A simple method for measuring calorie expenditures during sleep

Summary A simple method for the continuous determination of O₂ consumption and CO₂ production in sleeping subjects is described. Face masks and moth pieces have been eliminated and the method has been successful with babies and adults. The apparatus consists of a ventilated hood connected to a vacuum cleaner. The subject sleeps with his head on a pillow inside the hood and fresh air is sucked over his face. Diluted expired air leaves the hood through an outlet in the rear wall and its volume is measured with a gas meter. A side-arm fitted to the outlet enables continuous samples of diluted expired air to be diverted by an aquarium pump into a sampling bag. Safety circuits are incorporated into the hood.     

Direct determination of local oxygen consumption of the brain cortex in vivo

Summary A method is described to determine local oxygen consumption quantitatively in the brain cortex under in vivo conditions. Local oxygen consumption is calculated from the slope of local tissue P_{O 2} decrease during a few seconds of total ischemia of the brain for each second after the stop of circulation. The decrease of tissue P_{O 2} is recorded simultaneously at several measuring sites. To be independent of oxygen chemically bound to hemoglobin, tissue P_{O 2} values are raised above 100 Torr. The calculation of local oxygen consumption for each second during the short period of ischemia showed that the O₂ consumption remains constant only for a few seconds ranging from 5 to maximally 15 s at different locations. Then O₂ consumption decreases continuously although the tissue P_{O 2} values are still above the full saturation of hemoglobin. The rate of local oxygen consumption varies considerably at different measuring sites of the superficial layers of the brain cortex (cat). The mean value amounts to 3±1.5 ml O₂/100 g tissue and minute.     

A Mathematical Model to Predict CO2 Removal in Hollow Fiber Membrane Oxygenators

A mathematical model has been developed to predict CO₂ removal in hollow fiber membrane oxygenators. The model is analogous to one developed previously for predicting O₂ transfer. A mass transfer correlation was determined in water for O₂ and CO₂ exchange and collapsed onto one universal curve. The correlation was used to predict CO₂ removal in blood by incorporating a ‘facilitated diffusivity’ to account for the transport of CO₂ present as bicarbonate. The diffusion of bicarbonate greatly increased the ability of the oxygenator to remove CO₂ in blood compared to water. A fiber bundle module was fabricated to test the model predictions. The fiber bundle had a length of 13 cm and a bundle thickness of 0.2 cm. The module was tested in bovine blood at flowrates of 0.75, 1.5, and 2.2 L/min and CO₂ removal rate predictions were within 9% of experimental measurements at all flowrates. The O₂ transfer rate predictions were within 10% of experimental measurements. A second module was manufactured with a bundle of length 4 cm and thickness of 1 cm. The CO₂ removal predictions were within the standard deviation of the experimental measurements.     

The role of fitness on VO2 and VCO2 kinetics in response to proportional step increases in work rate

Summary The purpose of this study was to determine the effect of fitness and work level on the O₂ uptake and CO₂ output kinetics when the increase in work rate step is adjusted to the subject's maximum work capacity. Nine normal male subjects performed progressive incremental cycle ergometer exercise tests in 3-min steps to their maximum tolerance. The work rate step size was selected so that the symptom-limited maximum work rate would be reached in four steps at 12 min in all subjects. Oxygen consumption (VCO₂) and carbon dioxide production VCO₂ were calculated breath by breath. For the group, the time (mean, SEM) to reach 75% of the 3-min response (T _0.75) for VO₂ increased significantly (P<0.01) at progressively higher work rate steps, being 53.3 (5.5) s, 63.5 (4.6) s, 79.5 (5.0) s, and 94.5 (5.8) s, respectively. In contrast, T _0.75 for VCO₂ did not change significantly [74.9 (7.4) s,. 75.6 (5.0) s, 85.1 (5.3) s, and 89.4 (6.3) s, respectively]. VCO₂ kinetics were slower than VO₂ kinetics at the low fractions of the subjects' work capacities but were the same of faster at the high fractions because of the slowing of VO₂ kinetics. The first step showed the fastest rise in VO₂. While VO₂ kinetics slowed at each step, they were faster at each fraction of the work capacity in the fitter subjects. The step pattern in VO₂ disappeared at high work rates for the less fit subjects. The heart rate response paralleled that of VO₂. We conclude that VO₂ and VCO₂ kinetics are slower in the less fit subjects but only VO₂ kinetics are significantly attenuated in response to proportional step increases in work rate.     

Pancreatic O2 consumption and CO2 output during secretin-induced,exocrine secretion from the pancreas in the anesthetized dog

Secretin stimulates pancreatic water and CO₂ excretion as well as pancreatic blood flow. It has been questioned whether the production (i.e. water and CO₂ excretion) is reflected in the input-output difference of nutrients. In pentobarbital anesthetised dogs, pancreatic exocrine secretion was stimulated by secretin, (Karolinska), 1 U/kg injected as an i.v. bolus. Secretion was maximally increased at 2 min after the secretin shot and returned to a basal value at between 16 and 32 min after secretin. Blood flow was also maximally increased at 2 min, but decreased to the basal value at between 8 and 16 min. O₂ extraction first decreased (at 2 min) and then gradually increased until it was higher than the basal value (at 16 min) and then returned to the basal level (at 32 min). O₂ consumption increased quickly, reached a plateau, lasting from 1 to 16 min, and then decreased to the basal level (32 min). CO₂ transfer from blood to tissue reached a maximum at 4 min and then decreased to the basal value (at between 16 and 32 min). The curves for CO₂ transfer from tissue to pancreatic secretion and for CO₂ in the secretion had the same shape. It is concluded that the curve of production (of water and CO₂ excretion) parallels the curve of O₂ consumption fairly well. The O₂ consumption curve did not correlate either with the blood flow curve or with the O₂ extraction curve. About one quarter of the excreted CO₂ originated from pancreatic metabolism and the remaining three quarters were transferred from blood, through the pancreatic tissue into the secretion. The increase in O₂ consumption was achieved by an increase in blood flow, followed by an increase in O₂ extraction. The release of a vasodilator metabolite by the pancreatic cells upon arrival of the secretin molecules, may explain both the increase in blood flow and the successive increase in O₂ extraction. Therefore these data can be interpreted according to the model for metabolic control of tissue oxygenation.     

Massenspektrometrische Bestimmung des O2- und CO2-Gehaltes von Blut

Zusammenfassung CO₂ und O₂ werden durch einen Trägergasstrom (N₂ reinst) aus dem Blut extrahiert und die Konzentration dieser Gase im konstanten Gasstrom fortlaufend massenspektrometrisch gemessen. Das Integral der Konzentration über der Zeit ist proportional der ausgetretenen Gasmenge. Die Eichung erfolgt durch Eingabe abgemessener Volumina von O₂ und CO₂ in den Trägergasstrom. Analysendauer: 2 min.     

The effects of CO2 and pH on the spontaneous activity of the taenia coli of guinea-pig

Summary Mechanical and electrical activity of the isolated guinea-pig taenia coli were recorded in bicarbonate-, Tris- or phosphate-buffered test solutions. In normal solution (pCO₂ 5%, pH 7.4), typical minute-rhythmical fluctuations of activity occurred, whereby activity was present for approximately 40% of the total time (active time), the other 60% being activity-free intervals. At constant extracellular pH, low CO₂ content increased active time, high CO₂ content reduced it. Extracellular alkalinization at constantpCO₂ also diminished active time, even in CO₂-free medium, whereas acidification raised it, sometimes causing continuous activity. At constant bicarbonate-buffer, changes of CO₂ content, i.e. accompanied by corresponding changes in pH, affected the active time much less than did CO₂ alterations in an isohydric medium. The test solutions had no major effect on frequency of synchronized spike discharges, in contrast to their actions on the minuterhythmical activity. Since CO₂ can pass through the cell membrane more easily than ions regulating intra- and extracellular pH, the observed effects are best explained by changes in the transmembrane pH-gradient. A drop in active time would be due to a relative intracellular acidification; continuous activity, on the other hand, due to an opposite change in transmembrane pH-gradient, i.e. a relative intracellular alkalinization.This work was supported by the Deutsche Forschungsgemeinschaft     

Transmural gradient of tissue gas tensions in the canine left ventricular myocardium during coronary clamping and reactive hyperemia

Mass spectrometry was used for the continuous, simultaneous and quantitative measurement of oxygen (PO₂) and carbon dioxide (PCO₂) partial pressures in the subendocardial and subepicardial layers of the left ventricle in 11 anaesthetized ventilated dogs. Under control conditions,PO₂ was significantly lower in the subendocardium (13.5±4.5 mm Hg) than in the subepicardium (20.7±2.3 mm Hg), whereasPCO₂ did not differ significantly (43±8.8 and 51±9.2 mm Hg respectively). These variables were not correlated with blood pressure or coronary blood flow. Subendocardial and subepicardialPO₂ decreased less than 5 s after coronary occlusion. These changes were more rapid and severe in the subendocardium. After occlusion for 90 s: subendocardialPO₂ was 4.1±6.3 mm Hg while subepicardialPO₂ was 6.7±15.0 mm Hg (P<0.05).PCO₂ reached peak values of 56±25 mm Hg subendocardial and 82±22 mm Hg subepicardial at 2.67±0.71 min and 3.43±0.93 min after coronary clamping. A reactive hyperemia occurred after coronary unclamping with different time courses and amplitudes for systolic and diastolic stroke flows whilePO₂ recovered with different kinetics. SubendocardialPO₂ increased with a lower initial slope, probably in relation with the delay in the diastolic hyperemia. The observed delayed subendocardial hyperoxia, unrelated to the hyperemia, may indicate a delay in the recovery of normal work and metabolism in the inner layers of the myocardium.     

Das Verhalten der endexspiratorischen Atemgasdrucke,der O2-Aufnahme und CO2-Abgabe nach einfacher Apnoe im Wasser,an Land und apnoeischem Tauchen

Zusammenfassung Um zu untersuchen, wie ein Tauch- oder Atemanhalte-manöver den Sauerstoffverbrauch und die CO₂-Abgabe des Menschen beeinflußt, hielten 6 männliche Versuchspersonen 30, 60, 90, 120 und 165 sec ruhig an der Wasseroberfläche und an Land liegend den Atem an. In einer Vergleichsserie tauchten sie in 80 cm Tiefe gleich lange.Nach der Apnoe wurden der endexspiratorischeP _{O 2} undP _{CO 2}, und die Sauerstoffaufnahme und die CO₂-Abgabe pro Atemzug mit Hilfe eines Massenspektrometers und eines Pneumotachographen ermittelt.Es zeigte sich, daß die Sauerstoffschuld, die während der Apnoe eingegangen wird, beim Atemanhalten im Wasser bis zu 29%, an Land bis zu 38% unter der O₂-Schuld lag, die zu erwarten wäre, wenn die gemessene Ruheaufnahme angehalten hätte. Beim Tauchen sank die O₂-Schuld bis etwa 28% unter die erwartete Schuld. Der endexspiratorischeP _{O a} in der ersten Exspiration fiel mit Zunahme der Apnoezeit ab, lag jedoch bei gleich langen Apnoezeiten nach Tauchen signifikant unter dem Wert nach Atemanhalten.Die CO₂-Abgabe nach der Apnoe entsprach bis zu einer Apnoezeit von 90 sec etwa der in der Apnoe gebildeten Menge. Bei längeren Apnoezeiten trat eine deutliche CO₂-Retention ein. Beim Atemanhalten wurde bis zu 60% weniger CO₂ abgegeben als zu erwarten war.Der endexspiratorischeP _{CO 2} im ersten Exspirationsgas lag unabhängig von der Apnoezeit ziemlich konstant bei 45 mm Hg, die CO₂-Abgabe in der ersten Exspiration konstant bei etwa 150 ml ohne wesentlichen Unterschied zwischen Tauchen und Atemanhalten.     

A vasopressin-induced decrease in pancreatic blood flow and in pancreatic exocrine secretion in the anesthetized dog

Vasopressin decreases blood flow as well as secretory flow in the pancreas. The question raised was whether the blood flow decrease is the determinant of the decrease in secretion or quite the reverse. In pentobarbital anesthetized dogs, secretory flow was first increased to a steady level by infusion of secretin. At this steady state, O₂ consumption and O₂ extraction were increased, while blood flow remained at the control level, indicating an increase in the area available for exchange i.e. an increase in capillary density. At increasing doses of vasopressin, secretory flow decreased, arterial flow decreased, and O₂ extraction increased, while O₂ consumption decreased and venous-arterial CO₂ concentration difference was not changed. At the same time CO₂ transport decreased, CO₂ concentration in the secretion was unchanged and CO₂ output in the secretion was decreased. The decrease in blood flow was always seen about 25 s before the decrease in secretory flow, strongly suggesting that the decrease in blood flow induced the decrease in secretory flow. A higher dose of vasopressin was required to decrease the O₂ consumption (i.e. this effect was less sensitive) than to increase O₂ extraction. The decrease in secretory flow and the decrease in blood flow showed an intermediate sensitivity. So O₂ consumption seems to be preserved at a high level by the increase in O₂ extraction. It is concluded that the vasopressin-induced decrease in blood flow is the determinant of the decrease in secretory flow. This phenomenon is discussed in terms of the model for metabolic control of tissue oxygenation.     

A method for estimating bicarbonate buffering of lactic acid during constant work rate exercise

A method to estimate the CO₂ derived from buffering lactic acid by HCO₃ ^– during constant work rate exercise is described. It utilizes the simultaneous continuous measurement of O₂ uptake ( O₂) and CO₂ output ( CO₂), and the muscle respiratory quotient (RQ_m). The CO₂ generated from aerobic metabolism of the contracting skeletal muscles was estimated from the product of the exercise-induced increase in O₂ and RQ_m calculated from gas exchange. By starting exercise from unloaded cycling, the increase in CO₂ stores, not accompanied by a simultaneous decrease in O₂ stores, was minimized. The total CO₂ and aerobic CO₂ outputs and, by difference, the millimoles (mmol) of lactate buffered by HCO₃ ^– (corrected for hyperventilation) were estimated. To test this method, ten normal subjects performed cycling exercise at each of two work rates for 6 min, one below the lactic acidosis threshold (LAT) (50 W for all subjects), and the other above the LAT, midway between LAT and peak O₂ [mean (SD), 144 (48) W]. Hyperventilation had a small effect on the calculation of mmol lactate buffered by HCO₃ ^– [6.5 (2.3)% at 6 min in four subjects who hyperventilated]. The mmol of buffer CO₂ at 6 min of exercise was highly correlated (r = 0.925, P < 0.001) with the increase in venous blood lactate sampled 2 min into recovery (coefficient of variation = ±0.9 mmol·l^–1). The reproducibility between tests done on different days was good. We conclude that the rate of release of CO₂2 from HCO₃ ^– can be estimated from the continuous analysis of simultaneously measured CO₂, O₂, and an estimate of muscle substrate.     

Ventilatory response to oscillating and non-oscillatingPa CO 2 in the anesthetized cat

Summary In 11 adult cats, lightly anesthetized with chloralose-urethane, blood from both common carotid arteries was led into a plastic chamber of 15–20 ml and returned to the carotids at a point 1.5 cm more cranial. By doing so arterial blood was assumed to pool within the chamber and lose itsP _{CO 2} oscillations which are normally known to exist as a result of the respiratory cycle. In control periods blood bypassed the chamber, thus maintaining respiratoryP _{CO 2} oscillations. Spontaneous ventilation was measured spirometrically. The animals were breathing pure O₂.Results. 1. When the sinus (carotid) nerves were intact or sectioned there was no significant difference in ventilation before or after switching from non-oscillating to oscillatingPa _{CO 2}. 2. When the vertebral arteries were ligated a drop in ventilation occurred after turning to oscillatingPa _{CO 2} which was followed by a slight rise above control values after 30–50 sec. This phenomenon was independent of sinus nerve integrity. Thus in hyperoxie condition the smallPa _{CO 2} oscillations known to occur in phase with respiration do not seem to provide a respiratory stimulus to resting ventilation above that generated by the mean level ofPa _{CO 2}. The ventilatory depression after vertebral artery ligation must at this time remain unexplained.     

Interpretation of gas-blood ratio inP O 2polarography

Summary The differences inP _{O 2}readings between gas and blood were studied with a Clark-type electrode in the range of 38.5 to 713 mm HgP _{O 2}.The tonometered blood samples were taken in two different ways. The results showed that the gas-blood ratior _b(equilibrating gasP _{O 2}reading/equilibrated bloodP _{O 2}reading) depended not only on the sampling method but also on theP _{O 2}range: it varied from 1.005 to 1.032 for aP _{O 2}of 96.5 mm Hg, and from 1.040 to 1.081 for aP _{O 2}of 713 mm Hg according to the sampling procedure.A theoretical analysis demonstrated that the variation ofr _bwith the bloodP _{O 2}can be attributed to the influence of the degree of oxygen saturation of the hemoglobin on theP _{O 2}gradient existing in the blood diffusion boundary layer adhering to the electrode membrane.This work was supported by grants from the High Authority of the European Coal and Steel Community and from the Fonds de la Recherche Scientifique Médicale, Belgium.     

CO2 responsivity in the mouse measured by rebreathing

We have modified the rebreathing method to study CO₂ responsivity in very small mammals. Tidal volume (V _T) and frequency (f) of pentobarbital-anesthetized mice were measured during rebreathing from a closed circuit, primed with 95% O₂, 5% CO₂, through which the gas was constantly circulated at 0.5 l·min^–1. The circuit consisted of T-tube from a plethysmograph, Tygon tubing with compliant element, CO₂ analyzer and pump, in series. CircuitPCO₂ (PctCO₂), which was recorded continuously during spontaneous breathing, rapidly equilibrated with end-tidalPCO₂. CO₂ response curves were constructed from extrapolated minute ventilation ( ),V _T,f and parameters of breath-to-breath timing, respectively, onPctCO₂. Analyses of slopes of the response curves, change from onset of rebreathing to peak response, andPctCO₂ at which the response peaked revealed that CO₂ stimulates by increasingf andV _T and that this is effected by facilitation of central inspiratory-expiratory phase switching and inspiratory drive mechanisms. However, the stimulatory effect of CO₂ on phase switching was not sustained, with maximal effect occurring before peak . The advantages and facility of the modified rebreathing method make it suitable for studies of other small mammals, including neonates.     

Transcutaneous,noninvasiveP O 2 monitoring in adults during exercise and hypoxemia

A new, commercially available, transcutaneous (tc)P _{O 2} monitor was tested in adult females and in laboratory animals to assess its applicability in measuring arterial oxygen tension during physiological stress. Observed values on dogs correlated well with direct measurements of arterialP _{O 2} and with previous data obtained from measurements of arterial blood during exercise and hypoxemia. In our female subjects the unit responded rapidly to changes in inspired ambient oxygen and electrical stability was excellent during maximal exercise tests. TranscutaneousP _{O 2} decreased to an average of 87.8 Torr during maximum exercise breathing 20.9% O₂, and to 32 Torr while breathing 12.6% O₂ at maximum work. Two distinct patterns of response in tcP _{O 2} were observed during hypoxic and normoxic exercise. The technique appears to have substantial future application both in clinical and physiological investigation involving adult subjects.     

Exercise during hyperoxia and hyperbaric oxygenation

Physiological reactions during exercise were tested under hyperoxic and hyperbaric conditions. In 6 subjects walking and running at increasing speeds on a treadmill, maximum performance showed little change when the respired air was enriched with O₂. Maximum metabolism, measured by CO₂ production, increased by 3.2%. During exercise on a bicycle ergometer, maximum O₂ uptake increased by 3% in 5 subjects breathing pure O₂ at 1 ATA. During hyperoxia the maximum O₂ consumption measured at 2 and 3 ATA did not differ significantly from that measured at 1 ATA. Heart rate showed highly comparable maximum values under the various experimental conditions. During submaximal exercise, heart rate was consistently lower when the subjects breathed O₂. The O₂-linked difference became slighter with every increase in work load. Under hyperbaric and hyperoxic conditions, ventilation was invariably reduced during exercise.     

Comparison of arterial,end-tidal and transcutaneous PCO2 during moderate exercise and external C02 loading in humans

Summary Static relationships between arterial, transcutaneous[/p] and end-tidal PCO₂ (P _aCO₂, P _tc CO ₂, P _etCO₂) as well as the dynamic relationship between P _etCO₂ and P _tcCO₂ were studied during moderate bicycle ergometer exercise with and without external C0₂ loading. The exercise pattern consisted of 5-min intervals of constant power at 40 W and 100 W and 900 s of randomised changes between these two power levels. The external CO₂ loading was achieved by means of controlled variations of inspiratory gas compositions aimed at a constant P _etCO₂ of 6.5 kPa (49 mm Hg). The P_etO2 was regulated at 17.3 kPa (130 mm Hg). Under steady-state conditions all PCO₂ parameters showed close linear relationships. P _aCO₂/P _tcCO₂ was near to identity while the P _etCO₂ systematically overestimated changes in P _aCO₂. No relationship showed a significant influence of the exercise intensity. Transients of P _tcCO₂ are considerably slower than P _etCO₂ transients. The dynamic relationship between both parameters was found to be independent of whether internal or external C0₂ loadings were applied. It is concluded that the combination of P _etCO₂ and P _tcCO₂ measurements allows an improved non-invasive assessment of P _aCO₂. While P _etC0₂ better reflects the transients, P _tcCO₂ can be employed to determine slow changes of the absolute P _aCO₂.     

Three successive steady-state CO2-response curves

Three successive steady-state CO₂-response curves, with 10-minute intervals, were taken in seven healthy subjects at rest in normoxia. No systematic difference in the slopes between the three curves could be found. The results suggest that a previous steady-state CO₂-response curve does not change the sensitivity of the ventilatory controlling system for CO₂.     

Maximal breath-holding time and immediate tissue CO2 storage capacity during head-out immersion in humans

This study tested three possible mechanisms that could explain the prolonged breath-holds (BH) previously observed in humans during submersion in 35°C (thermoneutral) water, including a reduced metabolism, a decreased CO₂ sensitivity, and an increased CO₂ storage capacity. During immersed BH (n = 13), maximal BH time was prolonged by 20.3% (P < 0.05), the rate of rise of end tidal partial pressure of carbon dioxide (P _ETCO₂) was slower (P < 0.05) by 31 % (compatible with increased CO₂ storage capacity), but the breaking-pointP _ETCO₂ (CO₂ sensitivity) and the rate of decrease of end tidal partial pressure of oxygen (metabolism) were unchanged. During air breathing (n = 5), immersion resulted in a significant decrease in tidal volume (11%), but did not affect O₂ uptake, CO₂ elimination , or respiratory exchange ratio (R). During a 4-min CO₂-rebreathing (n = 9), the slope of the hypercapnic ventilatory response curve (CO₂ sensitivity index) was unchanged by immersion, but the significantly decreased ,R, and rate of rise inPETCO₂ during immersed rebreathing indicated an increase in the acute CO₂ storage capacity (SC). The estimated SC (n = 9), based on an assumed cellular respiratory quotient of 0.8, were 0.52 (SEM 0.03) ml · kg⁻¹ · mmHg⁻¹ for control and 0.66 (SEM 0.04) ml · kg⁻¹ · mmHg⁻¹ for immersion. A proposed mechanism for the increased SC during immersed BH and during immersed rebreathing is that immersion accelerated CO₂ redistribution in the body by increasing perfusion to some low-perfused, low-metabolism, and high-capacity tissues, such as resting skeletal muscle. The increased SC during immersion, however, did not correlate with the prolonged BH duration (n = 9,P > 0.05). The mechanism of the latter remains unclear.     

Enhancement of O2 and CO2 Transfer Through Microporous Hollow Fibers by Pressure Cycling

An intravascular gas exchange device for the treatment of respiratory failure consisted of a multitude of blind-ended hollow fibers glued in a pine-needle arrangement to a central gas supply catheter. It has previously been shown that gas desorption rates can be significantly enhanced by cycling gas pressure between a hypobaric level of 130 and an ambient level of 775 Torr. In this study, influences of the cycling frequency (f) and the cycle fraction during which hypobaric pressure is applied () were investigated. Rates of O₂ desorption from O₂-saturated water and CO₂ desorption from CO₂-saturated water into a manifold containing 198 fibers, 380 m in diameter, were measured over a range of f from 0.2 to 1.0 Hz, from 0.1 to 0.8, and fiber lengths from 4 to 16 cm. Relative to operation at ambient pressure, pressure cycling increased O₂ transfer 3–4 times and CO₂ transfer 4–6 times when the water flowed over the fiber manifold at 2.3 l/min. Transfer rates were relatively insensitive to f and with 80–90% of maximum enhancement obtained when was as low as 0.2. Transfer rates increased continuously with fiber length, implying that pressure cycling reduced the intra-fiber resistance to gas diffusion. A mathematical diffusion model which utilized only two adjustable parameters, a mass transfer coefficient for O₂ and for CO₂, simulated the trends exhibited by desorption data. © 1998 Biomedical Engineering Society. PAC98: 8745Hw, 8790+y