Functional magnetic resonance imaging of the lung

G. H. Mills*,1, J. M. Wild2, B. Eberle3 and E. J. R. Van Beek2

1 Unit of Academic Anaesthesia, University of Sheffield and The Directorate of Critical Care Medicine, Royal Hallamshire Hospital, Glossop Road, Sheffield, S10 2JF UK. 2 Unit of Academic Radiology, University of Sheffield, Royal Hallamshire Hospital, Sheffield, UK. 3 Dept of Anaesthesiology, Johannes Gutenberg University School of Medicine, Mainz, Germany

Corresponding author. E-mail: G.H.Mills@sheffield.ac.uk

Keywords: anaesthetics gases, xenon; gases non-anaesthetic, helium; measurement techniques, magnetic resonance imaging; oxygen, regional intrapulmonary

Limited imaging of lung function is possible with methods that use ionizing radiation, such as computed tomography (CT) and lung scintigraphy. Methods using magnetic resonance (MR) are preferable because they do not involve ionizing radiation. At first, MR methods were difficult because of the lack of protons and artefacts caused by inhomogeneity in the lungs. Recently, new techniques have allowed static and dynamic lung imaging with MR. We discuss the methods that allow functional imaging of the lung and describe new developments in functional MR imaging (MRI). From being almost invisible to MRI, lung images with MR may soon radically affect the way we assess the function of the lung and the pulmonary circulation.

Techniques available to functionally image the lung

The first attempts to obtain functional information of the lungs date back to the early days of x-ray fluoroscopy. However, these concentrated on the tissues and structures surrounding the lung. Fluoroscopy could show diaphragm motion during spontaneous breathing and sniffing, but gave variable results in detecting pathology.6 87 Ultrasonography allows useful imaging of structures around the lung, such as pleural effusions, diaphragm movement during sniffing or phrenic nerve stimulation, or even diaphragm thickness,30 48 55 56 79 80 82 but lung tissue itself is not readily penetrated by ultrasound.

Nuclear medicine and scintigraphy
Lung scintigraphy was developed in the 1960s, and has become the key test for assessment of lung perfusion and ventilation.110 This method uses inhaled radioactive substances and monitors the wash-in and wash-out of the isotopes with scintigraphic detection; however, the images do not have high temporal or spatial resolution. The inhaled radioactive gases include 133Xe, 127Xe and 81Kr. Radioactive aerosols including technetium 99m diethylenetriamine pentacetic acid (DTPA) suffer from problems of deposition in the larger airways. Technetium 99m Technegas uses carbon particles as a carrier as an ultrafine aerosol to allow deeper deposition. However, anatomical detail of the air passages remains poor. These techniques have been refined by combination with rotating gamma cameras to form single-photon emission CT (SPECT), which allows cross-sectional imaging with a slice thickness of 15 mm.

Positron emission tomography allows resolution down to 10 mm. However, temporal resolution is limited compared with the duration of the respiratory cycle, requiring about 30 s per image. Furthermore, a proximal cyclotron is needed for isotope production because the tracers have short half-lives.

Electrical impedance tomography (EIT)
EIT detects the impedance of the chest to very small electrical currents applied via skin surface electrodes. This allows virtually non-invasive imaging of axial slices of the lung throughout the respiratory cycle, by differentiat ing between regions where electrical resistance is high (inflated lung) and regions where current can flow more easily (collapsed lung). Although the spatial resolution is poor, EIT can give a real-time measure of lung tissue properties such as airway collapse, expansion or over-inflation.20 21 42 6769 99 This ability may be enhanced by using multiple electrical frequencies. A significant advantage of the technique is that it can be used at the bedside even in intensive-care patients.

Computerized tomography
Lung tissue can be imaged rapidly and volumetric data obtained by using slip-ring technology (spiral CT), multiple rows of detectors (multislice CT) and high-speed reconstruction algorithms, which speed image acquisition and increase resolution. High-resolution CT (HRCT) uses a narrow-beam collimation (1 mm), a narrow field of view and high spatial frequency reconstruction algorithms.78 Functional information can be gained in detection of air trapping, lung volume measurement (total and regional) and direct imaging of regional ventilation with the aid of inhaled xenon or aerosols.

Dynamic CT imaging can be applied to a single slice during a respiratory cycle, but even when using the latest multidetector technology, CT is not fast enough to image the whole lung during the respiratory cycle. Paired inspiratory and expiratory HRCT scans allow detection of air trapping.61 62 Assessment of lung volume is usually done by spirometry, with body plethysmography or helium-dilution methods in addition. However, volume information for specific lung regions or individual lungs is not available without imaging. CT scanning gives this information, although inspiratory helical CT consistently underestimated total lung capacity by 15%,61 and expiratory helical CT overestimated residual volume by almost 1 litre. Further refinements using three-dimensional reconstructions have reduced errors to about 3%.57

Dynamic lung CT requires very rapid imaging106 107 116 and potentially high radiation exposure if continuous imaging is used. To increase speed, the initial method was to fire electron beams at tungsten targets, generating X-rays that could be directed very rapidly by controlling the direction of the electron beams, so avoiding moving mechanical parts. Electron-beam CT scans are usually performed at three levels (aortic arch, lung base and carina), to reduce interference from lung motion.116

Dynamic upper-airway imaging can assess changes in the trachea during inspiration and expiration in normal subjects (35% decrease in area), tracheomalacia (82% decrease)105 and sleep apnoea (75%).44 Inspiratory and expiratory images can be obtained with a temporal resolution of 100 ms. This has allowed the assessment of conditions such as emphysema, where in addition to increased air space size and loss of interstitial tissue reducing attenuation, lung tissue remains hyperinflated during expiration and so does not undergo the expected increase in density seen in normal lungs.70 71 103 104 The technique has also allowed identification of patients with normal, restricted or obstructive lung diseases.63 Bronchial reactivity in asthmatics has also been assessed before and after methacholine and salbutamol administration.13 Diagnostically, the technique can show a variety of causes of main bronchus obstruction in children, such as dynamic changes in airway calibre, intraluminal polyps, tracheal atresia, compression resulting from mediastinal masses, and foreign bodies in the airway,17 and is improved when combined with HRCT.18 In pigs, continuous dynamic multiscan CT can show lung recruitment effects in experimental models of actue respiratory distress syndrome75 and assess ventilatory strategies during cardio pulmonary resuscitation.76

CT images show the lung parenchyma and from this, information can be inferred about the air spaces. However, to image the air spaces directly some form of inhaled contrast is needed.

Xenon is a noble gas. It is denser than air and will appear denser on CT scan depending on concentration. It is also non-radioactive and possesses anaesthetic properties, related to its lipid solubility, which become evident in inspired concentrations above 21% (MAC 71%).32 122 Equilibration of xenon within the air passages is not instantaneous and therefore wash-in and wash-out can be assessed. The greater the density of xenon in a particular area, the greater the ventilation in that region. Unfortunately, the anaesthetic effects and technical difficulties with imaging have limited the use of xenon. Early studies involved animals,53 normal subjects, or sedated and ventilated intensive-care patients. The technique was standardized by Simon and colleagues,101 and single-breath techniques using rapid electron-beam dynamic CT with subtraction techniques109 to compare images with and without xenon have reduced the problems of the anaesthetic effect. Xenon is soluble and passes from the lung into the circulation, which complicates interpretation. The rate of appearance of xenon in a region of the lung provides information regarding ventilation, but the xenon will then pass into the blood, giving information on perfusion. Precise estimation of xenon concentration has been investigated using monochromatic synchrotron radiation CT, which has allowed the assessment of two simultaneous images using different energy levels, which allows the calculation of local xenon concentration.11 Gravity can affect the distribution of perfusion and ventilation and cause areas of lung collapse. Image interpretation is complex,109 but microvascular regional blood flow and the effects of gravity and posture have recently been seen with xenon CT in animals.66 120 Xenon also passes in the circulation to the brain, where it is taken up because of its lipid solubility. Inhaled xenon may therefore have a role in assessing cerebral perfusion.43 66 96

Oxygen and sulphur hexafluoride (SF6) mixtures, and oxygen and helium mixtures, have been used during continuous-breathing CT studies of patients with chronic obstructive pulmonary disease (COPD). Images at end-inspiration showed changes with time, by comparing images taken during breath number two and number 60. Abnormal areas of lung with slow time constants were seen, with marked changes in attenuation between the two images. Early gas flow was poor compared with the late images.123

CT is versatile, and is being increasingly used to study pulmonary embolism, where it is gradually replacing the use of ventilation–perfusion scintigraphy. CT pulmonary angiography requires injection of 100–130 ml iodinated contrast agent into a peripheral vein, followed by imaging when opacification of the pulmonary arterial tree is optimal. Using state-of-the-art CT systems, a resolution of 1 mm is possible, with visualization of sub-sub-segmental pulmonary arteries.47 91 112

MRI of the lungs

Basic theory of MRI
To be sensitive to nuclear magnetic resonance (NMR), a material needs to contain atoms with odd numbers of nucleons (protons and neutrons). All nucleons possess a spin, which is an intrinsic property of atomic particles, as are mass and charge. Therefore, molecules containing hydrogen (1H) protons, such as water and numerous organic molecules, are sensitive to NMR and are candidates for MRI. Other nuclei are also sensitive, providing they possess unpaired nucleons in their nucleus, examples being 13C, 3He, 15O and 129Xe. In a strong static magnetic field (B0), these uncoupled spins can exist in one of two energy states aligned either parallel to or anti-parallel to B0. Conventional proton MRI uses the fact that in thermal equilibrium, a slightly larger proportion of the spins adopt the lower energy ground state (parallel to B0). This population can be excited or tipped over into the higher state by radio frequency (RF) excitation. The size of this fractional population imbalance is governed by Boltzmann thermal statistics and is directly proportional to the static field strength and the gyromagnetic ratio of the nucleus, and inversely proportional to the temperature. The gyromagnetic ratio of a given nucleus is the resonant frequency of its spin divided by the static field strength. The subsequent loss of excitation and return of spins to the lower ground state releases a coherent RF discharge, which we measure in MRI as the current induced in a coil as the spins decay to ground state. This is known as the free induction decay (FID). The detection of NMR was first observed by Bloch and Purcell in the 1950s, for which they were subsequently awarded a Nobel prize. This discovery of fundamental physics has allowed numerous aspects of pure and applied research in all areas of science, with many industrial and medical applications.

The relaxation of the spin population following excitation is characterized by its longitudinal relaxation rate (T1). T1 is related to the time taken for spins to relax back to their ground state and is very much dependent on the degree of energy interchange between the atom and its surroundings or lattice. Similarly, the FID itself represents the decay of spin magnetization (precession) after the spins have been tipped over by the RF pulse. This precession loses energy through relaxation processes characterized by a separate transverse spin–spin relaxation time (T2). T2 can be thought of as a measure of the coherence of the RF signal produced as the spins decay to the ground state. If the coherence is poor then the signals from each of the nucleons are not in phase with one another and so start to cancel each other out. If they remain in phase (coherent) they will summate, giving a large T2. For a sample to give an FID with sufficient signal intensity to provide a spatial map or MR image, the spin density needs to be sufficiently large, or the magnetic field strength extremely high. This is why conventional clinical MRI scanners that image protons in the body typically operate at field strengths of 1 T and above. By imaging with long repetition times, all spins with a range of T1 relaxation times have a chance to relax back to thermal equilibrium before they are subsequently tipped by the next RF discharge used to scan the next line of data, which is used to build up the image in reciprocal space. Such an image would represent a spatial map of absolute proton spin density and as such tends to show little contrast between adjacent tissue types of approximately the same spin density. Contrast can be introduced into the image by weighting the MRI pulse sequence towards specific ranges of T1 and T2 by careful timing of the repetition time and the echo time of the sequence. For example, water has a long T1 because water molecules are relatively free to tumble, having relatively few spin–lattice interactions and so they return to the ground state more slowly than, say, larger fat molecules. Thus, in clinical MRI it is possible to weight the image contrast so that T1 weighted images (with a short T1) are too short to record the water signal and so show water as dark, and fat as bright. The interested reader is referred to a recent overview of basic principles of MRI for further reading.90

Use of proton MRI in the lung
The density of the lungs is low compared with other tissues. They have a low proton spin density and therefore offer less MRI signal than other tissues such as brain parenchyma. Furthermore, the static magnetic fields within the thorax vary because of the magnetic susceptibility difference between paramagnetic oxygen in air and diamagnetic tissue, and this effect is compounded by the complicated geometry of the lungs. These field inhomogeneities rapidly dephase any transverse signal, and cause very short transverse relaxation times – short T2. Further problems arise from cardiac and respiratory motion and the lung is a technically challenging organ for 1H MR imaging. The very short T2 of lung tissue means that to provide anatomical detail of the fine structure in the parenchyma, 1H MRI needs to be done at extremely short echo times. The image has to be recorded before the signal produced by decay can become incoherent (i.e. dephased or damped). Therefore, radial projection gradient echo sequences14 100 or short-echo-time fast-spin echo sequences65 have given the best results to date at high field strengths of 1.5 T and above (Fig. 1).



View larger version (83K):
[in this window]
[in a new window]
 
Fig 1 The image on the left is an axial high-resolution CT image of the lungs of a patient with emphysema, acquired during breath-hold. It shows the chest wall anatomy and the fine structure in the parenchyma. On the right is an axial single-shot spin echo MRI scan obtained during a breath-hold on a 1.5 T clinical MRI system. The image delineates soft tissue and thorax, but demonstration of fine structure in the parenchyma is limited.

 
Recent developments with ultrafast steady-state free precession sequences show promise for parenchymal imaging at low field strengths (0.2 T). This is because the T2 of lung tissue is longer at low magnetic field strengths, the signal suffers less disruption of phase and more signal is available in the transverse plane for a longer period of time, with more opportunity to construct an image.86 115 To reduce motion artefacts, these images are best acquired either during a breath-hold with cardiac gating or with both cardiac and respiratory gating if the image sequence is not a single-shot acquisition.

Information on the fine structure of the parenchyma is gradually increasing with proton MRI as technical advances in pulse sequences are made. However, if the lung walls and pulmonary blood vessels are of primary interest then these limitations are less restrictive. For example, the fast acquisition times obtained with gradient and single-shot spin echo sequences lend themselves to dynamic imaging where bulk motion of the chest wall and diaphragm is of primary interest. The fat and bone in and around these structures have a short T1, which can therefore be imaged well with fast sequences (Fig. 2). The movements of these structures and their behaviour during spontaneous breathing have been assessed using proton MRI of the thorax with three-dimensional image reconstruction.29 Similar work has allowed the construction of a two-compartment mathematical model of the ‘inspiratory pump’,10 which could help predict the impact of diseases or therapeutic interventions on the respiratory system. Fetal lung volumes have been measured using proton MRI and an equation to predict normal lung size produced. Comparison with MR scans of fetuses with pulmonary hypoplasia in isolated congenital diaphragmatic hernia74 could predict post-natal pulmonary hypoplasia and poor outcome. Proton MRI has also been used to investigate the role of lung liquid in the presence of respiratory distress syndrome in human neonates. Increased lung liquid was detected in those requiring ventilatory support. At the same time, a reduction in the gravity-dependent gradient in the prone position was detected, suggesting a reduction in atelectasis.13 Dependent atelectasis during anaesthesia in the supine position, its volume and its resolution by a specific recruitment manoeuvre, but not by simple application of continuous positive airway pressure (CPAP), has been shown in children using proton MRI.114 MR has been used to detect lung water in animals. Comparison of dehydrated and hydrated porcine lung parenchyma showed differences in components of the spin relaxation T2 curves that were related to water content.41



View larger version (52K):
[in this window]
[in a new window]
 
Fig 2 A series of ultra-fast proton gradient echo MR images (four images s–1), acquired from a single coronal slice during a respiratory cycle, clearly showing chest wall and diaphragm movement.

 
Oxygen-enhanced 1H MRI
Difficulties in imaging the lung with proton MRI led to a search for alternatives. Oxygen-enhanced proton MRI has been studied as a means of measuring ventilation.46 47 The technique uses fast imaging sequences (such as single-shot fast-spin echo, with an inversion recovery pre-pulse) to introduce some T1 contrast. Molecular oxygen is paramagnetic and passes into the water in the lungs. The T1 of water in the lungs when breathing pure oxygen should be lower than when breathing air. Differences in contrast can be assessed by subtracting the signal intensities measured whilst breathing oxygen and air. The technique requires less equipment compared with hyperpolarized-gas MRI (see below), and can be performed with a standard proton MRI scanner. However, it is uncertain whether the oxygen enhancement truly represents ventilation, because the images are built up as a statistical average from multiple images. Furthermore, it is not clear whether solution of oxygen in interstitial lung water and the vessels, causing a T1 reduction, truly represents regional ventilation.

Contrast agents
The use of contrast agents is typically associated with MR angiography. Solutions of gadolinium-chelated contrast agents (e.g. Gd-DTPA) are injected in to the bloodstream, causing a shortening of the T1 of the blood and vessel enhancement when imaging with fast sequences at short repetition times. Much work has been published on MR pulmonary angiography and it is routinely used for diagnosis of pulmonary embolism and vessel disease.49 89 More recently, gadolinium-chelated contrast agents have been given through the respiratory tract to enhance the intrinsically weak signal from native protons. The mechanism of contrast enhancement is fundamentally the same as with molecular oxygen, causing a shortening of the parenchymal and interstitial T1 of water. These agents have been given by inhalation with and without i.v. contrast to look for matched defects in perfusion and ventilation in experimental models of pulmonary emboli,8 to examine pulmonary veins22 and to assess perfusion in transplanted lungs.15 Initial results in animals were poor because of failure to reach peripheral air spaces. More recent efforts to aerosolize Gd-DTPA demonstrate signal strength increased by more than 100%.54 108

The above is an overview of the advances in lung imaging using native protons in the lung and some of the functional information such as ventilation and perfusion that can be inferred from these images. The problems of intrinsic low spin density and field inhomogeneity will always limit 1H MRI, especially in accurate quantitative estimates of function, and alternatives to proton imaging have been sought, including MRI with noble gases.

MRI with noble gases

Isotopes of certain non-radioactive noble gases with an odd number of nucleons, such as 3He or 129Xe, can be imaged within the air passages using MR. However, the signal is extremely weak because of the low spin density, unless the population of spin states adopted by the nuclear spins can be enhanced by a process of ‘hyperpolarization’,12 yielding ‘magnetized’ noble gas (Fig. 3). This increases the potential MR signal by a factor of 100 000 beyond that achieved normally by thermal equilibrium in the presence of a strong static B0 field,4 as described above for 1H MRI.



View larger version (73K):
[in this window]
[in a new window]
 
Fig 3 The image on the left is a single-shot spin echo 1H MRI obtained while breath-holding on a 1.5 T clinical MRI system. The right-hand image is a low-flip-angle gradient echo image of inhaled 3He, acquired during a breath-hold from the same coronal slice. The image demonstrates the air spaces, rather than the parenchyma and so gives information concerning ventilation.

 
3He can be polarized by two techniques: (i) spin exchange (SE) with an optically pumped rubidium (Rb) vapour,16 or (ii) by metastability exchange (ME) scattering with optically pumped metastable 3He atoms.31 These techniques became feasible with the development of powerful pumping lasers for the necessary wavelengths: 795 nm for Rb and 1083 nm for 3He. The photons emitted by the laser are circularly polarized by filtering with a quarter wave plate that allows a net transfer of angular momentum from the light to the electron and nuclear spins of Rb or 3He.58

SE optical pumping
This technique uses the Rb vapour as an intermediary for transfer of spin to the 3He.83 The laser light excites the valence electron of Rb within a homogenous magnetic field, which collides with 3He, transferring polarization to the helium nuclear spin. This process occurs at a temperature of 80–160 °C under a pressure of 3 bar and the helium is mixed with a small amount of nitrogen. The apparatus can give polarizations of up to 40% and can be adapted to produce hyperpolarized 129Xe.

ME optical pumping
The intermediary Rb is not required. Instead, an RF discharge is used to excite electrons of 3He from their ground state to a metastable state. The laser light can then optically pump these electrons until they are polarized. This polarization is then passed on to the spin of the 3He nucleus of a ground-state atom. However, because this process takes place at low pressure (1 mbar) the gas has to be compressed without losing its polarization and then stored. This process can potentially produce gas that is more than 60% polarized.

Both SE and ME techniques are complex and require skilled maintenance and operation. ME is more difficult to maintain in less sophisticated environments. Although a local production facility is effective, the associated costs (housing, machine, personnel) are considerable. Central on-demand production could be an alternative to avoid start-up costs. Once produced, the polarization decays with time and is disrupted by stray magnetic fields. The helium is therefore transported in glass containers with low iron content and shielded within a stable magnetic field. This concept was successfully tested on a European scale, with delivery of hyperpolarized-3He gas from Mainz via air courier to Sheffield.119 Polarization decreased from 50% to 25–30% in transit, but this is acceptable for imaging.

Although hyperpolarized-xenon MR was applied first to lung air space imaging,5 hyperpolarized helium MR followed soon after.81 The handling of 3He was simpler and the signal change more pronounced. Thus, most subsequent clinical developments have used 3He gas.

Several 3He MRI methods have been developed, each providing specific information. This ‘four-step’ protocol can allow information in ventilation distribution, small airway size and its distribution pattern, gas in-flow and out-flow dynamics, and regional oxygen partial pressure within the ventilated lung spaces. Many of the data are new, since no other techniques give similar (often detailed) information in vivo.

Technical aspects of hyperpolarized-gas MRI
An external source of polarization by optical pumping makes the signal-to-noise ratio less dependent on field strength. Thus, hyperpolarized-3He imaging lends itself to low-field-strength MR applications.37 113 Although this would reduce artefacts during imaging, most of the developments of hyperpolarized-3He MRI have used 1.5 T systems. This is largely because this is the standard field strength in the clinic and the scanner manufacturers concentrate their MR technology developments at this field strength. There are two key constraints in imaging with inhaled hyperpolarized-3He gas. The first is the loss of hyperpolarization of the spin population once magnetization has been tipped into the transverse plane by the RF excitation pulse used in the imaging process. This limits both the size of flip angle and the number of applicable RF pulses (phase encode views) in the sampling of the reciprocal image space. The fact that thermal recovery of polarization is negligible in comparison with the size of the hyperpolarization allows ultra short repetition times. Furthermore, the introduction of oxygen will reduce the T1 relaxation time from several hours to a few seconds, which, combined with the finite time of a patient’s breath-hold, means that fast imaging sequences are required. The second constraint results from the high diffusion coefficient of 3He at atmospheric pressure (D{approx}0.8x10–4 m2 s–1 for 3He in air at standard temperature and pressure). This could potentially lead to significant signal dephasing as a function of the view-dependent MRI gradients.118 In conclusion, the pulse sequences used for standard proton MRI need to be adapted and optimized for efficient use of a hyperpolarized-gas signal.

Administering hyperpolarized 3He
Administration of hyperpolarized 3He can be either by self-inhalation from metered bags or by delivery of a preset volume from an ‘applicator’. Inhalation from metered bags involves a 330/670 mixture of 3He in nitrogen, giving a 1-litre anoxic breath. The nitrogen is used to prolong the lifetime of the 3He signal, since ambient molecular oxygen rapidly ‘relaxes’ hyperpolarization of 3He towards its normal thermal equilibrium. Alternatively, administration can be by a 50–500 ml bolus of 3He within a breath of air via nasal CPAP mask or mouthpiece, using a pneumatically driven, microprocessor-controlled dosing and delivery device (applicator). Self-inhalation from bags is simple, but the volumes and rate of delivery are not well known and it involves inhaling a hypoxic mixture without additional fresh gas. The use of an applicator is a little more complex but allows the administration of a measured 3He dose at an arbitrarily set point during inspiration. Pressures, flows and volumes of the 3He bolus and the carrier breath, as well as gas concentrations, are monitored. Supplemental oxygen, inspiratory pressure support or CPAP can be provided by an attached respirator machine for patients with impaired lung function. Additionally, the applicator device allows recovery of the exhaled 3He, allowing recycling of up to 95% of the inhaled dose of this rare isotope.50

Single-breath static distribution of ventilation in normal subjects, smokers and patients with asthma, cystic fibrosis and COPD
During a breath-hold immediately after inhalation of 300 ml hyperpolarized 3He, normal subjects always show a homogeneous distribution of polarization within the air spaces (Fig. 4). This consistency is important, as it allows the confident identification of abnormalities. Loss of signal implies absent inspiratory gas inflow and hence functional or structural pathology. This reduces inter-observer variation when assessing hyperpolarized-3He images.36



View larger version (93K):
[in this window]
[in a new window]
 
Fig 4 Hyperpolarized helium images: from a normal subject (top left), a smoker with normal lung function tests (bottom right), an asthmatic (top right) and a patient with emphysema (bottom left).

 
Typical patterns of abnormalities include oval defects, which may be predominantly alveolar, and wedge-shaped defects, which may relate to bronchial obstruction. A mixture of oval and wedge-shaped defects are seen in smokers, but may be found even in those with apparently normal lung function (Fig. 4).51 Transient, flitting, peripheral defects are also seen in otherwise healthy subjects with mild ‘hay fever symptoms’.52 3He imaging in asthma shows similar defects (Fig. 4), which resolve with bronchodilator therapy. Larger defects are present in asthmatics with more symptoms and abnormal lung function.7

Early studies comparing proton MR images with 3He images in patients with cystic fibrosis showed disproportionately severe ventilation defects compared with the morphological changes.36 This was particularly the case in upper lung zones. The 3He MR images appeared more sensitive than the Brasfield chest radiography score, and were related to abnormalities in pulmonary function tests.60 Multiple wedge-shaped defects are seen in the bronchiolitis obliterans of chronic rejection following lung transplantation.52

COPD produces defects in the spread of 3He59 60 64 because of restricted entry of a single breath of 3He into already overinflated areas, with long time constants for gas wash-in (Fig. 4). The diffusion characteristics of 3He and the rate of spread of 3He through the lungs add valuable information to the diagnosis and the assessment of severity of the condition.

Regional volumetric imaging with hyperpolarized 3He
Three-dimensional reconstruction of static 3He lung images allows split-lung and even regional volumetry of ventilated lung tissue. 3He volumetry and conventional global spirometry were well correlated, both in healthy subjects and in patients with pulmonary fibrosis and unilateral lung transplants. Only the image-based method, however, quantified the relative contributions of native and grafted lung to total aerated volume.77

Figure 5 shows a surface-rendered three-dimensional reconstruction of aerated lung tissue of a patient with idiopathic pulmonary fibrosis (native right lung) and a single left-lung graft. The larger ventilated volume in the grafted lung is obvious, and a change in calibre is depicted in the left main stem bronchus of the transplanted side.



View larger version (70K):
[in this window]
[in a new window]
 
Fig 5 A surface-rendered three-dimensional reconstruction of the aerated lung volume of a patient with idiopathic pulmonary fibrosis (native right lung) and a single left-lung graft, imaged with 3He 300 ml. The ventilated lung volume of the grafted lung is about 66% and of the native lung is about 34% of the total lung volume. A change in calibre can be seen in the left main bronchus of the transplanted side.77

 
Hyperpolarized 3He has also been used to image the air passages of the nasal cavity, pharyngeal space, trachea9 and has even been used to localize air leaks from pneumothoraces.92

Apparent diffusion coefficient (ADC) measurements with 3He

Atoms can be expected to move with Brownian motion and in a time interval, t, they will move a distance of (2D0t)1/2 along any axis, where D0 is the free diffusion constant, which for 3He in air is 0.88 cm2 s–1.121 This means that the 3He atoms will move about 0.4 mm in 1 ms if they are free to do so. However, in normal lungs the alveolar, acinar and bronchiolar walls restrict this movement, leading to an ADC in healthy human lungs of 0.2 cm2 s–1 and less. Emphysema results in abnormal, permanent enlargement of the air spaces distal to the terminal bronchioles, accompanied by destruction of their walls.102 This extra space in the acinar airways allows room for greater longitudinal and transverse diffusion paths. Destruction of some of the obstacles to diffusion in emphysema means that this value increases to 0.55 cm2 s–1 in severe emphysema.94 Alveoli and the related air passages are not spherical; indeed, if a molecule moves along the axis of an airway it can travel much further without reaching an obstacle compared with when it moves at right angles to the axis. Therefore, diffusion in the airway is anisotropic and it may be possible to model the diffusion behaviour to provide information on the anatomy at the acinar level.

Investigations in guinea-pig lungs confirmed that diffusion constants in the trachea were similar to values in free space; as expected, however, values were much less in distal pulmonary air passages (0.16 cm2 s–1).24 25 27 28 In rats, diffusion was less restricted during inspiration compared with expiration in the normal control rats. The greater inspiratory lung volumes allowed larger diffusion paths compared with expiration. However, in rats with elastase-induced emphysema, ADC was increased and the inspiratory-to-expiratory difference was reduced, suggesting that the walls of the small air spaces had been partially broken down.26 Studies in humans have suggested an increase in ADC with increasing age, which may be related to decreasing elastic recoil.19 In human emphysema, mean ADC may be 2.5 times that in healthy lungs.94 Diffusion MRI with 3He has shown that in normal and emphysematous subjects, 3He diffusivity is anisotropic. The mean longitudinal ADC is three times greater than the mean transverse ADC. However, in normal subjects, transverse diffusion is almost eight times less than the free-3He diffusion coefficient in air (0.1–0.2 cm2 s–1, compared with 0.88 cm2 s–1). In emphysema, both the longitudinal and transverse diffusion coefficients are increased, allowing greater movement in all directions and actually reducing anisotropy.121 An example of an ADC map in a normal volunteer is given in Figure 6; that in Figure 7 is from a patient with smoking-related centri-acinar emphysema.



View larger version (36K):
[in this window]
[in a new window]
 
Fig 6 Diffusion images from a volunteer show homogenous apparent diffusion coefficient values acquired with 3He MRI.

 


View larger version (39K):
[in this window]
[in a new window]
 
Fig 7 Abnormally bright upper zones in both lungs indicating abnormally high apparent diffusion coefficient values in a patient with smoking-related centri-acinar emphysema.

 
Dynamic ventilation MRI with 3He

Initial 3He imaging used ‘static’ breath-holding images. However, imaging during free-breathing ‘dynamic’ imaging can provide information on the dispersion of 3He gas through the lungs and allows visualization of the full respiratory cycle (Fig. 8). This is particularly important in emphysema (to demonstrate flow abnormalities such as air trapping), but can also provide information concerning the trachea and upper airways. In normal subjects, the timing of appearance of peak levels of 3He during MRI after application of a 300 ml bolus was 260 ms in the trachea and 910 ms in the parenchyma.98 In emphysema, dynamic analysis using transaxial slices at three different lung levels showed sequential filling of non-segmental lung regions of variable size. There were even areas where no filling occurred. The signal became more uniform during re-breathing and wash-out phases. Wash-out was relatively slow, with signals persisting in some poorly ventilated areas.46 This may have been compounded by low levels of oxygen in these lung zones, so reducing the rate of loss of 3He MR signal that occurs in the presence of oxygen. The areas of slow uptake and slow wash-out corresponded with severe defects seen on CT. This information, combined with ADC assessments, may be important for early diagnosis and assessment of patients with COPD, particularly those who may warrant bullectomy or even lung volume reduction surgery.



View larger version (16K):
[in this window]
[in a new window]
 
Fig 8 The upper series shows radial projection images (128 views per frame) selected at intervals from the first part of an inhalation in a healthy normal subject. The time course of passage of gas down the trachea into the bronchi is clearly resolved. The lower series shows a dynamic time sequence in a patient with COPD, with areas of slow ventilation in both lungs, particularly in the upper lobes.117

 
More recently, novel imaging sequences have made it possible to obtain images at very short intervals (between 120 ms and 5.4 ms per frame), allowing the analysis of a single breath in slow motion,46 85 95 98 117 which led to the first work to determine regional spirometry.45 77

Intrapulmonary measurement of PO2 with 3He MR

Measurement of partial pressure of oxygen (PO2) in inspired and expired gas, as well as oximetry of arterial and mixed venous blood, allows the study of oxygen uptake in the lung and the estimation of venous admixture (‘shunt’). A much more precise description of the equilibrium between ventilation and perfusion in the lung is possible using the multiple inert gas elimination technique (MIGET). However, none of the techniques that rely on global respiratory and blood gas analysis provides topographical allocation of their information. Split-lung gas analysis and particularly regional alveolar gas sampling and analysis are difficult. Therefore, alveolar and corresponding end-capillary PO2 are usually derived indirectly – from inspired PO2, arterial carbon dioxide partial pressure and the gas exchange ratio.73

Researchers soon noticed in the development of 3He MRI that during repetitive MRI of inhaled 3He, signal intensity disappeared rapidly, even if the breath is held and no 3He is exhaled. Two major reasons apply: first, the non-equilibrium (hyper-) polarization of 3He decays (depolarizes) during MRI because of the effect of the RF pulses required for image acquisition (the magnetic gas is ‘shaken’), and secondly, the interaction with paramagnetic oxygen molecules (dipoles) in the alveolar gas destroys 3He hyperpolarization progressively within about the time of a breath-hold (‘relaxation’). The rate of this signal decay R(t) depends on the ambient PO2, according to the proportional relationship R(t)=PO2/{zeta}, with a so-called oxygen-induced longitudinal relaxation time constant {zeta}=2.61 s.bar, known from in vitro experiments.93 During early attempts at lung imaging, signal loss resulting from alveolar oxygen was regarded as a nuisance, and self-experimenting physicist researchers breathed pure nitrogen before the 3He in order to get ‘nicer’ images of the lung. Soon, however, the potential to use this oxygen sensitivity of hyperpolarized 3He for instantaneous, non-invasive and regional intrapulmonary PO2 measurement was realized.35 40 An estimate of local alveolar partial pressure of oxygen (PAO2) is of physiological interest since the latter is uniquely dependent on the ventilation/perfusion ratio prevailing in the imaged region. Technically, the contribution of the PO2 effect to the total signal decay rate can be isolated quite simply from that of the imaging process, and can be quantified. For instance, the kinetics of signal decay may be obtained from a series of fast low angle shot (FLASH) images. During image acquisition, one of the two major determinants of relaxation (i.e. flip angle [RF voltage] or exposure time to oxygen [interval between consecutive images]) is varied methodically and its relative contribution to relaxation assessed.

To improve accuracy, other minor determinants of 3He signal attenuation must be taken into account or minimized, such as the relaxation properties of the surfaces that come into contact with the 3He40 and the convective and diffusive movement of 3He out of the imaged field. Meanwhile, several techniques have been described to determine regional PO2 at the very beginning of a breath-hold manoeuvre as well as the rate of its decrease during breath-holding.35 40 This rate of decrease is the net result of both alveolocapillary oxygen transfer and non-ventilatory mass movement of fresh gas to replace it. This has been developed in lung phantoms, pigs and human volunteers breathing various fractional inspired oxygen concentrations.35 40 In healthy subjects’ lungs, PO2 distribution determined by 3He MRI appears very homogeneous, with its mean close to end-tidal PO2 (Fig. 9).40 Patients with idiopathic pulmonary fibrosis and a single-lung transplant ation, serving as a model of disparate split-lung function, showed greater MRI-determined PAO2 in functioning grafts than in the remaining native lungs. This was compatible with the restrictive abnormalities and reduced ventilation/perfusion ratio in the native lung. Mean image-based PAO2 of these patients ranged between PO2 levels determined in end-tidal gas and arterial blood.38 The measurement of intrapulmonary PO2 may thus have further important clinical implications. If an area of lung is poorly perfused, then the rate of uptake of oxygen by the blood passing through vessels in that area will be low. Therefore, PO2 will be relatively high in such an area and would fall relatively slowly on breath-holding. This raises the possibility that this technique could detect areas of ventilation/perfusion mismatch, such as could be caused by pulmonary embolism. Indeed, in an animal model of pulmonary embolism, alveolar deadspace generated by pulmonary arterial obstruction is depicted topographically correctly by image-based PAO2 determination.39 Other researchers have also derived PO2 from intrapulmonary 3He relaxation in small animals in vivo and post mortem, and have estimated pulmonary oxygen uptake and cardiac output from their data.84 Meanwhile, intrapulmonary PO2 measurements have also been performed successfully in small animals using low-field MRI.88



View larger version (47K):
[in this window]
[in a new window]
 
Fig 9 Maps of intrapulmonary PO2 distribution are illustrated at the very start of breath-holding. Each breath-hold was preceded by wash-in of fresh gas with different fractional inspired oxygen concentrations, resulting in stable end-expiratory PO2 levels before the imaged breath-hold. End-expiratory PO2 is measured immediately before the imaged breath-hold. A 250 ml bolus of 3He has then been inhaled at the front of a tidal inspiration containing the respective fresh gas mixture. Image data are derived from serial 3He projection imaging using a single acquisition technique in a healthy subject.

 
Thus, in the future oxygen-sensitive 3He MRI may become a quick, non-invasive and easily repeatable method to assess ventilation/perfusion matching, and to estimate oxygen uptake from the lung.

Low/ultra-low field imaging with 3He
MR, and in particular MR at high field strength like that used in the clinical 1.5 T scanners, is complex and expensive. However, because of the artificially increased spin density, hyperpolarized gas imaging does not require strong B0 fields, and hence smaller and cheaper systems can be used, with permanent magnets less than 0.3 T field strength. It is therefore possible to image the lung using low-strength magnetic fields of around 0.1 T, with advantages in cost, size of equipment and patient comfort, with no worsening of signal-to-noise ratio.33 Work in this area is part of an ongoing EU sponsored project (www.phil.ens.fr).

3He microbubbles and vascular imaging
Hyperpolarized 3He provides a strong signal in the lung air spaces, but is not soluble in blood so other techniques or contrast media are used to image the vasculature. However, because the signal from 3He is very strong, attempts have been made to create suspensions of 3He microbubbles to enable vascular imaging.23 The rapid deterioration in polarization would remove the problem of recirculation of signal leading to a second pass. Initial animal studies suggest possible high quality vascular imaging.

Hyperpolarized xenon
129Xe is another inert noble gas that can be hyperpolarized and used for lung imaging.5 It behaves in a similar way to 3He in the presence of oxygen. It is hyperpolarized using a Rb SE process similar to that described with 3He. However, 3He is easier to polarize and it yields a stronger signal than xenon, although xenon is more abundant. Following initial studies in mice, xenon has been used in humans.34 An interesting feature of xenon is its solubility in blood, which allows it to pass around the body. Some will enter the nervous system because of its lipid solubility. A drawback is that the polarization decreases rapidly, thus allowing only 15 s of effective use. Models of the compartmental distribution of xenon have been developed, and could allow study of the brain and its behaviour. Intriguingly, because of the anaesthetic properties of xenon, there may be the possibility of gaining further insight into the mechanisms of anaesthesia.

Lung imaging with 19F
With the use of appropriately resonant MR coils and amplifier equipment, 19F atoms bound within highly fluorine-substituted molecules (perfluorocarbons [PFC], SF6 gas, and in principle, fluorine-containing volatile anaesthetics) can be imaged. Oxygen-carrying blood substitutes have been developed from PFC. In addition, since dissolved oxygen has paramagnetic properties and also affects the T1 of fluorine nuclei, this could allow MRI of PO2 effects. This opens up the possibility of using 19F as means of ventilation imaging of the lung, and of tracking oxygen levels in the lung through to other tissues such as liver and spleen.111 Inhaled SF6 has recently been used for morphological and ventilation imaging in pigs.97 Also, a PFC solution employed in partial liquid ventilation has been imaged and used for regional measurement of PO2 within the alveolar PFC phase.72

Conclusion

Functional imaging of the lung is now possible without the use of radioactive agents or ionizing radiation. These techniques give information on a macro scale, demonstrating areas of poor ventilation in the lung, and also show changes at the microscopic level. PO2 can be imaged at sites within the lung and this allows ventilation and perfusion information to be gathered using hyperpolarized gases. In the future, xenon could also allow investigation of other organs, in particular the brain, while MRI of 19F-substituted compounds may allow tissue oxygenation to be studied within a wide range of organs. Detailed information of previously inaccessible sites and non-invasive detection of pathology in the lungs before conventional lung function tests become abnormal is now almost with us. In contrast to previous years, the future of lung MRI is bright (see Fig. 1).

References

1 Adams EW, Counsell SJ, Cox P, et al. Investigation of the role of lung liquid in the pathogenesis of lung disease in the preterm infant using magnetic resonance imaging. Early Hum Dev 2000; 58: 76[CrossRef][Medline]

2 Adams EW, Counsell SJ, Hajnal JV, et al. Investigation of lung disease in preterm infants using magnetic resonance imaging. Biol Neonat 2000; 77 [Suppl. 1]: 17–20

3 Adams EW, Counsell SJ, Hajnal JV, et al. Magnetic resonance imaging of lung water content and distribution in term and preterm infants. Am J Respir Crit Care Med 2002; 166: 397–402[Abstract/Free Full Text]

4 Albert MS, Balamore D. Development of hyperpolarised noble gas MRI. Nucl Instrum Methods Phys Res A 1998; 402: 441–53[ISI][Medline]

5 Albert MS, Cates GD, Driehuys B, et al. Biological magnetic resonance imaging using laser-polarized 129Xe. Nature 1994; 370: 199–201[CrossRef][ISI][Medline]

6 Alexander C. Diaphrgam movements and the diagnosis of diaphragmatic paralysis. Clin Radiol 1966; 17: 79–83[Medline]

7 Altes TA, Powers PL, Knight-Scott J, et al. Hyperpolarized 3He MR lung ventilation imaging in asthmatics: preliminary findings. J Magn Reson Imaging 2001; 13: 378–84[CrossRef][ISI][Medline]

8 Amundsen T, Kvaerness J, Jones RA, et al. Pulmonary embolism: detection with MR perfusion imaging of the lung – a feasibility study. Radiology 1997; 203: 181–5[Abstract]

9 Bachert P, Schad LR, Bock M, et al. Nuclear magnetic resonance imaging of airways in humans with use of hyperpolarized 3He. Magn Reson Med 1996; 36: 192–6[ISI][Medline]

10 Basso Ricci S, Cluzel P, Constantinescu A, Similowski T. Mechanical model of the inspiratory pump. J Biomech 2002; 35: 139–45[CrossRef][ISI][Medline]

11 Bayat S, Le Duc G, Porra L, et al. Quantitative functional lung imaging with synchrotron radiation using inhaled xenon as contrast agent. Phys Med Biol 2001; 46: 3287–99[CrossRef][ISI][Medline]

12 Becker J, Bermuth J, Ebert M, et al. Interdisciplinary experiments with polarised 3He. Nucl Instrum Methods Phys Res A 1998; 402: 327–36[ISI]

13 Beigelman-Aubry C, Capderou A, Grenier PA, et al. Mild intermittent asthma: CT assessment of bronchial cross-sectional area and lung attenuation at controlled lung volume. Radiology 2002; 223: 181–7[Abstract/Free Full Text]

14 Bergin CJ, Noll DC, Pauly JM, Glover GH, Macovski A. MR imaging of lung parenchyma: a solution to susceptibility. Radiology 1992; 183: 673–6[Abstract]

15 Berthezene Y, Croisille P, Bertocchi M, et al. Lung perfusion demonstrated by contrast enhanced dynamic magnetic resonance imaging. Application to unilateral lung transplantation. Invest Radiol 1997; 32: 351–6[CrossRef][ISI][Medline]

16 Bouchiat MA, Carver TR, Varnum CM. Nuclear polarisation in 3He gas induced by optical pumping and dipolar exchange. Phys Rev Lett 1960; 5: 373–5[CrossRef][ISI]

17 Brasch RC, Gould RG, Gooding CA, Ringertz HG, Lipton MJ. Upper airway obstruction in infants and children: evaluation with ultrafast CT. Radiology 1987; 165: 459–66[Abstract]

18 Brody AS, Kuhn JP, Seidel FG, Brodsky LS. Airway evaluation in children with use of ultrafast CT: pitfalls and recommendations. Radiology 1991; 178: 181–4[Abstract]

19 Brookeman JR, Mugler JP, 3rd, Knight-Scott J, et al. Studies of 3He diiffusion coefficient in the human lung. Age related distribution patterns. Eur Radiol 1999; 9: B21

20 Brown BH, Barber DC, Seagar AD. Applied potential tomography: possible clinical applications. Clin Phys Physiol Meas 1985; 6: 109–21[CrossRef][ISI][Medline]

21 Brown BH, Barber DC, Wang W, et al. Multi-frequency imaging and modelling of respiratory related electrical impedance changes. Physiol Meas 1994; 15 [Suppl 2a]: A1–12[CrossRef][ISI]

22 Carriero A, Marano R, Fossaceca R, Magarelli N, Bonomo L. Pulmonary veins: magnetic resonance angiography anatomy. Magn Reson Mat Biol Physics Med 1997; 6: 2–6

23 Chawla MS, Chen XJ, Moller HE, et al. In vivo magnetic resonance vascular imaging using laser-polarized 3He microbubbles. Proc Natl Acad Sci USA 1998; 95: 10832–5[Abstract/Free Full Text]

24 Chen XJ, Chawla MS, Cofer GP, et al. Hyperpolarized 3He NMR lineshape measurements in the live guinea pig lung. Magn Reson Med 1998; 40: 61–5[ISI][Medline]

25 Chen XJ, Chawla MS, Hedlund LW, et al. MR microscopy of lung airways with hyperpolarized 3He. Magn Reson Med 1998; 39: 79–84[ISI][Medline]

26 Chen XJ, Hedlund LW, Moller HE, et al. Detection of emphysema in rat lungs by using magnetic resonance measurements of 3He diffusion. Proc Natl Acad Sci USA 2000; 97: 11478–81[Abstract/Free Full Text]

27 Chen XJ, Moller HE, Chawla MS, et al. Spatially resolved measurements of hyperpolarized gas properties in the lung in vivo. Part I: diffusion coefficient. Magn Reson Med 1999; 42: 721–8[CrossRef][ISI][Medline]

28 Chen XJ, Moller HE, Chawla MS, et al. Spatially resolved measurements of hyperpolarized gas properties in the lung in vivo. Part II: T*(2). Magn Reson Med 1999; 42: 729–37[CrossRef][ISI][Medline]

29 Cluzel P, Similowski T, Chartrand-Lefebvre C, et al. Diaphragm and chest wall: assessment of the inspiratory pump with MR imaging-preliminary observations. Radiology 2000; 215: 574–83[Abstract/Free Full Text]

30 Cohen MA, Benditt JO, Eveloff S, McCool FD. Diaphragm thickness during inspiration. J Appl Physiol 1997; 83: 291–6[Abstract/Free Full Text]

31 Colegrove FD, Schearer LD, Walters GK. Polarisation of 3He gas by optical pumping. Physiol Rev 1963; 132: 2561–72[CrossRef][ISI]

32 Cullen SC, Gross EG. The anesthetic properties of xenon in animals and human beings, with additional observations on krypton. Science 1951; 113: 580–2[ISI]

33 Darrasse L, Guillot G, Nacher PJ, Tastevin G. Low field 3He nuclear magnetic resonance in human lungs. C R Acad Sci Paris 1997; 324: 691–700

34 de Lange EE, Brookeman JR, Mugler JP, 3rd, et al. Proc RSNA, 82nd annual meeting, 1996

35 Deninger AJ, Eberle B, Ebert M, et al. Quantification of regional intrapulmonary oxygen partial pressure evolution during apnea by (3)He MRI. J Magn Reson 1999; 141: 207–16[CrossRef][ISI][Medline]

36 Donnelly LF, MacFall JR, McAdams HP, et al. Cystic fibrosis: combined hyperpolarized 3He-enhanced and conventional proton MR imaging in the lung – preliminary observations. Radiology 1999; 212: 885–9[Abstract/Free Full Text]

37 Durand E, Guillot G, Darrasse L, et al. CPMG measurements and ultrafast imaging in human lungs with hyperpolarized helium-3 at low field (0.1 T). Magn Reson Med 2002; 47: 75–81[CrossRef][ISI][Medline]

38 Eberle B, Markstaller K, Lill J, et al. Oxygen-sensitive 3He magnetic resonance imaging of the lungs in patients after unilateral lung transplantation. Am J Resp Crit Care Med 2000; 161: A718

39 Eberle B, Markstaller K, Stepniak A, Viallon M, Kauczor H-U. 3Helium-MRI-based assessment of regional gas exchange impairment during experimental pulmonary artery occlusion. Anesthesiology 2002; 96: A1309[CrossRef]

40 Eberle B, Weiler N, Markstaller K, et al. Analysis of intrapulmonary O(2) concentration by MR imaging of inhaled hyperpolarized helium-3. J Appl Physiol 1999; 87: 2043–52[Abstract/Free Full Text]

41 Estilaei M, MacKay A, Roberts C, Mayo J. 1H NMR measurements of wet/dry ratio and T1 T2 distributions in lung. J Magn Reson 1997; 124: 410–9[CrossRef][ISI][Medline]

42 Frerichs I, Hinz J, Herrmann P, et al. Regional lung perfusion as determined by electrical impedance tomography in comparison with electron beam CT imaging. IEEE Trans Med Imaging 2002; 21: 646–52[CrossRef][ISI][Medline]

43 Furukawa M, Kashiwagi S, Matsunaga N, et al. Evaluation of cerebral perfusion parameters measured by perfusion CT in chronic cerebral ischemia: comparison with xenon CT. J Comput Assist Tomogr 2002; 26: 272–8[CrossRef][ISI][Medline]

44 Galvin JR, Rooholamini SA, Stanford W. Obstructive sleep apnea: diagnosis with ultrafast CT. Radiology 1989; 199: 825–9

45 Gast KK, Puderbach MU, Rodriguez I, et al. Dynamic ventilation (3)He-magnetic resonance imaging with lung motion correction: gas flow distribution analysis. Invest Radiol 2002; 37: 126–34[CrossRef][ISI][Medline]

46 Gierada DS, Saam B, Yablonskiy D, et al. Dynamic echo planar MR imaging of lung ventilation with hyperpolarized (3)He in normal subjects and patients with severe emphysema. NMR Biomed 2000; 13: 176–81[CrossRef][ISI][Medline]

47 Goodman LR, Lipchik RJ, Kuzo RS, et al. Subsequent pulmonary embolism: risk after a negative helical CT pulmonary angiogram – prospective comparison with scintigraphy. Radiology 2000; 215: 535–42[Abstract/Free Full Text]

48 Gottesmore E, McCool FD. Ultrasound evaluation of the paralysed diaphragm. Am J Respir Crit Care Med 1997; 155: 1570–4[Abstract]

49 Goyen M, Laub G, Ladd ME, et al. Dynamic 3D MR angiography of the pulmonary arteries in under four seconds. J Magn Reson Imaging 2001; 13: 372–7[CrossRef][ISI][Medline]

50 Grossmann T. Realisierung des 3He-Kreislaufs zur 3He-Magnet-Resonanz-Tomographie. Mainz: Institute of Physics, Johannes Gutenberg University, 2000

51 Guenther D, Eberle B, Hast J, et al. (3)He MRI in healthy volunteers: preliminary correlation with smoking history and lung volumes. NMR Biomed 2000; 13: 182–9[CrossRef][ISI][Medline]

52 Guenther D, Hanisch G, Kauczor HU. Functional MR imaging of pulmonary ventilation using hyperpolarized noble gases. Acta Radiol 2000; 41: 519–28[CrossRef][ISI][Medline]

53 Gur D, Drayer BP, Borovetz HS, et al. Dynamic computed tomography of the lung: regional ventilation measurements. J Comput Assist Tomogr 1979; 3: 749–53[ISI][Medline]

54 Haage P, Adams G, Karaagsac S, et al. Mechanical delivery of aerosolised gadolinium-DTPA for pulmonary ventilation assessment in MR imaging. Invest Radiol 2001; 36: 240–3[CrossRef][ISI][Medline]

55 Houston JG, Angus RM, Cowan MD, McMillan NC, Thomson NC. Ultrasound assessment of normal hemidiaphragmatic movement: relation to inspiratory volume. Thorax 1994; 49: 500–3[Abstract]

56 Houston JG, Fleet M, Cowan MD, McMillan NC. Comparison of ultrasound with fluoroscopy in the assessment of suspected hemidiaphragmatic movement abnormality. Clin Radiol 1995; 50: 95–8[ISI][Medline]

57 Hu S, Hoffman EA, Reinhardt JM. Automatic lung segmentation for accurate quantitation of volumetric X-ray CT images. IEEE Trans Med Imaging 2001; 20: 490–8[CrossRef][ISI][Medline]

58 Kastler A. Quelques suggestions concernant la production optique et la detection optique d‘une inégalité de population des niveaux de quantification spatiale des atomes. Application à l‘expérience de Stern et Gerlach et à la resonance magnétique. J Phys Radium 1950; 11: 255–65[ISI]

59 Kauczor H, Surkau R, Roberts T. MRI using hyperpolarized noble gases. Eur Radiol 1998; 8: 820–7[CrossRef][ISI][Medline]

60 Kauczor HU, Ebert M, Kreitner KF, et al. Imaging of the lungs using 3He MRI: preliminary clinical experience in 18 patients with and without lung disease. J Magn Reson Imaging 1997; 7: 538–43[Medline]

61 Kauczor HU, Hast J, Heussel CP, et al. Assessment of lung volumes using helical CT at inspiration and expiration: comparison with pulmonary function tests. Am J Roentgenol 1998; 171: 1091–5[Abstract]

62 Kauczor HU, Hast J, Heussel CP, et al. Focal airtrapping at expiratory high resolution CT: comparison with pulmonary function tests. Eur J Radiol 2000; 10: 1539–46[CrossRef]

63 Kauczor HU, Hast J, Heussel CP, et al. CT attenuation of paired HRCT scans obtained at full inspiratory/expiratory position: comparison with pulmonary function tests. Eur Radiol 2002; 12: 2757–63[ISI][Medline]

64 Kauczor HU, Kreitner KF. MRI of the pulmonary parenchyma. Eur Radiol 1999; 9: 1755–64[CrossRef][ISI][Medline]

65 Knight-Scott J, Keilholz-George SD, Mai VM, Christopher JM. Temporal dynamics of blood flow effects in half-Fourier fast spin echo (1)H magnetic resonance imaging of the human lungs. J Magn Reson Imaging 2001; 14: 411–18[CrossRef][ISI][Medline]

66 Kreck TC, Krueger MA, Altemeier WA, et al. Determination of regional ventilation and perfusion in the lung using xenon and computed tomography. J Appl Physiol 2001; 91: 1741–9[Abstract/Free Full Text]

67 Kunst PW, Bohm SH, Vazquez de Anda G, et al. Regional pressure volume curves by electrical impedance tomography in a model of acute lung injury. Crit Care Med 2000; 28: 178–83[CrossRef][ISI][Medline]

68 Kunst PW, de Vries PM, Postmus PE, Bakker J. Evaluation of electrical impedance tomography in the measurement of PEEP-induced changes in lung volume. Chest 1999; 115: 1102–6[Abstract/Free Full Text]

69 Kunst PW, Vazquez de Anda G, Bohm SH, et al. Monitoring of recruitment and derecruitment by electrical impedance tomography in a model of acute lung injury. Crit Care Med 2000; 28: 3891–5[ISI][Medline]

70 Lamers RJ, Kemerink GJ, Drent M, van Engelshoven JM. Reproducibility of spirometrically controlled CT lung densitometry in a clinical setting. Eur Respir J 1998; 11: 942–5[Abstract/Free Full Text]

71 Lamers RJ, Thelissen GR, Kessels AG, Wouters EF, van Engelshoven JM. Chronic obstructive pulmonary disease: evaluation with spirometrically controlled CT lung densitometry. Radiology 1994; 193: 109–13[Abstract]

72 Laukemper-Ostendorf S, Scholz A, Burger K, et al. 19F-MRI of perflubron for measurement of oxygen partial pressure in porcine lungs during partial liquid ventilation. Magn Reson Med 2002; 47: 82–9[CrossRef][ISI][Medline]

73 Lumb AB. Nunn’s Applied Respiratory Physiology. Oxford: Butterworth Heinemann, 2000

74 Mahieu-Caputo D, Sonigo P, Dommergues M, et al. Fetal lung volume measurement by magnetic resonance imaging in congenital diaphragmatic hernia. Br J Obstet Gynaecol 2001; 108: 863–8[CrossRef][ISI]

75 Markstaller K, Eberle B, Kauczor HU, et al. Temporal dynamics of lung aeration determined by dynamic CT in a porcine model of ARDS. Br J Anaesth 2001; 87: 459–68[Abstract/Free Full Text]

76 Markstaller K, Karmrodt J, Doebrich M, et al. Dynamic computed tomography: a novel technique to study lung aeration and atelectasis formation during experimental CPR. Resuscitation 2002; 53: 307–13[CrossRef][ISI][Medline]

77 Markstaller K, Kauczor HU, Puderbach M, et al. 3He-MRI-based vs. conventional determination of lung volumes in patients after unilateral lung transplantation: a new approach to regional spirometry. Acta Anaesthesiol Scand 2002; 46: 845–52[CrossRef][ISI][Medline]

78 Mayo JR, Webb WR, Gould R, et al. High-resolution CT of the lungs: an optimal approach. Radiology 1987; 163: 507–10[Abstract]

79 McCool FD, Benditt JO, Conomos P, et al. Variability of diaphragm structure among healthy individuals. Am J Respir Crit Care Med 1997; 155: 1323–8[Abstract]

80 McCool FD, Conomos P, Benditt JO, et al. Maximal inspiratory pressures and dimensions of the diaphragm. Am J Respir Crit Care Med 1997; 155: 1329–34[Abstract]

81 Middleton H, Black RD, Saam B, et al. MR imaging with hyperpolarized 3He gas. Magn Reson Med 1995; 33: 271–5[ISI][Medline]

82 Mills GH, Kyroussis D, Hamnegard CH, et al. Unilateral magnetic stimulation of the phrenic nerve. Thorax 1995; 50: 1162–72[Abstract]

83 Miron E, Schaefer S, Schreiber D, et al. Polarisation of the nuclear spins of noble-gas atoms by spin exchange with optically pumped alkali-metal atoms. Phys Rev 1984; A29: 3092–110

84 Moeller H, Hedlund L, Chen X, et al. Measurements of hyperpolarized gas properties in the lung. Part III: 3He T1. Magn Reson Med 2001; 45: 421–30[CrossRef][ISI][Medline]

85 Mugler JP, Brookeman JR, Knight-Scott J, et al. Interleaved echo-planar imaging of the lungs with hyperpolarized 3He. Proc ISMRM 1998; 448

86 Muller CJ, Loffler R, Deimling M, Peller M, Reiser M. MR lung imaging at 0.2 T with T1-weighted true FISP: native and oxygen-enhanced. J Magn Reson Imaging 2001; 14: 164–8[CrossRef][ISI][Medline]

87 Newsom-Davies J, Goldman M, Loh M, Casson M. Diaphragm function and alveolar hypoventilation. Q J Med 1976; 45: 87–100[ISI][Medline]

88 Olsson L, Magnusson P, Deninger A, et al. Intrapulmonary pO2 measured by low field MR imaging of hyperpolarized 3He. Proc Int Soc Magn Reson Med 2002; 10: 2021

89 Oudkerk M, van Beek EJ, Wielopolski P, et al. Comparison of contrast-enhanced magnetic resonance angiography and conventional pulmonary angiography for the diagnosis of pulmonary embolism: a prospective study. Lancet 2002; 359: 1643–7[CrossRef][ISI][Medline]

90 Paley MNJ, Wilkinson ID, van Beek EJR, Griffiths PD. Magnetic Resonance Imaging. In: Grainger RE, Allison DJ, Adam A, Dixon AK, eds. Grainger and Allison’s Textbook of Diagnostic Radiology. London: Churchill Livingstone, 2001; 101–36

91 Prokop M. Multislice CT angiography. Eur J Radiol 2000; 36: 86–96[CrossRef][ISI][Medline]

92 Roberts DA, Rizi RR, Lipson DA, et al. Detection and localization of pulmonary air leaks using laser-polarized (3)He MRI. Magn Reson Med 2000; 44: 379–82[CrossRef][ISI][Medline]

93 Saam B, Happer W, Middleton H. Nuclear relaxation of 3He in the presence of O2. Phys Rev 1995; 52: 862[ISI]

94 Saam BT, Yablonskiy DA, Kodibagkar VD, et al. MR imaging of diffusion of (3)He gas in healthy and diseased lungs. Magn Reson Med 2000; 44: 174–9[CrossRef][ISI][Medline]

95 Salerno M, Altes TA, Brookeman JR, de Lange EE, Mugler JP, 3rd. Dynamic spiral MRI of pulmonary gas flow using hyperpolarized (3)He: preliminary studies in healthy and diseased lungs. Magn Reson Med 2001; 46: 667–77[CrossRef][ISI][Medline]

96 Sase S, Honda M, Kushida T, et al. Quantitative cerebral blood flow calculation method using white matter lambda in xenon CT. J Comput Assist Tomogr 2002; 26: 471–8[CrossRef][ISI][Medline]

97 Schreiber WG, Eberle B, Laukemper-Ostendorf S, et al. Dynamic (19)F-MRI of pulmonary ventilation using sulfur hexafluoride (SF(6)) gas. Magn Reson Med 2001; 45: 605–13[CrossRef][ISI][Medline]

98 Schreiber WG, Weiler N, Kauczor HU, et al. Ultraschnelle MRT der Lungenventilation mittels hochpolarisiertem Helium-3. ROFO Fortschr Geb Röntgenstr Neuen Bildgeb Verfahr 2000; 172: 129–33[CrossRef][ISI][Medline]

99 Serrano RE, de LB, Casas O, et al. Use of electrical impedance tomography (EIT) for the assessment of unilateral pulmonary function. Physiol Meas 2002; 23: 211–20[CrossRef][ISI][Medline]

100 Shattuck MD, Gewalt SL, Glover GH, Hedlund LW, Johnson GA. MR micro imaging of the lung using volume projection encoding. Magn Reson Med 1997; 38: 938–42[ISI][Medline]

101 Simon BA, Marcucci C, Fung M, Lele SR. Parameter estimation and confidence intervals for Xe-CT ventilation studies: a Monte Carlo approach. J Appl Physiol 1998; 84: 709–16[Abstract/Free Full Text]

102 Snider GL, Kleinerman J, Thurlbeck WM, Bengali ZH. The definition of emphysema. Report of a National Heart and Lung and Blood Institute, Division of Lung Diseases Workshop. Am Rev Respir Dis 1985; 132: 182–5[ISI][Medline]

103 Stern EJ, Frank MS. CT of the lung in patients with pulmonary emphysema: diagnosis, quantification, correlation with pathologic and physiologic findings. Am J Roentgenol 1994; 162: 791–8[Abstract]

104 Stern EJ, Frank MS. Small-airway diseases of the lungs: findings at expiratory CT. Am J Roentgenol 1994; 163: 37–41[Abstract]

105 Stern EJ, Graham CM, Webb WR, Gamsu G. Normal trachea during forced expiration: dynamic CT measurements. Radiology 1993; 187: 27–31[Abstract]

106 Stern EJ, Webb WR. Dynamic imaging of lung morphology with ultrafast high-resolution computed tomography. J Thorac Imaging 1993; 8: 273–82[Medline]

107 Stern EJ, Webb WR, Gamsu G. Dynamic quantitative computed tomography. A predictor of pulmonary function in obstructive lung diseases. Invest Radiol 1994; 29: 564–9[ISI][Medline]

108 Suga K, Ogasawara N, Okada M, Matsunaga N, Arai M. Regional lung functional impairment in acute airway obstruction and pulmonary embolic dog models assessed with gadolinium-based aerosol ventilation and perfusion magnetic resonance imaging. Invest Radiol 2002; 37: 281–91[CrossRef][ISI][Medline]

109 Tajik JK, Chon D, Won C, Tran BQ, Hoffman EA. Subsecond multisection CT of regional pulmonary ventilation. Acad Radiol 2002; 9: 130–46[CrossRef][ISI][Medline]

110 Taplin GV, Poe ND. A dual lung-scanning technique for evaluation of pulmonary function. Radiology 1965; 85: 365–8[ISI]

111 Thomas SR, Pratt RG, Millard RW, et al. In vivo pO2 imaging in the porcine model with perfluorocarbon F19 NMR at low field. Magn Reson Imaging 1996; 14: 103–14[CrossRef][ISI][Medline]

112 Tillie-Leblond I, Mastora I, Radenne F, et al. Risk of pulmonary embolism after a negative spiral CT angiogram in patients with pulmonary disease: 1-year clinical follow-up study. Radiology 2002; 223: 461–7[Abstract/Free Full Text]

113 Tseng CH, Wong GP, Pomeroy VR, et al. Low-field MRI of laser polarized noble gas. Phys Rev Lett 1998; 81: 3785–8[CrossRef][ISI][Medline]

114 Tusman G, Böhm SH, Tempra A, et al. Effects of recruitment maneuver on atelectasis in anesthetized children. Anesthesiology 2002; 98: 14–22[CrossRef]

115 Wagner M, Bowing B, Kuth R, et al. Low field thoracic MRI – a fast and radiation free routine imaging modality in children. Magn Reson Imaging 2001; 19: 975–83[CrossRef][ISI][Medline]

116 Webb WR, Stern EJ, Kanth N, Gamsu G. Dynamic pulmonary CT: findings in healthy adult men. Radiology 1993; 186: 117–24[Abstract]

117 Wild JM, Paley MN, Kasuboski L, et al. Dynamic radial projection MRI of inhaled hyperpolarized 3He gas. Magn Reson Med 2003; 49: 991–7[CrossRef][ISI][Medline]

118 Wild JM, Paley MN, Viallon M, et al. k-space filtering in 2D gradient-echo breath-hold hyperpolarized 3He MRI: spatial resolution and signal-to-noise ratio considerations. Magn Reson Med 2002; 47: 687–95[CrossRef][ISI][Medline]

119 Wild JM, Schmiedeskamp J, Paley MN, et al. MR imaging of the lungs with hyperpolarized helium-3 gas transported by air. Phys Med Biol 2002; 47: N185–90[CrossRef][ISI][Medline]

120 Won C, Chon D, Tajik J, et al. CT-based assessment of regional pulmonary microvascular blood flow parameters. J Appl Physiol 2003; 94: 2483–93[Abstract/Free Full Text]

121 Yablonskiy DA, Sukstanskii AL, Leawoods JC, et al. Quantitative in vivo assessment of lung microstructure at the alveolar level with hyperpolarized 3He diffusion MRI. Proc Natl Acad Sci USA 2002; 99: 3111–16[Abstract/Free Full Text]

122 Yagi M, Mashimo T, Kawaguchi T, Yoshiya I. Analgesic and hypnotic effects of subanaesthetic concentrations of xenon in human volunteers: comparison with nitrous oxide. Br J Anaesth 1995; 74: 670–3[Abstract/Free Full Text]

123 Yamaguchi K, Soejima K, Koda E, Sugiyama N. Inhaling gas with different CT densities allows detection of abnormalities in the lung periphery of patients with smoking-induced COPD. Chest 2001; 120: 1907–16[Abstract/Free Full Text]





This Article
Full Text (PDF)
E-Letters: Submit a response to the article
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (2)
Disclaimer
Request Permissions
Google Scholar
Articles by Mills, G. H.
Articles by Van Beek, E. J. R.
PubMed
PubMed Citation
Articles by Mills, G. H.
Articles by Van Beek, E. J. R.