Nanomedicine, Volume I: Basic Capabilities

© 1999 Robert A. Freitas Jr. All Rights Reserved.

Robert A. Freitas Jr., Nanomedicine, Volume I: Basic Capabilities, Landes Bioscience, Georgetown, TX, 1999


 

8.4.1 Thermographic Navigation

8.4.1.1 Thermography of the Human Body

The human body presents a complex and temporally varying spatial temperature field. The external parts of the body have a lower mean temperature than the internal parts, with temperature decreasing along the longitudinal axis of the extremities, producing both axial and radial temperature gradients. The differing heat production of individual organs, geometric irregularities, changes in insulation and evaporation, convective heat transport via the blood, and the diurnal and other periodic variations3327 add further complexities to the thermal map.890

The homeothermic core of the body is distinguishable from the shell, which most readily responds to environmental fluctuations. The core generally consists of the interior of the thorax and abdomen, the brain, and part of the skeletal muscles. With moderate changes in ambient temperature, the shell normally comprises the outermost 20%-35% of the human body.894,895 However, during extreme chilling, the shell may enlarge to ~50% of total body volume, equivalent to a mean layer thickness of 2.5 cm.890 Figure 8.28 shows the overall distribution of tissue temperatures as a series of isotherms.

Skin temperatures display the greatest thermographic variability in response to external factors. For example, nude humans standing for 3 hours in a cold room (5C, 50% relative humidity, 0.1-0.2 m/sec wind speed) experience skin temperature differentials up to 15C (= 13C to 28C), with the lowest temperatures in fingers and toes, the highest in trunk and forehead, and average core/surface gradient ~15C;896 heat loss is ~10% lower for females due to their thicker layer of subcutaneous fat, making their cold-room skin temperature slightly lower than for males.898 After 3 hours in a hot room (50C), skin temperature differentials amounted to only 2.5C (= 35C to 37.5C), with an average core/surface gradient of ~1C.896 With normal clothing in a room at 15-20C, mean skin temperature is 32-35C.

Skin thermography of the human head was first reported by Edwards and Burton;897 skin vs. rectal temperatures at various ambient temperatures are well-studied.893,898 Skin temperature patterns in neonates reflect near uniform heat conduction through the tissues.917 In childhood, specific patterns develop into a stable, permanent adult pattern.916 The dermothermal patterns differ in lean and obese patients, and exhibit a continual state of small rhythmic change. These changes -- probably a result of active vasodilation due to sympathetic innervation over most of the human skin area919 -- are observed over the arms, hands, trunk and head, but are not all in phase with each other, nor even of the same amplitude.918,919,3334-3338 Skin thermology, including infrared thermography, is now an important branch of medical diagnostic imaging.899,912 Subcutaneous (shell) temperatures generally increase with depth.

Human core (rectal) temperature averages 37.0C,894 but this simple number hides considerable natural variation (Table 8.11). The temperatures of inner organs vary by 0.2-1.2C under normal room conditions, and by up to 0.9C within individual organs.890 Temperature gradients within the brain amount to 1.4C; the cortex is cooler than the basal regions, with incoming blood cooler than the central brain tissue.900 Brain temperature also decreases during sleep and rises during periods of emotional arousal.901 Cooling or warming the skin of the head causes temperature changes in the tympanic membrane (Fig. 7.3) of up to 0.4C due to the returning venous blood.909 Oral temperature can have wide variation, depending for example on the thermal character of recently ingested food and drink, and thus has little direct correlation with swings in rectal temperature.910 Liver temperatures lie 0.2-0.6C subrectal,902 possibly due to heat loss via the respiratory tract through the diaphragm. The temperature of the testes in the scrotum is ~3C lower than that of the abdominal cavity.890 The temperature of the hypothalamus, the seat of human thermoregulation, might be the most logical choice as a reference value.890

Human blood temperature also displays considerable variation. Outflowing blood is warmer than inflowing blood in organs with high metabolic rates, so the blood has a cooling effect in these organs, the opposite of the skin. Venous blood temperature in the extremities may be ~3C lower than the arterial supply, but even the arterial blood can sink to 22C in human extremities at cold room temperatures.903 Blood in the common carotid artery remains within 0.2C of oral temperature, but falls 0.2-0.5C as it cools the face and possibly also due to cooling from the internal jugular vein.905 However, inhalation of cold air has little effect on pulmonary blood temperature; even for patients in rooms cooled to 18C, blood in the pulmonary vein and artery differ at most by 0.03C.906

There are also periodic variations in body temperature. The diurnal fluctuation is the best known. This cycle is absent in neonates, but develops during the first weeks of life.913 Children have higher daytime temperatures than adults and a daily range of 1.7C,890 with the daily range peaking near the age of 1 year.894 Precisely standardized tests on young adults found that rectal core temperature fluctuated between 36.83-38.32C in males (range 1.49C) and 37.16-38.36C (range 1.20C) in females,908 with temperatures normally reaching their lows near 6 AM local time and their highs near 6 PM.915 (The timing is keyed to environmental influences such as light/dark cycles, not personal activity.) The menstrual cycle also involves a complex long period fluctuation. During this monthly cycle, morning core temperature falls from 37.0C to 36.7C just before and during menstruation, after which there is an abrupt dip to ~36C at mid-period near ovulation, followed by a return to normal levels for the remaining ~2 weeks. Menstrual-related skin temperature pattern variations are also observed.3329

Finally, there are various irregular influences on body temperature. Rectal temperature may rise up to 3.5C during the most extreme exercise,821 or may fall 2C after ~1 hour immersion in water at 23C. Ingesting food elevates skin temperature, and to a lesser extent the rectal temperature, for 1-2 hours. Alcohol ingestion raises skin temperature but lowers core temperature, while nicotine ingestion (e.g., smoking) lowers skin temperature in the extremities by 2-7C. In persons with fever, the temperature may rise from 37.0C up to 41.5C,916 with daily fluctuations of 2.8-5.0C, usually peaking in late afternoon. Transient elevations to 44-45C have been recorded but are very rare.

In view of this measurable spatial and temporal variability, maintaining a high-resolution whole-body thermal map in real-time is impractical using in vivo nanodevices. Given the availability of ~microkelvin sensors with ~microsecond measurement cycles (Section 4.6.1), sensor technology is not the limiting factor. Map stability is the main concern. A volumetric map with only ~1 mm3 resolution, comparable to the best thermographic images, requires 108 voxels and 17 bits/voxel for 0.001C accuracy. A hand placed on a stove produces thermal gradients >10C/sec in the shell, but periodic temporal variations are 0.1-40 microkelvins/sec in the core and 0.0051C/sec in the shell during changes in physical exercise levels (~3C in 10-100 sec), movement between hot and cold environments (~10C in 10-100 sec), and fevers (~5C in ~1000 sec). Thus to maintain a map that reports all changes of >~1 millikelvin requires the entire field to be resampled every 0.001C / (1C/sec) = 1 millisec, generating ~2 terabits/sec. This dataflow possibly could be accommodated by a dedicated high-capacity fiber network (Section 7.3.1) but would rapidly overwhelm both a mobile acoustic network and the data storage capacity of individual bloodborne nanorobots. Maintaining a hematothermographic map employing capillary-volume resolution to millikelvin accuracy generates an even more intractible >1015 bits/sec.

 


Last updated on 20 February 2003