Nội dung

(2021) 21:85 Laghi Jr. et al. BMC Pulm Med https://doi.org/10.1186/s12890-021-01441-6 Open Access REVIEW Ultrasound and non‑ultrasound imaging techniques in the assessment of diaphragmatic dysfunction Franco A. Laghi Jr.1, Marina Saad2 and Hameeda Shaikh3,4* Abstract Diaphragm muscle dysfunction is increasingly recognized as an important element of several diseases including neuromuscular disease, chronic obstructive pulmonary disease and diaphragm dysfunction in critically ill patients. Functional evaluation of the diaphragm is challenging. Use of volitional maneuvers to test the diaphragm can be limited by patient effort. Non-volitional tests such as those using neuromuscular stimulation are technically complex, since the muscle itself is relatively inaccessible. As such, there is a growing interest in using imaging techniques to characterize diaphragm muscle dysfunction. Selecting the appropriate imaging technique for a given clinical scenario is a critical step in the evaluation of patients suspected of having diaphragm dysfunction. In this review, we aim to present a detailed analysis of evidence for the use of ultrasound and non-ultrasound imaging techniques in the assessment of diaphragm dysfunction. We highlight the utility of the qualitative information gathered by ultrasound imaging as a means to assess integrity, excursion, thickness, and thickening of the diaphragm. In contrast, quantitative ultrasound analysis of the diaphragm is marred by inherent limitations of this technique, and we provide a detailed examination of these limitations. We evaluate non-ultrasound imaging modalities that apply static techniques (chest radiograph, computerized tomography and magnetic resonance imaging), used to assess muscle position, shape and dimension. We also evaluate non-ultrasound imaging modalities that apply dynamic imaging (fluoroscopy and dynamic magnetic resonance imaging) to assess diaphragm motion. Finally, we critically review the application of each of these techniques in the clinical setting when diaphragm dysfunction is suspected. Keywords: Diaphragm dysfunction, Ultrasound imaging, Static imaging techniques, Dynamic imaging techniques, Phrenic nerve, Mechanical ventilation, Neuromuscular disorders Background The diaphragm is the main muscle of respiration during resting breathing [1]. When respiratory demands are increased or diaphragm function is impaired, rib cage muscles and expiratory muscles are progressively recruited [2]. In some patients with diaphragm dysfunction, this compensation is associated with minimal *Correspondence: Hameeda.Shaikh@va.gov 3 Division of Pulmonary and Critical Care Medicine, Hines Veterans Affairs Hospital (111N), 5th Avenue and Roosevelt Road, Hines, IL 60141, USA Full list of author information is available at the end of the article or no respiratory symptoms [1]. In other patients, this compensation is associated with significant respiratory symptoms. Diaphragm dysfunction can cause alveolar hypoventilation and, in the most severe cases, respiratory failure requiring mechanical ventilation [1, 2]. The ultimate causes of diaphragmatic dysfunction can be broadly grouped into three major categories: disorders of central nervous system or peripheral neurons, disorders of the neuromuscular junction and disorders of the contractile machinery of the diaphragm itself [3] (Table 1). Early diagnosis of diaphragmatic dysfunction is essential, © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​iveco​mmons​.org/licen​ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat​iveco​ mmons​.org/publi​cdoma​in/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Page 2 of 29 Table 1 Pathophysiological causes of diaphragmatic dysfunction Group Examples Neuronal disorders a) Disorders of CNS Poliomyelitis, ALS, multiple sclerosis, spinal cord injury b) Disorders of peripheral neurons Guillain–Barre syndrome, chronic inflammatory demyelinating polyneuropathy, critical illness polyneuropathy, compression of the phrenic nerve by neighboring structures Disorders of the neuromuscular junction Myasthenia gravis, Lambert-Eaton syndrome, botulism, organophosphates Disorders of the muscle Muscular dystrophies, critical-illness myopathy, acid maltase deficiency (Pompe disease), dermatomyositis CNS central nervous system, ALS amyotrophic lateral sclerosis because it may be responsive to therapeutic intervention [4–7]. A number of static and dynamic imaging techniques are used in the evaluation of patients suspected of diaphragm dysfunction [8]. Static imaging techniques are used to assess the position, shape and dimensions of the diaphragm and include chest radiography [9], brightness mode (B-mode) ultrasound [10, 11], computed tomography (CT) [12], and static magnetic resonance imaging (MRI) [13]. Dynamic imaging techniques are used to assess diaphragm motion in one or more directions. This group of imaging techniques include fluoroscopy [14], motion mode (M-mode) ultrasonography [10, 11, 15], and dynamic MRI [16]. The purpose of this review is to present the accumulated knowledge on imaging techniques used in the evaluation of diaphragm dysfunction. A MEDLINE search of articles published between 2010 and 2020 was undertaken. Searches of the bibliographies of articles resulted in several additional articles and book chapters. Information was selected on the basis of scientific quality and potential relevance to patients suffering from a pulmonary disease or a disorder requiring admission to an intensive care unit; in all, a total of 128 sources were included in this review. In this review we will first discuss ultrasound imaging techniques and then non-ultrasound imaging techniques. We will then discuss the clinical applications of these techniques. Finally, we will provide a critical appraisal of the limitations of ultrasound and non-ultrasound imaging in the evaluation of diaphragmatic dysfunction and considerations about future directions. investigators reported a close correlation between diaphragm thickness measured in cadavers using ultrasound imaging and thickness measured with a ruler (Fig. 1). Since the seminal work of Wait et al. [18], investigators have published a growing number of studies on the use of ultrasonography to monitor the thickness of the diaphragm in the zone of apposition, the motion of the dome and to estimate diaphragm strength and recruitment during voluntary contractions [19, 20]. Ultrasound measurement of diaphragm thickness (zone of apposition) Operators use linear ultrasound probes to measure diaphragm thickness [15] (Fig. 2). These probes use high frequency ultrasound waves (7–18 Hz) to create high resolution images of structures near the body surface [15]. To measure diaphragm thickness, operators place the ultrasound probe longitudinally parallel to the long axis of the body, usually between the eighth to tenth Ultrasound imaging of the diaphragm Diaphragm ultrasonography was first described in the late 1960s as a means to determine position and size of supra- and subphrenic mass lesions, and to assess the motion and contour of the diaphragm [17]. Two decades later, Wait et al. [18] developed a technique to measure diaphragm thickness based on ultrasonography. The Fig. 1 Relationship of diaphragm thickness measured in situ in 10 human cadavers by ultrasound and in vitro by a ruler. Ultrasound measurements of diaphragm thickness in situ are as accurate as measurements in vitro with a ruler. Reproduced with permission from The American Physiological Society: Wait et al. J Appl Physiol 1989;67(4):1560–1568 Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Page 3 of 29 Fig. 2 Ultrasound image of the zone of apposition of the diaphragm. In brightness-mode (B-mode; left panel), the diaphragm appears as a three-layer structure. In motion-mode (M-mode; right panel), the diaphragm is thinnest at end-exhalation and thickest at end-inhalation; stronger diaphragmatic inspiratory efforts are associated with greater tidal thickening of the diaphragm. Reproduced with permission from The American Association for Respiratory Care: Shaikh et al. Respir Care 2019;64:1600–2. The entirety of both images was obtained by Dr. Shaikh intercostal space [21, 22], at the anterior axillary line [15] or midway between the anterior- and mid-axillary lines [23]. The costo-phrenic sinus [24] is identified as the transition between lung and liver (right) or between lung and spleen (left). The zone of apposition, where the diaphragm is opposed to the rib cage, is located caudal to the costo-phrenic sinus [24]. To identify the diaphragm, subjects are asked to inhale while operators select B-mode imaging. As the lung comes between transducer and diaphragm, it creates an hyperechoic bright artifact (“lung curtain sign”) [25] that obliterates the muscle’s image [18] (Additional file 1). The diaphragm is identified as a threelayer structure (two echogenic layers of peritoneum and pleura sandwiching a more hypoechoic layer of the muscle itself ) underneath the intercostal muscles [26] that reappear as lung artifact recedes [15, 18] (Fig. 2). Occasionally, an additional bright layer due to connective tissue and vessels can be seen within the muscle layer itself [27]. To reduce lung artifact operators can move the transducer towards the anterior axillary line [25] or to the next (caudal) intercostal space. It is easier to visualize the right than the left hemidiaphragm [24, 28]. Intra- and interobserver agreement of measurements of diaphragm thickness obtained at a single sitting in healthy adults [26, 29] and in ventilated patients [28] are high as long as the operator marks the site and all subsequent images are recorded from that mark [28]. This caveat is critical because measurements of diaphragm thickness are highly variable depending on the chosen intercostal space: in a study of 150 healthy subjects [26] investigators reported as much as a 6-mm change in resting thickness from one intercostal space to another. Diaphragm thickness while healthy subjects rest at functional residual capacity (FRC) varies widely from 1.2 to 11.8 mm among individuals, with group mean values ranging from 1.6 to 3.4 mm [18, 25, 26, 29–32]. The lower limit of normal (LLN) in adults can range from 0.80 to 1.60 mm [26, 29–31]. Some of these wide variations may be due to failure to standardize body position [33], lung volume or the intercostal space where the thickness of the diaphragm is measured. The thickness of the diaphragm varies, with more inferior portions of the diaphragm being thicker than more superior portions [15]. Measurements of diaphragm thickness can be difficult in some individuals: a poor acoustic window occurs in 2–10% of ambulatory subjects [34, 35] and 5–15% of intensive care unit (ICU) patients [28, 34]. Adiposity has a detrimental effect on the quality of ultrasound imaging [27]. Ultrasound estimation of diaphragm strength and recruitment (zone of apposition) Recordings of diaphragm thickening during voluntary contractions, two-dimensional speckle tracking imaging and shear wave elastography are ultrasound-based techniques that have been used to estimate diaphragm strength. Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Diaphragm thickening The contracting diaphragm shortens and thickens [18]. This thickening can be quantified as thickening fraction (change in thickness from end exhalation to peak inhalation divided by thickness at end exhalation × 100 [18]) (Fig. 3) or as thickening ratio (thickness at peak inhalation divided by thickness at end exhalation [27]). According to some [27, 28] but not all investigators [36], diaphragm thickening during voluntary contractions correlates with inspiratory pressures [27, 28] (Fig. 4). When a correlation has been reported, there is high inter-individual variability in the relationship between diaphragmatic thickening and changes in airway pressure (Paw) [27], transdiaphragmatic pressure (Pdi) or electrical activity of the diaphragm (EAdi) [28]. A likely contributor for the high inter-individual variability is shifting recruitment of the various inspiratory muscles during inspiration [37]. Two‑dimensional speckle tracking imaging The measurement of diaphragm thickening does not assess contraction-associated longitudinal muscle shortening—the plane of muscle fiber motion [25]. Speckle tracking has the potential to describe this longitudinal shortening during diaphragm contractions [25, 36]. This technique takes advantage of the fact that ultrasound images are made up of different grey-scale pixels called speckles. A speckle- Fig. 3 Diaphragmatic thickening fraction. (Upper panel) Schematic representation of the points used to measure diaphragm thickness and formula used to calculate thickening fraction (TF). (Lower panels) Relationship of thickening fraction to lung volume expressed as a percent of inspiratory capacity (IC) in two healthy subjects. Each point represents the mean of three measurements taken from one breath. The slope of the relationship between thickening fraction and lung volume has high interindividual variability. Reproduced with permission from The American Physiological Society: Wait et al. J Appl Physiol 1989;67(4):1560–1568 Page 4 of 29 Fig. 4 Relationship between the thickening of the diaphragm recorded with ultrasonography and changes in transdiaphragmatic pressure (∆Pdi; left panel) and diaphragmatic electrical activity (∆EAdi; right panel) in five healthy subjects during a series of inspiratory maneuvers. Diaphragmatic thickening increased as transdiaphragmatic pressure (left panel) and electrical activity of the diaphragm (right panel) increased. The correlation was weak ­(r2 = 0.32 and 0.28, respectively, p < 0.01). Adapted with permission from Springer Nature: Goligher et al. Intensive Car Med 2015;41(4):642–9 tracking software follows unique groups of these pixels (known as ‘kernels’) to measure their displacement in relation to one another (deformation) [36, 38] (Fig. 5). The extent of deformation is known as ‘strain’. Negative strain values indicate kernels are coming closer together (Additional file 2). For example, a strain value of -30% indicates local muscle fiber shortening of 30%. The more negative a number, the greater the deformation and the greater the contraction (Fig. 5). During loaded breathing, strain is closely correlated with Pdi (r2 = 0.72) and EAdi (r2 = 0.60), whereas diaphragmatic thickening is not [36]. The performance of this technique under different loading conditions (isometric contractions, high inspiratory volume) is still unknown. Shear wave elastography Ultrasound shear wave elastography is an imaging method that allows real-time quantification of tissue mechanical properties [39]. Shear wave elastography relies on the estimation of the propagation velocity of shear waves generated inside tissues [39]. With this technique it is possible to calculate the shear modulus (SM) of the tissue being studied [40]. (The shear modulus, or modulus of rigidity, is defined as the ratio of shear stress to the shear strain where shear stress refers to the deforming force applied on an object, and shear strain refers to the change in size or shape that object.) In limb muscles, local muscle stiffness measured using shear wave elastography provides estimates of muscle force [41]. Chino et al. [42] were the first to report that the diaphragm’s shear modulus (SMdi) increases along with increases in Paw. The rate of increase of the shear module slowed when the pressure reached higher levels. In a Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Page 5 of 29 Fig. 5 Two-dimensional speckle tracking imaging of the diaphragm’s zone of apposition at end-exhalation (left panel) and end-inhalation (right panel). The image of the diaphragm has a granularity caused by an inherent ultrasound artifact known as speckle. A cluster of speckles form a kernel. The stronger the contraction of the diaphragm, the closer kernels come together (strain). With speckle-tracking software, it is possible to quantify the strain of the diaphragm as: 100 multiplied by the difference of the distance between two representative kernels at end-inhalation (D2) minus the distance between the same kernels at end-exhalation (D1) divided by D1. In the example above, the distance between two representative kernels at end-exhalation (left panel) is 10 mm (D1) and at end-inhalation (right panel) 6 mm (D2), yielding a strain of -40%. Reproduced under Open Access Creative Commons License: Orde et al. BMC Anesthesiol 2015;16(1):43 study by Bachasson et al. [22] investigators determined whether shear wave elastography could be used as a surrogate of Pdi in healthy subjects. In that study, mean Pdi was related to mean SMdi during closed-airways maneuvers and during inspiratory threshold loading (Additional file 3). The intra- and inter-rater agreement of SMdi measurements are yet to be determined. Ultrasound measurement of diaphragm motion (dome) The cranio-caudal movement of the dome of the diaphragm during quiet breathing [31, 43–45] and during forceful inspiratory efforts such as sniff maneuvers or maximal inspirations [43, 45, 46] can be monitored using curvilinear ultrasound probes. Curvilinear probes use low frequency ultrasound waves (2–6 Hz) [15] that penetrate deeply in the body giving a wide depth of field. On the right, operators position the probe longitudinally in the subcostal area between the mid-clavicular and anterior axillary lines using the liver as acoustic window. The probe is directed medially, cephalad and dorsally so that the ultrasound beam reaches the right dome of the diaphragm perpendicularly. On the left side, operators use the spleen as an acoustic window [45]. (Less often, operators may use the right or left lateral view (midaxillary lines) [47] or the posterior subcostal view or the subxiphoid view [15].) Once a good quality B-mode image is obtained, operators adjust the M-mode interrogation line as to be perpendicular to the movement of the hemidiaphragm [45, 48]. With M-mode ultrasonography, the diaphragm appears as a single thick echogenic line (Fig. 6). During inhalation the contracting diaphragm moves towards the ultrasound probe (Additional file 4). Diaphragm excursion are greater in men than in women [43, 45, 46, 49]. In up to 28% of patients, it is impossible to record maximal diaphragmatic excursions with M-mode ultrasonography [50]. Reported normal values of diaphragm motion during quiet breathing and deep breathing range from 2.6 to 30 mm [43, 45, 46], and 16.7 to 110.0 mm [43, 45, 46], Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Page 6 of 29 Fig. 6 Ultrasound image of the dome of the diaphragm in brightness-mode (B-mode; left panel) and motion-mode (M-mode; right panel). As the diaphragm contracts, the dome moves towards the ultrasound probe. The larger the caudal displacement of the diaphragm, the greater the diaphragmatic contribution to tidal breathing. Reproduced with permission from The American Association for Respiratory Care: Shaikh e al. Respir Care 2019;64:1600–2. The entirety of both images was obtained by Dr. Shaikh respectively. Diaphragm motion is greater posteriorly than anteriorly and greater laterally than medially [46]. The reported LLN of diaphragm excursion can range from 6.8 to 9.1 mm during resting breathing and from 29.3 to 61.8 mm during deep breathing [31, 43]. Some of these wide variations may be due to failure to standardize body position [13, 14], lung volume, and gender distribution among study participants [44, 45]. The association between diaphragm excursion and diaphragm thickening is very weak [25] and that between diaphragm excursions and diaphragm pressure output [31, 51] is weak-to-absent. A likely contributor for the limited [31] or absent [51] correlation between inspiratory pressure output and diaphragm excursions is the high inter-individual variability in recruitment of various inspiratory muscles during inspiration [37]. Non‑ultrasound imaging of the diaphragm: static imaging techniques Chest radiography Chest radiography is used to assess the position of each hemidiaphragm. An elevated hemidiaphragm suggests unilateral phrenic nerve paralysis (Fig. 7). This, however, is a nonspecific finding that can occur in several other conditions including atelectasis, pneumonia, lobectomy and pulmonary fibrosis (Fig. 8). Another limiting factor of chest radiography is the moderate interobserver agreement to detect the presence of unilateral hemidiaphragm elevation (kappa value ranging from 0.48 to 0.59) [52]. It can be difficult to detect diaphragm elevation in patients with bilateral paralysis unless chest radiographs before the onset of the paralysis are available for comparison [53]. Computed tomography CT images obtained while subjects maintain different lung volumes have been used to assess diaphragm position [54], and diaphragm dimensions in terms of thickness [55–57], surface area [58] and volume [59]. In 1987, Whitelaw [60] was the first to generate a threedimensional reconstruction of the diaphragm in one healthy subject using serial CT images. Ten years later, Pettiaux et al. [61] validated a technique of using spiral CT in four healthy subjects. Using this technique, Cassart et al. [58] studied the effect of chronic hyperinflation on diaphragm length and surface area in 10 patients with severe chronic obstructive pulmonary disease (COPD) (forced expiratory flow in one second (­FEV1) = 27 ± 6% (S.D.) predicted) with severe hyperinflation (functional residual capacity (FRC) = 225 ± 2% predicted) and 10 healthy subjects matched for age, sex, and height (Fig. 9). They concluded that patients with COPD have marked reductions in the diaphragm’s total surface area and surface area of the zone of apposition at FRC. At similar absolute lung volumes, however, diaphragm dimensions of patients were similar to those of healthy subjects. CT imaging has been proposed as a means to measure crural diaphragm thickness in ventilated patients [55] Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Page 7 of 29 Fig. 7 Phrenic nerve injury and diaphragm dysfunction in a patient after coronary artery bypass surgery (CABG). (Top panels) Initial posteroanterior (left) and lateral (right) films demonstrate both hemidiaphragms in a relatively normal position. (Bottom panels) After CABG, posteroanterior (left) and lateral (right) films demonstrates new elevation of the left hemidiaphragm, suggestive of postoperative phrenic nerve injury. Reproduced under Open Access Creative Commons License: Kokatnur et al. Diseases 2018:6(1):16 (Fig. 10) and in patients with suspected diaphragm paralysis [56]. Unfortunately, there is no consensus on which area of the muscle should be measured and at which lung volume [57]. Accordingly, the role of CT measurement of the crural diaphragm thickness remains uncertain. Spiral CT has been used to calculate the volume of the diaphragm [59]. With this technique, Jung et al. [59] reported that, upon ICU admission, the diaphragm’s volume in 23 critically ill patients (14 of whom were septic) was not different from the volume of the diaphragm in 17 control patients. Twenty-five days after admission, the volume of the diaphragm had decreased by 11 ± 13% in nonseptic patients and by 27 ± 12% in septic patients (p = 0.01) (Fig. 11). Upon ICU admission, diaphragm volume only weakly correlated with diaphragm strength as measured by airway twitch pressure (PawTw) elicited by magnetic stimulation of the phrenic nerves. To date, no investigator has validated the accuracy of spiral CT imaging to calculate the volume of the diaphragm. Static magnetic resonance imaging Static MRI obtained while subjects maintain different lung volumes can be used to assess the diaphragm’s shape [62], position [63], thickness [54] and surface area [58, 64]. Using static MRI in four healthy subjects, Paiva et al. [65] concluded that the shape of the dorsal half of the relaxed diaphragm in the supine position at FRC can be explained by the Laplace law—i.e., the lung and abdominal contents do not impose their shape on the Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Page 8 of 29 Fig. 8 Right hemidiaphragm elevation in a patient with idiopathic pulmonary fibrosis and dyspnea. Posteroanterior (a) and lateral (b) radiographs show pulmonary fibrosis with low lung volume and elevation of the right hemidiaphragm, concerning for paralysis. Axial (c) and coronal (d) CT images demonstrate diaphragm crura (arrows) without thinning, arguing against paralysis. CT images with lung window views in the axial (e) and coronal (f) plane show more lung fibrosis on the right than on the left. Right hemidiaphragm elevation likely reflects the greater degree of fibrosis and volume loss on the right. Reproduced with permission from Wolters Kluwer Health: Sukkasem et al. Journal of Thoracic Imaging 2017;32(6):383– 390 diaphragm. In a subsequent static MRI study, Gauthier et al. [64] reported that, with lung inflation, the dimension of the zone of apposition is a function of diaphragm shortening and, to a lesser extent, widening of the lower rib cage. The investigators concluded that the diaphragm is more accurately modeled by a "widening piston" (Petroll’s model) than a simple "piston in a cylinder" model. More recently, Cluzel et al. [62] evaluated the use of three-dimensional reconstruction of the static MRI of the thorax in five healthy subjects while supine. Compared to measurements of residual volume (RV) and total lung capacity (TLC) obtained using a spirometer and helium dilution technique, measurements obtained with static MRI tended to overestimate RV and underestimate TLC. Between RV and TLC, the mean volume under the dome of the diaphragm decreased by 66%, and the mean volume of the cavity delimited by the rib cage increased by 23%. The diaphragm contributed to 60% of the inspiratory capacity (Fig. 12). In patients with late-onset glycogen storage disease type II, also known as late-onset POMPE disease, Gaeta et al. [54, 66] reported that diaphragm atrophy and maximal diaphragm excursions assessed with static MRI, correlate with forced vital capacity (FVC) in the supine position, standing-to-supine decrease in FVC, peak cough flow and maximal inspiratory pressure. Similar results were reported by Wens et al. [67] (Fig. 13). In summary, chest radiograph provides limited information pertaining to diaphragm morphology [52]. The more advanced static imaging techniques, CT and MRI, provide useful information regarding the surface area and positioning of the entirety of the diaphragm within the thorax. As such, using CT and MRI, it is possible to correlate changes in diaphragm position with changes in lung volume [58, 62], and elucidate the role of the diaphragm in pulmonary disease states [54, 58]. In addition, Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Page 9 of 29 diaphragm paralysis. In patients with bilateral diaphragm weakness, the abdominal wall muscles relax at the onset of inspiration. The diaphragm descends due to the outward recoil of the abdominal wall; this movement can be misinterpreted as normal diaphragm contraction. Dynamic magnetic resonance imaging Fig. 9 Three-dimensional reconstructed images of the diaphragm at functional residual capacity in (a) a control subject and (b) a patient with COPD. The numerical units represent centimeters. Compared to the control subject, the patient with COPD has a markedly smaller muscle surface area. Reproduced with permission from The American Thoracic Society: Cassart et al. Am J Respir Crit Care Med 1997;156:504–508 images obtained with CT and MRI can be sufficiently detailed to detect clinically important changes in the thickness of the muscle itself; CT and MRI may become a useful tool in assessing for diaphragm atrophy [55, 56, 66]. Non‑ultrasound imaging of the diaphragm: dynamic imaging techniques Fluoroscopy Fluoroscopic imaging during a sniff maneuver is the traditional methodology used to diagnose unilateral diaphragm paralysis. During a sniff maneuver in patients with unilateral diaphragm paralysis the healthy hemidiaphragm descends, (Additional file 5) whereas the affected hemidiaphragm paradoxically ascends (Additional file 6). For the study to be considered abnormal, the affected hemidiaphragm must ascend at least 2 cm. This test has a number of limitations: it is not highly specific (6% of healthy subjects demonstrate paradoxical motion) [68] and results can be misleading in cases of incomplete paralysis or bilateral diaphragm weakness [69] (Additional file 7). Newsom-Davis et al. [69] observed paradoxical motion in less than 20% of patients with bilateral Dynamic MRI has been used to study the mechanisms responsible for the diaphragm’s shape at FRC [65], diaphragm motion in different body postures [13, 70] and to quantify the volume displaced by the contraction of the diaphragm [62]. In 1995, Gierada et al. [63] tested the feasibility of using dynamic MRI to assess diaphragm motion during slow vital capacity maneuvers. Dynamic MRI was obtained in ten healthy volunteers in the supine position. The mean excursion of the hemidiaphragm dome was 4.4 ± 0.4 (S.E.) cm on the right and 4.2 ± 0.3 cm on the left. Diaphragm displacement revealed a gradient of excursion that increased from anterior, to middle, to posterior (p < 0.05). Excursion of the lateral aspect of both hemidiaphragms was greater than that of the corresponding medial aspect (p < 0.001). Takazakura et al. [13], expanded on these findings by recording dynamic MRI in ten healthy men while sitting and while supine. The investigators reported that the movement of the diaphragm during a slow vital capacity (VC) maneuver in the supine position was greater than in the sitting position, most notably, posteriorly. The mean craniocaudal excursion of the posterior portion of the right hemidiaphragm was 10.25 ± 1.96 (S.E.) cm in the supine position and 7.95 ± 2.25 cm in the sitting position. The corresponding values for the left hemidiaphragm were 9.23 ± 2.05 cm and 8.04 ± 2.41 cm, respectively. The coordinated contraction of diaphragm and abdominal muscles increases intra-abdominal pressure. This increase stiffens the lumbar spine and contributes to spinal stabilization during trunk and voluntary limb movements [71]. Dynamic MRI has been used to examine this stabilizing function of the diaphragm during postural limb activities [71]. In thirty healthy subjects, Kolar et al. [71] obtained dynamic MRI during tidal breathing while subjects relaxed all four extremities along the torso, and while maintaining isometric flexion of the upper or of the lower extremities against external resistance. Tidal excursions of the diaphragm were greater during upper and lower extremity contraction than during relaxed condition (p < 0.05). In addition, the position of the diaphragm at end inhalation during upper or lower extremity contraction was lower (more caudal) than during relaxed conditions (p < 0.01). The position of the diaphragm at end exhalation was lower (more caudal) during lower Laghi Jr. et al. BMC Pulm Med (2021) 21:85 Page 10 of 29 Fig. 10 Assessment of crural diaphragm thickness by chest CT using axial and coronal images. (Left panel) With axial imaging, the thickness of the right and left crural hemidiaphragms (arrows) can be measured at level of the origin of the celiac artery (arrowhead). Note the nodularity of the left crus, a normal variant of the diaphragm’s shape. (Right panel) With coronal imaging, the thickness of the crural hemidiaphragms (arrows) can be measured at level of the first lumbar vertebra Fig. 11 Computed tomography measurements of diaphragm volume in a critically ill, septic patient on admission to the intensive care unit (left panel) and ten days later (right panel). Sepsis was associated with a decrease in diaphragm volume. The images were provided, with permission, by Drs. Boris Jung, Stephanie Nougaret and Samir Jaber, University Hospital of Montpellier, France extremity contraction than during upper extremity contraction or during relaxed conditions (p < 0.01). The latter finding suggests that during lower extremity contraction, the diaphragm does not relax fully and remains in higher tonic state of activity. In turn, the higher tonic state of activity supports critical involvement of the diaphragm in stabilizing the spine during postural activity. Dynamic MRI has been used for the evaluation of diaphragm movements in patients with COPD. Unal et al. [72], reported that during slow VC maneuvers, the cephalocaudal excursion of the dome of the diaphragm in 26 patients with COPD was less than half that recorded in 15 healthy subjects. Later, these same investigators reported an increase in diaphragm excursion following