| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Technical Developments |
1 From the Department of Diagnostic Radiology, University Hospital Kiel, Arnold-Heller-Strasse 9, 24105 Kiel 1, Germany. Received July 26, 2001; revision requested September 12; revision received February 8, 2002; accepted February 28. Address correspondence to J.B. (e-mail: juergen.biederer@rad.uni-kiel.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
© RSNA, 2002
Index terms: Animals • Experimental study • Lung, MR, 60.121411, 60.121412 • Model, anatomical
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
|
Materials and Methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
|
A flexible membrane replaced the true diaphragm at the "caudal" part of the model, closing the lower thoracic space outlet. This artificial diaphragm consisted of a silicone layer in the shape of a porcine diaphragm in midinspiration. Intermittent evacuation of the space below this membrane led to collapse and extension. The volume of the thoracic cavity was rhythmically extended and diminished, simulating a breathing cycle with partial expansion and collapse of the lung. Shapes and sizes of the chest wall and the diaphragm were obtained from the midinspiration CT scan of a living anaesthetized pig.
Heart-Lung Preparation
The specimens consisted of fresh heart-lung preparations from adult pigs (about 80–100 kg) and were obtained from a local slaughterhouse. The objectives of the first experiments were to standardize the preparation procedure and to evaluate if the material would be satisfactory for further imaging studies with the artificial thorax. No animals were sacrificed for the particular purpose of this study. The preparations passed the regular veterinarian controls of a licensed slaughterhouse and were treated with the same hygiene precautions as fresh meat. The present series included five fresh preparations and two lungs that were preserved in the freezer at -32°C for 7 and 21 days, respectively. The time needed for the preparation of the specimens and for setting up the artificial thorax was recorded for each experiment.
After the specimens were inspected and the lesions sutured, if necessary, the preparation was collocated into the lower shell of the artificial thorax. A 6.5-mm diameter tracheal tube (Portex; SIMS Portex, Kent, England) was introduced into the trachea and connected to the dedicated outlet through the artificial chest wall. The heart was left in situ to preserve intact pulmonary veins for circulation experiments. Catheters were placed in the main pulmonary artery and the left ventricular outlet. For the pulmonary artery, we used a 26-F balloon catheter (Biocath; Bard, West Sussex, England) inflated with 10 mL of water or air to tighten the vessel. Additional tightening was provided by the flaps of the intact pulmonary valve that closed around the catheter shaft. The left ventricular outlet was accessed with a 22-F double-balloon catheter (Silkolatex; Rüsch, Kernen, Germany) via the common carotid artery trunk. The distant balloon was inflated with 20 mL of water inside the left ventricle to prevent the catheter from dislocation. The second balloon was inflated inside the ascending aorta itself (10-mL volume). A silicone tube cuff tightened around the outer aortic wall guaranteed tight fitting of the balloon. The catheters were connected with tubes passing through the double wall of the container. These tubes could be connected to an external pump to establish circulation (Fig 1). For MR imaging, we used long tubes to place the evacuator and the pump outside the scanning room.
After the artificial thorax was closed, the pressure was lowered until the lung was maximally inflated. This pressure was maintained during the imaging tests. Between the experiments, when the model was moved from one room to another, the evacuation was switched off and the lungs were deflated. All specimens were dissected after the experiments and inspected for macroscopic changes.
Imaging Tests
A range of imaging studies that included radiography, CT, MR imaging, and digital subtraction angiography were performed to gain experience with inflated porcine lungs and different imaging modalities. The studies were performed in simulated inspiration, with imaging parameters as would be used in human patient studies.
All seven lungs were initially examined with radiographs and CT scans after being mounted inside the artificial thorax. The radiographs were obtained at 80 and 120 kVp (small focus, antiscattering grid, film-focus distance of 110 cm) by using a conventional x-ray unit (Maximus CM 80; Philips Medizinsysteme, Hamburg, Germany). The CT scans were obtained with a commercial single-section spiral CT scanner (Somatom Plus-S; Siemens, Erlangen, Germany). The whole lung was scanned with a beam collimation of 10 mm and a table speed of 15 mm/sec (rotation time of 1 second, pitch factor of 1.5, 210 mAs, 137 kVp). Images were reconstructed in a section thickness of 8 mm (overlapping reconstruction), with a slim reconstruction profile and an edge-enhancing algorithm at a matrix of 512 x 512 pixels.
After initial CT and radiography, four of seven lungs were taken for MR imaging. This group included the two former frozen lungs, which were also used for angiography and perfusion experiments after MR imaging. One of the freshly prepared lungs remained in the MR imager for a perfusion experiment.
All MR studies were performed with a commercial 1.5-T unit (Magnetom Vision; Siemens), with dedicated phased-array coil for cardiopulmonary imaging. For T1-weighted imaging, we used a gradient-echo sequence (Flash2d; Siemens) as would be used in a patient, with slabs of 19 images (222 x 256 matrix, 350-mm field of view) acquired within 23 seconds, a repetition time msec/echo time msec of 100/2.2, and a flip angle of 50° (5–7). For T2-weighted imaging, we used a rapid acquisition with relaxation enhancement (RARE) sequence with half-Fourier acquisition (HASTE; Siemens) (2,000/43, 350-mm field of view, 224 x 256 matrix, 7-mm coronal sections) (15) and a turbo spin-echo sequence (5,000/120, 350-mm field of view, 270 x 512 matrix, 8-mm transverse sections) (16,17).
Three of seven lungs were taken directly for angiographic and perfusion experiments after initial CT and radiography. Angiograms were obtained with a single C-arm digital subtraction angiographic unit with a 1,024 x 1,024 pixel image intensifier (Multistar TOP; Siemens) during perfusion with saline (0.9% NaCl). The perfusion rate was adjusted to a pulmonary arterial pressure of 40 mm Hg. Between one and three studies with digital subtraction angiographic technique were acquired with 10–15-mL boluses of radiopaque nonionic contrast material (iopentol [Imagopaque 300]; Nycomed, Munich, Germany) with iodine concentration of 300 mg/mL. The image frequency of one per second was adapted to the slow perfusion. A special connector for the placement of an 8-F percutaneous sheath (Radiofocus Introducer II; Terumo, Tokyo, Japan) at the access to the pulmonal artery was designed for the use of catheters. To overcome the friction of angiographic catheters within the soft silicone tube of the balloon catheter, an additional 90-cm flexible 7-F sheath (Super Arrow-Flex; Arrow International, Reading, Pa) was introduced into the first sheath and the balloon catheter, with the tip reaching the pulmonary arterial bifurcation. During and after angiographic examination, perfusion was maintained for at least 30 minutes with intermittent fluoroscopic controls (one lung, 30 minutes; two lungs, 60 minutes). Then CT scanning was repeated and the lungs were examined with MR imaging.
The imaging studies were evaluated by two observers (J.B., M.H.), with consensus. Radiographic, CT, and MR images were interpreted to describe the inflation of the lung and the presence of infiltration with liquid (edema) before and after perfusion studies. The angiograms were evaluated for the visualization of pulmonary arteries and veins; that is, for complete or incomplete filling with liquid containing contrast material. The fluoroscopic controls were used to detect lung opacification in case of edema.
One specimen was subjected to continuous perfusion with 200 mL/min of 0.9% NaCl solution at MR imaging to quantify perfusion-dependent changes of the lung parenchyma signal intensity. Transverse T2-weighted RARE and turbo spin-echo images from the protocol were obtained before perfusion and after 10, 30, and 90 minutes. To quantify the signal-to-noise ratio (SNR), we used the software of the MR unit. Six transverse sections were defined at the level of the tracheal carina and in 20-mm steps below this section. The regions of interest for signal intensity measurement included the lung parenchyma on these sections but excluded big vessels and bronchi. Additional measurements outside the artificial chest wall were used to calculate the SNR, as proposed by Kaufman et al (18).
|
Results |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
hours. Repair of small cuts in the lung parenchyma was feasible with surgical suturing of the visceral pleura. The remaining leakage of air as a result of small injuries was tolerable in all cases, as it was fully compensated by the continuous evacuation of the artificial thorax cavity. The negative pressures required to maintain the lungs inflated ranged between -2 x 103 and -3 x 103 Pa (Fig 2). All lungs could still be inflated after 9 hours of experiments at room temperature.
|
|
|
|
|
The angiograms demonstrated complete filling of central and segmental pulmonary arteries in all three specimens, but only in few subsegmental and smaller vessels. Missing contrast enhancement in peripheral subsegmental pulmonary arteries and in the ventricles of the heart was interpreted as small thrombi or air (Fig 5). In all specimens, the pulmonary veins showed contrast enhancement but appeared only partially filled. The perfusion rates did not suffice to fill the venous system completely.
After the perfusion studies, findings from all modalities demonstrated liquid inside the pulmonary vessels and, to a variable extent, inside the airspace and the interstitial tissue. On CT and MR images of the two lungs, which were perfused for 60 minutes, the appearance of liquid retention resembled human disease. The elevated water content was consistent with findings of ground-glass opacities on CT scans and high signal intensity on T2-weighted MR images, mimicking pulmonary infiltration, as it is known from human patient studies. Interlobular interstitial thickening indicated water uptake into the lymphatic system (Fig 6). Dissection of the lungs after the experiments confirmed the high water content.
Without perfusion, the SNR of the porcine lung parenchyma on T2-weighted MR images appeared to be lower than that in patient studies (3.94 on RARE and 1.8 on T2-weighted turbo spin-echo images). After starting perfusion, the signal intensity on T2-weighted RARE images increased visibly and the average SNR reached 7.08 (increase of 79.7%) and 8.22 (increase of 108.6%) after 10 and 60 minutes, respectively. After 90 minutes, the images demonstrated water uptake into the alveolar spaces and the interstitial tissue (Fig 6). At this time, the SNR reached 18.86 (increase of 378.7%) on T2-weighted RARE and 19.65 (increase of 991%) on T2-weighted turbo spin-echo images. The findings were similar to those obtained after perfusion and angiography.
Macroscopically, the two formerly frozen lungs looked similar to the fresh preparations, but the visceral pleura was more fragile and difficult to suture. The images from these lungs did not differ from those of fresh preparations, until perfusion was started. The perfusion rates at 40 mm Hg did not exceed 100 mL/min, and no backflow through the pulmonary veins could be achieved. The fluoroscopic controls showed rapidly increasing opacities of both lungs within the first few minutes of perfusion. After 20 minutes, the experiments were stopped. Dissection of the two specimens confirmed a massive uptake of liquid into the parenchyma.
|
Discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
Inflation of a lung explant can be achieved with either continuous evacuation of the space around the heart-lung preparation or with a respirator connected to the trachea to apply positive pressure. In both cases, the lung would be expanded by the higher pressure within the bronchial system compared with the airspace around the lung itself; that is, inside the artificial thorax. The principal idea behind inflating the lung by lowering the surrounding pressure with the trachea open to normal atmospheric air pressure levels is well known and has been used in the early days of thoracic surgery and for whole-body respirators (21). We decided to use this approach, as it facilitates the whole procedure. No MR imaging-compatible respirator is needed. Moreover, continuous evacuation simulates the in vivo situation, where the same principle of negative pressure in the pleural space is keeping the lung expanded. The pressure levels of -2 x 103 to -3 x 103 Pa needed for expansion of the lung in our experiments are near the range of human physiologic values (-1.5 x 103 to -2.5 x 103 Pa in maximum inspiration) (21).
The first imaging experiments demonstrate the feasibility of the model experiment. All lungs could be kept inflated with continuous evacuation of the artificial thoracic cavity. The lung parenchyma appeared with the specific properties of the different modalities. Angiography at low flow rates was feasible. The use of a transparent material for the artificial chest wall facilitates observation of function and possible malfunction of the experiment. Incomplete inflation of the lung and leaks can be detected easily and correlated to the imaging findings.
Further development of the model will have to concentrate on the heart-lung explants. In the first series, the preparations were used without any revascularization procedure. Only residual thrombi were found in the peripheral vessels and the heart ventricles after exsanguination of the animals during sacrifice. These thrombi and air bubbles within the capillary vessels may represent the reason for the high peripheral resistance that limited the feasible perfusion rates. Better conditions will be achievable with immediate establishment of a perfusion with anticoagulant agents (eg, heparin) after explantation (22). Nevertheless, our test findings show that a low flow perfusion can be achieved even under these simplified conditions.
The possible range of experiments to be performed with this model is wide. Mechanical and pharmacologic manipulations of lung specimen are easy to realize without the use of living animals. One example would be ventilation-perfusion experiments under controlled conditions (9). In particular, ventilation studies with hyperpolarized noble gases could be easily performed (23–26). For these types of experiments, the flexible diaphragm can be used to simulate a breathing cycle. Whole-animal lung explant systems have also been demonstrated to be applicable to pharmacologic experiments for simulating the in vivo response to lung toxins, such as inorganic and organic fibrogens (19). The model thorax might facilitate use of some of these experiences for imaging studies.
Overall, the advantages of the presented lung model are clear. The model is based on fresh material to simulate conditions that are very close to the in vivo situation. The experimental set-up was successfully tested for different imaging modalities, including perfusion studies. The flexibility with respect to planning and preparing studies is very high, and standardized material from 100-kg pigs can be purchased from local slaughterhouses at low cost (about $4.00 for a heart-lung preparation) in virtually unlimited quantities. No specific animal breed is needed. Even formerly frozen material can be used for many studies, although we cannot recommend its use in perfusion experiments.
The artificial thorax provides an easy-to-prepare experimental platform for a variety of MR imaging lung studies under close to in vivo conditions. It has been tested successfully and may be also used for any other imaging modality, including angiography. A particular advantage is the high flexibility of the model.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, J.B.; study concepts and design, J.B., M.H.; literature research, J.B.; experimental studies, J.B.; data acquisition, J.B.; data analysis/interpretation, J.B., M.H.; statistical analysis, J.B.; manuscript preparation and editing, J.B.; manuscript definition of intellectual content, revision/review, and final version approval, J.B., M.H.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H Bolte, C Riede, S Muller-Hulsbeck, S Freitag-Wolf, G Kohl, T Drews, M Heller, and J Bieder Precision of computer-aided volumetry of artificial small solid pulmonary nodules in ex vivo porcine lungs Br. J. Radiol., June 1, 2007; 80(954): 414 - 421. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Biederer, A. Schoene, S. Freitag, M. Reuter, and M. Heller Simulated Pulmonary Nodules Implanted in a Dedicated Porcine Chest Phantom: Sensitivity of MR Imaging for Detection Radiology, May 1, 2003; 227(2): 475 - 483. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
