also depends on the thermometer’s field of view and depth of insertion. Substituting (2) into (1) we receive the core equation: ear

Tc = Tb(1 + Bc) - Taf3~ = Tb + f3~(Tb-Ta)

(3)

The above equation indicates that the core temperature be derived from ear temperature by adding to it an offset A =13~(Tb Ta) whose value is a function of a difference between the ear and ambient temperatures. The hypothesis which was tested below is that the equivalent circuits for the oral To and rectal T temperatures are analogous to that of the core with the corresponding substitutions of To or Tr for Tc. Thus, the coupling coefficients for the oral and rectal estimates (13o and 6) may be determined by the equations similar to (3).

can

Coupling Coefficients constant for a particular state of

Determination of Site Coefficient

6

is

a

particular patient but may somewhat vary from person to person and in a person during the day. The coefficients, if averaged over the sufficiently representative population, can be put to use in the practical IR thermometers to estimate temperatures in the body locations which are not easily accessible for direct measurements.~ Thus, knowing the appropriate coupling coefficients 13, by measuring temperature of the ear canal, one could predict what oral, rectal and

core

contact thermometers would read.

inconvenient for the B calculation as thermal resistances are very difficult to measure in vivo.

Equation (2) is

68

The above-mentioned assumption #2 demands that the model is linear, hence B should be independent of both body and ambient temperatures. To prove that assumption, the stability of 13 under varying ambient temperatures should be verified. The experiment was set in the State University of New York Health Science Center in Syracuse and is covered in detail elsewhere.15 Briefly, a group of 21 volunteers was subjected to three ambient temperatures (18°C, 24°C and 35°C) in a walk-in environmental chamber. The oral, rectal and ear temperatures were measured and 13 was calculated for each subject. On the average, at 35°C, the differences between both oral, rectal and ear readings become smaller (respectively 0.5’ and 0.8°C versus 1.0° and 1.6°C of control). This was predicted by the linear model. The experiment has resulted in the following coupling coefficients: 6~=0.047 and B 0 =0.017. Variations in 13 at different ambient temperatures were within the measurement error range and generally didn’t exceed +/- 8%. Thiss proves that the above described linear model may be accepted for practical purposes. The above experiment was designed to prove the model. However, the actual numbers for the coupling coefficients must be determined for the more representative groups of patients. r

Results and The

Analysis

differences between body sites and ear temas taken by the Thermoscan® PRO-1 Instant peratures Thermometers were used to calculate 13 with equations (4)(6). The basis for the calculations were the experimental data collected by several independent research groups whose reports are presented at this symposium.14.16-19 The mean differences between the site temperatures are summarized in Table 1. It is apparent that a difference between a selected body site and the ear canal is specific for a mean

(°C) between body sites and ear temperatures as measured by various researchers. temperature 23°C. Age groups: &dquo;adults&dquo; (3 years and older), &dquo;infants&dquo; (under 3 years).

Table 1 Mean differences

Average

room

selected group of patients. Unfortunately, B can’t be averaged over a general population as variations are too large. Presently, oral temperatures are routinely taken from the cooperative patients who generally include children over three years of age and adults. Rectal readings are taken from infants and the uncooperative adults while core are not routine measurements in any age Sometimes they are used in the operating and recovery rooms and ICUs. It is important to note that invasive temperature readings from bladders and pulmonary arteries were considered core and taken in the OR and

temperatures group.

settings. The specific conditions of the patients, use of anesthetics, vasodilators, etc. may noticeably affect the results. That limits use of the core estimates for the broader ICU

population whose conditions are different from the experiment.

To define 13 for practical purposes, we have established the following criteria on the analysis of published data: 1,2,6-11 1). The Oral Equivalents are to be derived from the experimental data taken from children over three years

of age and adults of any age.

2). Rectal Equivalents are to be established from the experimental data taken from children under three years of age. 3). Core Equivalents are to be established as an average of the experimental data taken in the clinical settings from bladder, pulmonary artery and rectally from adults. 4). Temperatures of newboms are to be determined in the Equal Mode (actual ear canal temperature Tb) as the overall body temperature is quite homogeneous for those subjects and mean ear and axillary temperatures are proven to equal one another (see Table 1). Table 2 summarizes the selected values of 13 and, as an example, also gives offsets A° (site-ear) which are calculated by the PRO-1 Instant Thermometer at room temperature. Fig. 3 shows how three offsets change with ambient

temperatures. Table 2

* Average values of B and corresponding offsets A° at

room

temperature (23°C)

FIG. 3. Offsets as functions of ambient temperatures. Calculated for the ear canal temperature of 37°C.

69

In conclusion, we want to stress that the calculated offsets are based on the statistical data analysis. The oral, core or rectal equivalents are estimated transpositions of the ear reading onto the corresponding body sites. The estimates which are made by the ear thermometer may be more accurate than the actual readings from the sites as representatives of true core body temperature, especially in dynamic situations where the patient temperature is

changing. These sites (primarily rectal) are subject to physiological dynamics and their temperatures may lag significantly from those taken from the ear canal. 1,2 The ear is the best orifice for the body temperature determination and we believe that it will be used future.

as a

standard in the

Acknowledgement Authors want to thank Thermoscan@ Inc. personnel for lending the body sites and John Hyle whose suggestions were

always stimulating.

References 1.

Benzinger M, Tympanic clinical temperature. Fifth symposium on temperature, Washington, D.C. 1971; 2089-102.

2.

Benzinger M. Tympanic thermometry in surgery and anesthesia. JAMA 1969; 209:1207-11. Brinnel H, Cabanac M. Tympanic temperature is acore temperature in humans. J Therm Biol 1989; 14:47-53. Fraden J, Lackey RP. Corrective measurement system for a medical thermometer, 1990. U.S. Patent Appl. Fraden J. Non-contact temperature measurements in medicine. In: Bioinstrumentation and biosensors. Marcel Dekker Inc., 1991; 511-49. Nadel ER, Horvath SM. Comparison of tympanic temperature and deep body temperatures in man. Life Sci 1970; 9:869-75. Rhoades F, Grandner J. Assessment of an aural infrared sensor for body temperature measurement in children. Clin Pediatr 1989; 29:112-5. Shinozaki T, Deane R, Perkins F. Infrared tympanic thermometer: evaluation of a new clinical thermometer. Critical Care Med 1988; 16:148-50. Terndrup T, Allegra J, Kealy J. A comparison of oral, rectal and tympanic membrane-derived temperature changes after ingestion of liquid and smoking. Am J Emerg Med 1989; 7:150-4. Wilson R, Knapp C, Traber D, et al. Tympanic thermometry: a clinical evaluation of a new technique. Southern Med J 1971; 61:1452-5. Benzinger T. Cranial measurements of internal temperature in man. In: J Hardy, (ed) Temperature. Reinhold Publishing Corp. 1963; 3:111-20. Talo H, Macknin ML, S VanderBrug Medendorp. Tympanic membrane temperatures compared to rectal and oral temperatures. Clin Ped 1991; suppl: 30-33. Sharkey A, Elliott P, Lipton JM, et al. The temperature of the air within the external auditory meatus compared with esophageal temperature during anesthesia and surgery. J Therm Biol 1987; 12:11-3.

3. 4. 5.

6. 7.

8.

9.

10.

11.

12.

13.

70

14. Milewski A, Ferguson KL, Temdrup TE. Comparison of pulmonary artery, rectal, and tympanic membrane temperatures in adult intensive care unit patients. Clin Ped 1991; suppl: 13-16. 15. Zehner J, Temdrup TE. The impact of moderate ambient temperature variance on the relationship between oral, rectal and tympanic membrane temperatures. Clin Ped 1991; suppl: 61-64. 16. Shenep JL, Adair JR, Hughes WT, et al. Infrared, thermistor and glass-mercury thermometry for measurement of body temperature in children with cancer. Clin Ped 1991; suppl: 36-41. 17. Kelly B, Alexander D. Effect of Otitis media on infrared tympanic thermometry. Clin Ped 1991; suppl: 46-48. 18. Chamberlain JM, Grander J, Rubinoff JL, et al. Comparison of a tympanic thermometer to rectal and oral thermometers in a pediatric emergency department. Clin Ped 1991; suppl: 24-29.

Estimation of body sites temperatures from tympanic measurements.

For many years, measurement of body temperature in routine medical practice was limited to oral, rectal and axillary sites. Recent introduction of inf...
481KB Sizes 0 Downloads 0 Views