A radiographic method to estimate lung volume and its use in small mammals

In this paper we develop a method to estimate lung volume using chest x-rays of small mammals. We applied this method to assess the lung volume of several rodents. We showed that a good estimator of the lung volume is: VL= 0.496 · VRX≈1/2·VRX, where VRX is a measurement obtained from the x-ray that represents the volume of a rectangular box containing the lungs and mediastinum organs. The proposed formula may be interpreted as the volume of an ellipsoid formed by both lungs joined at their bases. When that relationship was used to estimate lung volume, values similar to those expected from allometric relationship were found in four rodents. In two others, M. musculus and R. norvegicus, lung volume was similar to reported data, although values were lower than expected. Key terms: Lung volume, x-ray, method, rodents


INTRODUCTION
Lung volume is one of the most relevant structural parameters related to the aerobic capacity of mammals.It is approximately linearly related to body mass (Stahl, 1967;Gehr et al., 1981), constituting about 5% of the body mass (Calder, 1996;Schmidt-Nielsen, 1984).The volume of the lungs is related to oxygen diffusing capacity (D LO2 ) and oxygen consumption (V O2 ) (Maina 2000) by the relationship: V O2 =D LO2 •∆P O2 , where ∆P O2 is the gradient of partial pressures between alveoli and capillaries.The oxygen diffusing capacity may be described as: D LO2 = where κ is the Krohg's diffusion coefficient, dSA is the alveolar surface density (m 2 /l), V L is the lung volume, and τ h is the thickness of the blood-gas barrier (Maina 2000).Thus, larger lung volume may be a way to have   et al. (1981) reported that lung volume m e a s u r e d b y w a t e r d i s p l a c e m e n t underestimates the total lung capacity (TLC) by approximately 10-30%.

• •
Methods for estimating lung volume are demanding because they usually require the removal of the lungs to directly measure the volume through water displacement methods (Scherle, 1970).Moreover, this procedure frequently produces a variable athelectasis or elastic retraction that has been estimated to be approximately 5% in the first few hours (Weibel et al., 1981).Other techniques, such as nitrogen washout, helium dilution technique, plethysmography, and radiographicplanimetric methods have been developed to estimate functional lung capacities (Kilburn et al., 1993).In this regard, the Barnhard-Loyd radiographic method has good correlations with the other methodologies in estimating total lung capacity in humans (Barnhard et al., 1960, Loyd et al., 1966, Kilburn et al., 1993, Pappas et al., 1998).It considers the thorax to be a stack of five elliptical sections that allow an estimation of the lung volume by integration from measures of transverse diameters and heights in anteroposterior and lateral chest x-rays.Although the Barnhard-Loyd radiographic method appears to be as accurate as pletysmography, it does not seem to have gained wide use, probably because it is a lengthy process requiring as much as 30 minutes for the determination of total lung capacity (Rodgers and Tannen, 1983).The use of this method in small animals is limited by the size of the lungs, and the idea of correlate lung volume with chest xrays appears to be superior.
This paper aims to correlate the lung volume measured by water displacement with fixed measurable parameters in chest x-rays of small rodents.We hope that these parameters will be useful in estimating lung volume.This method could be useful for studying the ontogeny of lung volume and making inter-specific comparisons.
Animals were sacrificed by means of CO 2 exposure, and chest x-rays were taken.All experiments comply with the current laws of Chile and the Ethical Committee of the Faculty of Sciences of the University of Chile, where the experiments were performed.All x-rays were obtained with a standardized technique of 300 mA, 0.05s and 50 kV and with 1m of object-focus distance, a sufficient distance to make magnification negligible.
On the images of the lungs, a straight line (RL) was traced between the two costophrenic recesses, and the following magnitudes were measured: i) the width of RL (W) as a transversal measure; ii) the height between RL and the top of the left (H 1 ) and the right (H 2 ) lung; and iii) left (w 1 ) and right (w 2 ) lung width on the middle point of diaphragm domes.Both w 1 and w 2 were used as estimators of the anteroposterior diameters of lungs, assuming circular lung sections (Fig. 1).
These measurements were used to build a rectangular box with a volume of: V RX = to contain the lungs and mediastinum organs.Given that V RX is a dimensional volumetric characteristic (L 3 ) that contains the actual lung volume (V L ), we expected both variables be tightly correlated.
Next we performed a small tracheotomy and instilled the lungs with a 2.5 % Glutaraldehide solution with a plastic catheter.The trachea was ligated to keep the intrapulmonary fixative volume and a small quantity of latex or auto-curable acrylic was introduced to fill the airway and avoid collapse.The next day the lungs were removed by thoracotomy, and their actual volumes (V L ) were estimated by means of a water displacement method (Scherle, 1970).
Correlation between V L and V RX was tested with student t test: . The ratio R VL =V L /V RX (x± 1Sd) was calculated as an estimator of the proportion of V RX that is filled by the lungs.Thereafter V L * = R VL • V RX would be a good estimator of lung volume (V L ).The ratio between the true volume and the estimated volume, R = V L /V L * , was computed for all individuals in this group, testing the hypotheses Ho: R = 1 vs H 1 :R ≠ 1 1 with .

RESULTS
The lung-filling method produced good results (Fig. 2), yielding, by water displacement, lung volume values similar to those observed by other authors for rats and mice: Mus musculus 0.69 ± 0.08 ml; Rattus norvegicus 5.76 ± 0.69 ml and Cricetus auratus 1.56 ± 0.34 ml (Table I).
A good correlation between V RX and V L was found (r = 0.992, t 10 = 24.85,p < 0.01).The ratio R VL was 0.496 ± 0.138, and we could then estimate the lung volume from thorax x-rays by means of: V * L = 0.496•V RX • The ratio between the actual lung volume and the radiographically-estimated lung volume, R = V L /V L * = 0.997 ± 0.279 was not significantly different from the unit (p >0.05).A good correlation between V RX and V L was found (r = 0.992, t 10 = 24.85,p << 0.01).
The estimations of the lung volume (V L * ) from x-rays and the expected values from Gehr's relationships for all individuals (the original 12 plus the additional 16) are summarized in Table II.

DISCUSSION
The method of filling the lungs yielded good results and avoided a significant retraction of the parenchyma when the lungs were removed.Previous reports suggest that the lung may lose approximately 5% of its volume in the first hours after removal due to elastic recoil, but then remains constant for weeks.Our  between the water displacement lung volume (V L ), the X-ray box volume (V RX ), estimated lung volume (V L * ) from chest x-rays and the expected lung volume (EV L ) from Gehr's allometric relationship for twelve individuals of three small mammal species.Mb is body mass and R  lung volume measurements were made the day after removal, so we did not expect recoils larger than this value.Our estimations of lung volume were within the range of what has been previously reported for these species.For example, our values of lung volume for R. norvegicus and M. musculus, 5.76 ± 0.69 ml and 0.69 ± 0.08 ml respectively, are within the range of 6.3 ± 0.2 ml and 0.74 ± 0.075 ml reported for rats and mice (Weibel, 1973).However, when body mass was introduced into the allometric relationship between body mass and lung volume: V L = 0.046 • M 1.06 b , where M b is in kilograms and V L in liters (Gehr et al., 1981), the expected values for the lung's volume were larger than those obtained in our sample of M. musculus and R. norvegicus.However, it is known that mice and rats have smaller lungs than those expected as a result of Gehr's relationship.For example, Standaert et al (1985) studied 6 rats (286.67 ± 184.1g) and obtained an average lung volume of 9.59 ± 4.58 ml, both high oxygen diffusion capacity and high maximum oxygen flow.T o e s t i m a t e o x y g e n d i f f u s i o n capacity, the lung volume (tissue) is p r e f e r r e d t o f u n c t i o n a l c a p a c i t i e s .Although related, they are not the same.F o r e x a m p l e , S t a h l f o u n d t h a t i n mammals lung volume and vital capacity (VC) were related to the body mass (M b ) by V L = 53.5•M 1.06 b and V L = 56.7•Mb 1.03

Figure 1 .Figure 2 .
Figure 1.Thorax x-ray of one individual of Rattus norvegicus, showing the reference lines for the computation of V RX .