Overview of MRI for pulmonary functional imaging

Abstract

Morphological evaluation of the lung is important in the clinical evaluation of pulmonary diseases. However, the disease process, especially in its early phases, may primarily result in changes in pulmonary function without changing the pulmonary structure. In such cases, the traditional imaging approaches to pulmonary morphology may not provide sufficient insight into the underlying pathophysiology. Pulmonary imaging community has therefore tried to assess pulmonary diseases and functions utilizing not only nuclear medicine, but also CT and MR imaging with various technical approaches. In this review, we overview state-of-the art MR methods and the future direction of: (1) ventilation imaging, (2) perfusion imaging and (3) biomechanical evaluation for pulmonary functional imaging.

Introduction

Morphological evaluation of the lung is important for the evaluation of patients with pulmonary diseases because of the close relationship between structural changes of the lung and pathophysiology of such diseases. These correlations between pulmonary morphology and functions are fundamental for diagnostic imaging modalities including chest radiography and CT, which play a significant role in the diagnosis and management of patients with pulmonary disease. However, the underlying disease process, especially in its relatively early phases, may result primarily in changes in pulmonary function without substantial changes in pulmonary morphology or structure. In such cases, the traditional pulmonary imaging approaches may not provide enough insight into the underlying pathological process. Therefore, we have tried to assess pulmonary diseases in terms of gas exchange using nuclear medicine, CT and MR imaging combined with various technical approaches.

The principal function of the lung is gas exchange. For the lung to perform gas exchange effectively, the flow of air to the alveoli (ventilation) and the flow of blood to the capillary (perfusion) have to occur in a coordinated manner. Therefore, pulmonary function can be evaluated as a combination of ventilation and perfusion imaging with reference to spatial and temporal information derived from time-resolved anatomical imaging. A great advantage of pulmonary functional imaging is its ability to examine these processes on a regional basis, which is different from the traditional measures of pulmonary functions, such as spirometry, which evaluates the whole lung as a single unit. As many diseases begin regionally, their functional effects are often masked by normal lung elsewhere until the disease has progressed to involve substantial portions of the lung parenchyma. Pulmonary functional imaging may thus significantly improve our ability to evaluate these diseases before they progress to global changes in function and morphology. In addition to the evaluation of pulmonary disease, functional imaging may have a significant impact upon our understanding of pulmonary physiology. As already stated, many of the traditional methods for the evaluation of physiology lack spatial or temporal resolution. Pulmonary functional imaging, on the other hand, offers new, non-invasive, methods to evaluate regional physiology with high spatial and temporal resolution. In the last a few decades, new techniques for not only CT and MR imaging have been introduced, while new tracers for PET are expected to become available within the next decade, and be tested for clinical relevance for not only radiology, but also pulmonary medicine. Furthermore, MR imaging has long been considered the leading radiological technique in this setting and can be applied as a substitute for nuclear medicine studies, although for routine clinical practice, morphological assessment of lung parenchyma with MR imaging is still considered inferior to CT.

In this review, we describe and discuss state-of-the art MR methods as well as the future direction of (1) ventilation imaging, (2) pulmonary perfusion and hemodynamic imaging, and (3) biomechanical evaluation for pulmonary functional imaging.

Ventilation imaging

Hyperpolarized noble gas MR imaging

Hyperpolarized noble gas MR imaging and oxygen-enhanced (O₂-enhanced) MR imaging have been studied as potential MR-based ventilation imaging techniques since the 1990s. ^1–4 Hyperpolarized noble gas MR imaging can be performed with helium-3 (³He) and xenon-129 (¹²⁹Xe). It has been extensively tested for assessment of disease severity and therapeutic effect evaluation of various pulmonary diseases. ^5–29 With considering the hyperpolarized noble gas MR techniques, hyperpolarized noble gas MR imaging by using ³He and ¹²⁹Xe can provide ventilation images as spin-density images and gas molecule diffusion within air space, which is different from oxygen diffusion from alveoli to capillary bed, as diffusion-weighted imaging (DWI). Spin density images during breath-hold represent a snapshot of gas distribution. On the other hands, DWI exploits the high free diffusion of ³He and ¹²⁹Xe gases. In the context of restriction by the lung microstructure, this allows indirect measurement of the average dimensions of the lung airspaces. When ³He or ¹²⁹Xe gas is restricted by tissue boundaries, the diffusivity is referred to as the apparent diffusion coefficient (ADC). A semi-quantitative ADC, which is considered as a surrogate for airspace size, map is easily derived. It shows differences in regional ventilation for cystic fibrosis, for asthma and chronic obstructive pulmonary disease (COPD), ^5–12 for dynamic gas movement by means of dynamic gradient-echo sequencing ^5–9 and for regional gas diffusion to determine alveolar size, ^{5,6,9,12–15} all with higher spatial resolution or temporal resolution than can be attained with nuclear medicine studies (Figure 1). Moreover, hyperpolarized noble gas MR imaging with ¹²⁹Xe can also provide additional information such as xenon polarization transfer contrast (XTC) and a depolarization map, which are believed to better reflect the regional lung parenchyma density as can be obtained with hyperpolarized noble gas MR imaging with ³He. ^5,6,16 The soluble “dissolved phase” fraction of ¹²⁹Xe in blood and tissues is ~2% of the total signal after accounting for the lower density of lung tissue and blood in the lungs. The dissolved phase of ¹²⁹Xe has a different chemical shift frequency than the gas phase. This fact can be exploited to quantify gas exchange. Polarized gas nuclei diffuse rapidly between gas, blood cells, and plasma/tissue compartments, with different chemical shifts in each compartment: 0 ppm, 222 ppm, and 198 ppm, respectively. ^5,16 Compartmental modeling of these components has advanced relatively rapidly. ¹⁸ Simple and robust single voxel MR spectroscopy readily resolves tissue and blood enables the calculation of blood/tissue ratio as a possible biomarker of diffusion block. ^19,20 More quantitative measures such as saturation transfer time allow the kinetics of ¹²⁹Xe recovery to be modeled. These more advanced method scans provide direct or indirect estimates of average septal wall thickness and alveolar surface area to volume ratio. ^16,18–21 Therefore, XTC is considered as one of the advantages of hyperpolarized noble gas MR imaging with ¹²⁹Xe, when compared with that by ³He. Recently, pO₂ mapping by hyperpolarized noble gas MR imaging is introduced as a new promising tool for pulmonary functional MR imaging. ^22–24 The paramagnetic effects of oxygen reduce the T1 and T2 of hyperpolarized gases as well as proton signals and can therefore be exploited to provide a quantitative estimate of regional pO₂. ^22–24 Typically, the same slice is imaged repeatedly at different delay times following gas inhalation to separate RF saturation of signal from signal loss due to local O₂ concentration. ²⁵ The measurement is performed within a single breath-hold. ^22,26 Modified centric and reverse-centric view orders mitigate effects due to locally variable flip angle (B1-field variation). ²⁶ Gas flow effects within the lungs during a breath-hold using ³He have recently been described, underscoring that pO₂ is a mixture of regional pO₂ and local collateral ventilation effects. ^27–29 Regional pO₂ measures can potentially be used to calculate the ratio of ventilation to perfusion (V/Q). ²⁵ In fact, pO₂ measurements have recently been used as a marker of disease severity in a study of V/Q heterogeneity in smokers. ^27–29 For these reasons, hyperpolarized noble gas MR imaging has been viewed as useful for not only ventilation imaging, but also the quantitative assessment of changes in lung structure in patients with some pulmonary diseases. However, these techniques have not yet been clinically established. Drawbacks of these techniques are their dependence on polarizer and multinuclear technology which is not widely available. ⁶ The global supply of ³He is also very limited, thus leading to high cost. ¹⁷ These drawbacks account for the shift to the more widely available ¹²⁹Xe nucleus. The technical challenges of this shift have been made easier by advances in ¹²⁹Xe SEOP polarization. ⁶ Hyperpolarized ¹²⁹Xe MR imaging typically uses enriched ¹²⁹Xe isotope with a natural abundance of 26% to ≤85% to compensate for its lower gyromagnetic ratio and achievable polarization compared to those of ³He gas. ⁶ However, ¹²⁹Xe also dissolves in water/blood and therefore lung signal also derives from ¹²⁹Xe uptake. ⁶ Moreover, the anesthetic effects of ¹²⁹Xe is main causes of the limitation for administrating ¹²⁹Xe gas. All the major types of hyperpolarized ³He MR imaging have now been replicated robustly with enriched hyperpolarized ¹²⁹Xe MR imaging for future incorporation in the clinical setting, although the above-mentioned limitations of ¹²⁹Xe have to be considered.

Figure 1.

Ventilation and PRM of (1) a 55-year-old male without COPD (FEV₁, 83% of predicted value; FEV₁/FVC, 77%; residual volume to total lung capacity ratio [RV/TLC], 45%) (2) a 69-year-old male with GOLD I disease (FEV₁, 89% of predicted value; FEV₁/FVC, 69%; RV/TLC, 39%; DL_CO, 67% of predicted value) (3) an 84-year-old male with GOLD II disease (FEV₁, 52% of predicted value; FEV₁/FVC, 44%; RV/TLC, 62%; DL_CO, 47% of predicted value), and (4) a 67-year-old female with GOLD III disease (FEV₁, 33% of predicted value; FEV₁/FVC, 39%; RV/TLC, 72%; DL_CO, 28% of predicted value). First row: ³He MR images co-registered with ¹H MR images (grayscale) show static ventilation (blue areas). Second row: ³He MR imaging ADC maps show that ex-smokers with more advanced COPD (GOLD II/III disease) have elevated ADC values. Third row: CT attenuation masks show areas of less than −950 HU (yellow areas). Fourth row: PRMs show areas of healthy tissue (green), gas trapping (yellow), and emphysema (red). Permission was obtained from reference number 12 (RSNA provide the permission). ADC, apparent diffusion coefficient; COPD, chronic obstructive pulmonary disease; HU, Hounsifield unit; PRM, parametric response map

Oxygen-enhanced MR imaging

O₂-enhanced MR imaging has been recommended since 1996 as having potential for MR-based ventilation imaging. ^2,6,30–46 The basic feature of the molecular oxygen used for O₂-enhanced MR imaging is that it is weakly paramagnetic with a magnetic moment of 2.8 Bohr magnetons. O₂ modulates the MR imaging signal of blood and fluid through two different mechanisms: the paramagnetic property of deoxyhemoglobin and the paramagnetic property of molecular oxygen itself. ^47,48 Because deoxyhemoglobin is compartmentalized in red blood cells (RBC), tissue water protons do not have access to coordination sites, which is required for the spin-lattice interactions causing T1 shortening. ^47,48 Therefore, deoxyhemoglobin with RBC has a T2*-shortening effect with little T1-shortening. ^47,48 Because molecular oxygen is paramagnetic, dissolved molecular oxygen will shorten the T1 relaxation time of pulmonary venous blood. Therefore, the areas of the lung affected by pulmonary emphysema are pathologically characterized by a decrease in pulmonary capillary beds, abnormal permanent enlargement of the air spaces distal to the terminal bronchioles by destruction of their walls, while there is no obvious fibrosis. O₂-enhanced MR imaging can demonstrate not only regional oxygen enhancement based on oxygen diffusion from alveoli to capillary bed, but also oxygen uptake based on ventilation as well as respiration. For these reasons, it has been suggested as useful for evaluation of various pulmonary diseases such as asthma, COPD, cystic fibrosis, interstitial lung disease, lung cancer ^32–46 (Figure 2). O₂-enhanced MR imaging is a proton-based technique and thus considered to be more clinically useful than hyperpolarized noble gas MR imaging and one of the key techniques for MR-based ventilation imaging. No other imaging method can demonstrate oxygen uptake based on respiration itself, although O₂-enhanced MR imaging is not able to directly visualize O₂ itself. However, its drawbacks are changes in low signal intensity due to oxygen inhalation, and multiple acquisitions or increased number of excitations are warranted for compensate low signal intensity change. A certain amont of time is needed for allowing the oxygen to wash in and out, both together resulting in very long acquisition time. It is time consuming to cover the entire lung because the 2D sequence rather than the 3D sequence has been applied in the past studies. Many studies to assess this technique used semi-quantitative assessment, and only a few study studies used quantitative evaluation. ^2,32–46 Therefore, further developments of O₂-enhanced MR imaging for 3D acquisition are needed to reduce the time needed to cover the entire lung and improve quantitative analysis for oxygen uptake evaluation. Finally, these new techniques should be further investigated to enhance the clinical relevance of O₂-enhanced MR imaging.

Figure 2.

Images show a 55-year-old female with moderate persistent asthma. (a) Thin-section coronal MPR image (left) shows no emphysema, and MLD was assessed as −890 HU. On the quantitative coronal CT image (right), areas with attenuation values less than −950 HU appear in blue, and the low-attenuation area was assessed as 6.7%. (b) On thin-section CT at the vertical plane to the third level of the bronchus, outer and inner areas of the bronchus are circled and the green area represents the inner lumen area of the bronchus. Quantitatively assessed wall area was 58.6%. (c) Source image of dynamic oxygen-enhanced MR imaging (left), relative enhancement map (center), and wash-in time map (right). Regional distributions of relative enhancement and wash-in time for each pixel were expressed as maps color coded from blue to red. The pretherapeutic mean RER was 0.18, and mean wash-in time was 26.0 s. (d) Source image of dynamic oxygen-enhanced MR imaging (left), relative-enhancement map (center), and wash-in time map (right). Regional distributions of relative enhancement and wash-in time for each pixel were expressed as maps color coded from blue to red. Post-therapeutic mean RER was 0.23, and mean wash-in time was 17.7 s. Permission was obtained from reference number 42 (RSNA provide the permission). HU, Hounsfield unit; MPR, multiplanar reconstruction; RER, relativeenhancement ratio.

Fluorinated gas MR imaging

As part of the continuing developments in hyperpolarized noble gas MR imaging and O₂-enhanced MR imaging during the last few decades, some research groups have been testing fluorine-19 (¹⁹F) MR imaging as another promising approach for MR-based ventilation imaging. Inert ¹⁹F gases, on the other hand, are not hyperpolarized but are still able to be used for ventilation imaging. Fluorinated gases are an attractive choice because the ¹⁹F isotope is 100% naturally abundant, with a high gyromagnetic ratio on the order of that of ¹H. ^6,49 The T₁ relaxation is dominated by spin-rotation interactions, T₁ ~ T₂. ^1,50 Both MR decay parameters depend on the concentration of the gas itself and are generally on the order of a few milliseconds for the ¹⁹F nucleus. Therefore, signal intensity on ¹⁹F MR imaging is directly obtained from ¹⁹F itself, and not from hyperpolarization of the gas. Therefore, ¹⁹F MR imaging uses inhaled inert fluorinated gases for MR-based ventilation imaging and as such may constitute a more economical alternative to hyperpolarized noble gas MR imaging. ^51–54 Inert fluorinated gases are nontoxic, abundant, relatively inexpensive, and the technique can be performed on any MRI scanner with broadband multinuclear imaging capabilities. ^51–54 Feasibility studies for human lungs have been conducted and reportedly generated breath-hold images with gas mixtures containing 79% sulfur hexafluoride (SF₆) or perfluoropropane (C₃F₈ or PFP) and 21% oxygen (O₂). ^52,55 Since they are inert fluorinated gases, SF₆ and PFP are non-toxic, abundant, and relatively inexpensive. One report has estimated costs for ³He, ¹²⁹Xe and ¹⁹F gases. ⁶ The per-liter cost of ³He and enriched ¹²⁹Xe is much higher than that of ¹⁹F gases, but the cost of naturally abundant xenon is similar to that of SF₆ and PFP. Since a typical imaging session using ¹⁹F MR imaging might use several liters of SF₆ or PFP, the total cost of a single study may approach the cost of an enriched hyperpolarized noble gas MR imaging study using ¹²⁹Xe; however, the advantage of ¹⁹F gas MR imaging is that the gases are imaged using thermal equilibrium polarization so that a polarizer system is not required. ⁵³ In addition, ¹⁹F gas MR imaging can visualize regional differences in ventilation as well as air trapping and thus may provide functional information similar to that obtained from images acquired with hyperpolarized noble gas MR imaging(Figure 3). ^51–54 In addition, ¹⁹F MR imaging is also suggested as having the capabilities for demonstration of gas molecule diffusion and ventilation/ perfusion (V/Q) mapping in animal studies. ^51,56–59 On the whole, ¹⁹F gas MR imaging is expected to complement existing proton-based structural imaging techniques, and the combination of structural and functional lung MR imaging to provide useful outcome measurements for future management of pulmonary diseases based on not only academic, but also clinical findings. However, this technique has to be overcome several aspects such as equipment, scanner time and personnel, etc. Therefore, extensive further investigations are expected to be performed during the next decade or two for clinical application as one of the pulmonary functional imaging techniques used in routine clinical practice.

Figure 3.

Images show examples of morphologic imaging, initial and late ¹⁹F gas wash-in MR imaging with ventilation defect (red masks), and maps of wash-out times and FV assessed with dynamic ¹⁹F gas wash-out MR imaging for the different severities of COPD indicated by percentage predicted FEV₁ (FEV₁% predicted; corresponding to GOLD stages I–IV). VDP at initial gas wash-in MR imaging and median wash-out time increase and correspondingly median FV decreases with an increase in severity of COPD. Additionally, gas wash-out parameter maps show an increase in heterogeneity of regional lung ventilation with progression of disease. Permission was obtained from reference number 54 (RSNA provides the permission). COPD, chronicobstructive pulmonary disease; FV, fractional ventilation; VDP, volume defectpercentage.

Although we discuss MR-based ventilation imaging such as hyperpolarized noble gas MR imaging, O₂-enhanced MR imaging and ¹⁹F MR imaging for providing different ventilation-based information, dynamic contrast-enhanced MR imaging based on perfusion MR techniques is currently the most frequently used technique to assess pulmonary ventilation by exploiting the effect of hypoxic vasoconstriction in the lung. ⁶⁰ With considering this fact, dynamic contrast-enhanced MR imaging with perfusion MR techniques is also applied for not only perfusion, but also ventilation evaluations in patients with various pulmonary diseases in routine clinical practice.

Pulmonary perfusion and hemodynamic imaging

Non-contrast-enhanced and contrast-enhanced MR angiography

As early as the 1990s, it was suggested that non-contrast-enhanced and contrast-enhanced MR angiography are useful for visualization of pulmonary vasculature and blood flow by means of various methods. ^61–69 Currently, there are two major established methods for assessment of pulmonary vasculature and blood flow. One is non-contrast-enhanced (CE-) MR examination, which uses a proton within blood as an endogenous tracer, and another is CE-MR examination which uses a gadolinium contrast medium as an intravenously injected tracer. As other methods for MR angiography, it has been suggested that the balanced steady-state free precession (bSSFP) sequence, the 3D electrocardiogram-gated (ECG-gated) fresh blood imaging (3D FBI) using 3D fast spin-echo (SE)–based sequence and the arterial spin labelling technique with SE and gradient-recalled echo (GRE) type sequences are useful for assessment of pulmonary vasculatures ^65,66 (Figure 4). Results obtained thus far indicate that these techniques should be used in routine clinical practice for pulmonary vascular disease patients with low renal function or contraindication of gadolinium contrast media.

Figure 4.

50-year-old male with adenocarcinoma. (a) Thin-section CT shows lung cancer (arrow) in right posterior segment. (b). Thin-section contrast-enhanced multiplanar reformatted images obtained at mediastinal (left) and lung (right) window settings clearly show that posterior segmental artery (arrows) branches directly from pulmonary arterial trunk (visualization score, 5; variation probability score, 5). (c) Pulmonary arterial phase (left; time, 4.0 s), parenchymal phase (middle; time, 8.0 s), and venous phase (right; time, 12.0 s). 4D contrast-enhanced MR angiography images clearly show that posterior segmental artery (arrows) branches directly from pulmonary arterial trunk (visualization score, 5; variation probability score, 5). (d) Anterior (left) and posterior (right) cine MRI shows that posterior segmental artery (arrow) branches directly from pulmonary arterial trunk (visualization score, 4; variation probability score, 5). (e) Fresh blood imaging clearly shows that posterior segmental artery (arrow) branches directly from pulmonary arterial trunk (visualization score, 5; variation probability score, 5). (f) Time-spatial labeling inversion pulse image clearly shows that posterior segmental artery (arrow) branches directly from pulmonary arterial trunk (visualization score, 5; variation probability score, 5). Permission was obtained from reference number 65 (ARRS provides the permission).

On the other hand, CE-MR angiography examination has been advocated as useful and clinically applicable for diagnosis or assessment of not only pulmonary vascular diseases, but also other diseases including lung cancer during the past two decades. ^63–69 This technique is usually known as “time-resolved CE-MR angiography” or “4D CE-MR angiography” after clinical installation of parallel imaging techniques, and is more widely used in routine clinical practice than non-CE-MR angiography. ^63–69 Moreover, time-resolved CE-MR angiography is also used as a substitute or in a complementary role for CE-CT angiography and perfusion scanning, SPECT or SPECT/CT, in routine clinical practice because this technique can clearly show not only pulmonary vasculature, but also pulmonary parenchymal perfusion in a single examination. ^63–69 Therefore, this examination is considered to be clinically applicable and to be most effective for patients with cardiopulmonary diseases in routine clinical practice.

Non-contrast-enhanced and contrast-enhanced perfusion MR imaging

As for non-CE-perfusion MR imaging, signal targeting with an alternating radiofrequency (STAR) technique with a half-Fourier single shot turbo spin-echo (HASTE) sequence, ^70,71 arterial spin tagging imaging, ⁷² arterial spin labelling imaging such as flow sensitive alternating inversion recovery (FAIR) and FAIR with an extra radiofrequency pulse (FAIRER) imaging, ⁷³ ECG-gated fast-SE perfusion MR imaging generated as subtraction imaging between the systolic and diastolic phases ^74,75 were tested using not only healthy volunteers, but also patients with different pulmonary diseases. ^66,70–75 In contrast to ECG-gated fast-SE perfusion MR imaging, all other techniques are based on arterial spin labeling (ASL) MR perfusion is an MR perfusion technique which does not require intravenous administration of contrast. Instead, it exploits the ability of MR imaging to magnetically label arterial blood below the imaging slab. ASL images are acquired with MR pulse sequences that have two distinct and independent components, a preparation component and an acquisition component. The preparation component labels the inflowing blood in different magnetic states for the control and label images. Once the blood has been labeled, the acquisition component acquires the data. The acquisition component typically uses a fast acquisition method such as HASTE technique. The distinct and independent nature of the preparation and acquisition components has allowed researchers to choose the combination that is best suited for their specific application. However, all these techniques used 2D acquisition and were time consuming. These techniques have therefore not been used in the clinical setting since they were first introduced.

Since 2009, non-CE perfusion and ventilation assessment of the human lung by means of the Fourier decomposition (FD) technique for proton MRI has been tested using both healthy volunteers and patients with pulmonary diseases. ^76–81 FD MR imaging has been proposed as one of the non-CE MR techniques to obtain regional lung perfusion and ventilation-related information during a single acquisition series. ^76–81 The method utilizes a rapid free-breathing acquisition of time-resolved MR data combined with the compensation of respiratory motion by non-rigid image registration. Then, the Fourier analysis of the acquired data allows for a spectral separation of respiratory and cardiac signal variations, which are caused by changes of the regional proton density due to the lung parenchyma contraction and by flow-dependent signal dephasing. Thus, FD MR imaging provides an indirect approach for the assessment of ventilation and perfusion. ^76–81 Although the FD technique uses 2D acquisition, it can provide perfusion-weighted and ventilation-weighted images at the same time and performs like a proton-based MR technique.

In contrast to the above-mentioned 2D methods, the utility of 3D ECG- and respiratory-gated non-CE-perfusion MR imaging for prediction of post-operative lung function in non-small cell lung cancer patients has been demonstrated and it has shown similar potential for assessment of regional perfusion as dynamic CE-perfusion MR imaging. ⁸² Although this technique is considered to be only one method for 3D non-CE-perfusion MR imaging, several investigators are currently trying to have it accepted in the clinical setting and argue that it should be used for various pulmonary diseases, since the results of their studies have demonstrated its clinical relevance.

From the semi-quantitative and qualitative assessment of pulmonary circulation assessment, dynamic first-pass CE-perfusion MR imaging was extensively tested for patients with various pulmonary diseases by many investigators during the last two decades. ^{61–64,66–69,83–91} With the aid of indicator dilution therapy and deconvolution analysis, quantitatively assessed pulmonary blood flow, pulmonary blood volume and mean transit time can be measured on dynamic first-pass CE-perfusion MR imaging. ^{61–64,66–69,83–91} By using the above-mentioned methods, quantitatively assessed perfusion parameters can be employed as image-based biomarkers, not only for diagnosis, but also for disease severity assessment and patient management in various pulmonary diseases (Figure 5).

Figure 5.

A 30-year-old male volunteer a: Image maps (L to R: ventral to dorsal) of PBF shows regional changes of PBF in the gravitational and isogravitational directions. b: Image maps (L to R: ventral to dorsal) of PBV show regional changes of PBV in the gravitational and isogravitational directions. c: Image maps (L to R: ventral to dorsal) of MTT show regional changes of MTT in the gravitational and isogravitational directions. Permission was obtained from reference number 86 (John Wiley & Son provides the permission). MTT, mean transit time; PBF,pulmonary blood flow; PBV, pulmonary blood volume.

All MR scanners with field strengths equal to or higher than 1.5 Tesla from every vendor can currently perform dynamic first-pass CE-perfusion MR imaging with and without parallel imaging technique. On the other hand, there are no commercially available software for quantitative assessment of pulmonary perfusion parameter, and all previously reported studies used the investigators’ proprietary software. Therefore, vendors should be encouraged to provide appropriate sequences, bolus injection protocols and commercially available software for broadening the clinical application of this technique as soon as they can.

Velocity-encoding MR imaging for pulmonary hemodynamic evaluation

As a substitute for cardiac echography, velocity-encoding (or phase-contrast) MR imaging with velocity sensitizing gradients can be used to quantitatively map the blood flow in the major vessels in the lungs in two or three dimensions. It can thus provide insights into pulmonary vascular resistance and non-steady and turbulent flow in the pulmonary arteries, mainly in pulmonary arterial hypertension due to various cardiopulmonary diseases. ^92–96 Recent developments involving 3D view-shared phase contrast methods allow for 3D time-resolved imaging of blood flow in the pulmonary arteries and the acquisition of 4D flow in the pulmonary arteries ^94–97 (Figure 6). With applying the 4D velocity-encoded MR technique, 3D pulmonary vascular flow and advanced hemodynamic parameters are able to be assessed, and the degree of vortical flow in the pulmonary trunk may be suggested as one of the key aspects that reflects pulmonary hypertension. ^94–97 Therefore, 4D velocity-encoded MR imaging can highlight the value of MR imaging instead of cardiac echogram for the clinical diagnosis and management of patients with pulmonary hypertension.

Figure 6.

Process of determination of the duration of vortical blood flow in the main pulmonary artery A vortex of the blood flow in the main pulmonary artery is defined as being present in a cardiac phase if velocity vectors in the pulmonary artery form closed concentric ring-shaped curves that are parallel to the right ventricular outflow tract orientation, which is shown in the middle two images. The duration of the vortical blood flow in the main pulmonary artery is specified as the number of cardiac phases with vortical blood flow divided by all 20 reconstructed cardiac phases (bottom). Permission was obtained from reference number 94 (RSNA provides the permission).

Biomechanical evaluation for pulmonary functional imaging

Recent advances in MR technology have made it possible to assess diaphragmatic and chest wall motion, static and dynamic lung volumes, as well as regional lung function. Even though existing studies have shown major heterogeneity in design and applied methods, it has become evident that MR imaging can visualize pulmonary function as well as diaphragmatic and thoracic wall movement, thus providing new insights into lung physiology.

Several studies were designed to answer basic physiological questions about diaphragmatic configuration and function visualized as static thoracic MR images in sagittal and coronal orientations. ^98–101 Lung volumes were calculated from the analyzed contours of thorax and diaphragm and judged to be in good agreement with spirometric measurements. However, all these studies have a limitation, namely, that image acquisition was performed with the subject in a supine position, leaving the question open as to whether the findings would be transferable to a normal upright position. One study addressed this question by using an open MR system to dynamically image subjects in the sitting as well as supine position during slow maximal respiration and showed that diaphragmatic excursion is significantly greater when in the sitting than in the supine position, especially in the posterior region. ¹⁰² For static lung volume evaluations, a few studies were conducted to demonstrate methods for estimation of lung volume in normally breathing non-sedated infants with wheezing and bronchitis ¹⁰³ and emphysema patients scheduled for lung volume reduction surgery as well as healthy volunteers. ^104,105

In contrast to static lung volume assessment, dynamic lung volume evaluation could be attained by means of parallel imaging and the use of fast sequences with high temporal resolution and real-time visualization of lung and chest wall motion. ^106–109 Compared to healthy controls, patients with emphysema showed a flattened diaphragm and a reduced, irregular respiratory motion of both chest wall and diaphragm. Repetition of dynamic MR imaging after lung volume reduction surgery showed significant improvement in these parameters. Other studies used dynamic MRI to visualize chest wall and diaphragmatic movement for instantaneous spirometric measurements. ^107,108 In addition, the influence of tumor location on the local chest wall and diaphragmatic movement was also assessed as evident when this method was used for lung cancer and malignant mesothelioma. ^110–114 Further, the introduction of dynamic 3D MR imaging, using an isotropic time-resolved 3D-gradient echo pulse sequence for volumetric assessment of the breathing cycle, ¹¹⁵ improved the correlation between spirometric data and MR-volumetric data. However, the lower temporal resolution of the sequence allowed for a dynamic examination during slow maximum inspiration and expiration. Therefore, 3D sequences were used with parallel imaging techniques to improve correlations between MR-based volumetric evaluations and pulmonary function test results. ¹¹⁵

Pulmonary parenchymal strain and mechanical properties of the lung have been suggested for regional functional assessment, and much effort has been made to study these aspects with the aid of imaging. The thoracic wall and the diaphragm provided sufficient signal to be relatively easy to track. However, motion analysis of these structures is only a substitute for pulmonary motion and function, and approaches to assess pulmonary tissue motion directly are challenging due to the limited signal provided. Nevertheless, it has been possible to evaluate two promising techniques. The first uses grid tagging, a technique widely employed in MR motion analysis of not only heart, but also lung. In the lung, the fast signal decay of the grid due to the large number of susceptibility artefacts at the tissue/air interfaces presents the main challenge. Despite these difficulties, the feasibility of this technique has been shown by several studies. ^116–118

The second approach, registration of serial images, has to cope with the limited signal of pulmonary tissue and thus low signal-to-noise ratio, especially for dynamic imaging. So far, this technique has been evaluated in not only healthy volunteers and mice, but also interstitial lung disease. ^{115,119–123} However, its ability to detect motion changes over time appears promising for serial imaging of patients for the assessment of disease progression. The major limitations of registration methods used for MR imaging remain the poor signal and contrast from solid tissue and blood vessels. If this limitation could be overcome, registration would even be possible on 3D images, which would eliminate the problem of through-plane motion. Another advantage of the registration approach is that it is not dependent on a specific imaging sequence and thus should profit directly from new developments in MR sequences.

Currently, MR-based biomechanical assessment has been mainly attempted for radiation oncology rather than pulmonary functional imaging. The lack of ionizing radiation makes MR-based biomechanical assessment suitable for experimental work with healthy subjects. For this purpose, MR-based biomechanical assessment is just being put into practice and the optimization of the dedicated sequence protocols are strongly encouraged. Moreover, further investigations are warranted in the near future for demonstrating the clinical relevance of this technique for not only motion tracking in radiation oncology, but also pulmonary functional imaging.

Conclusion

This report reviewed state-of-the-art pulmonary ventilation imaging, perfusion imaging and biomechanical evaluation and discussed their future challenges. Imaging techniques are gradually but steadily shifting from nuclear medicine to MR imaging, and from those with qualitative assessment to those with quantitative assessment. In addition, they are still being developed and evaluated for clinical relevance in not only radiology, but also in the multidisciplinary fields of pulmonology, pulmonary physiology, pathophysiology, thoracic surgery, radiation oncology. In addition, the combining of imaging method with various image analyses by means of bioengineering is also important for future clinical settings. Finally, new MR imaging techniques that are under development and/or are being tested can be expected to materialize in the foreseeable future.