The present work targets scientists with an interest in quantitative measurements of microscopic parameters in peripheral lung tissue and also addresses researchers interested in experimental and clinical aspects of pulmonary MRI and signal formation.The line shape in human lung tissue is mainly affected by strong magnetic field inhomogeneities that are caused by the susceptibility difference between air-filled alveoli and the surrounding tissue. However, the diffusion process of spin-bearing particles around alveoli influences the signal formation as well. A detailed analysis of the interplay of susceptibility and diffusion effects allows connecting the specific line shape to microscopic tissue parameters. Thus, local line shape measurements can be used to examine the respective local lung microstructure.

Line shapes in human lung tissue in inspiration and expiration were measured in two healthy volunteers at 1.5 Tesla (Siemens Aera) and 3 Tesla (Siemens Prisma) with a PRESS-sequence for cubic voxels (15mm x 15mm x 15mm) and $$$TR = 1.5 s$$$, $$$TE = 30 ms$$$ and 20 averages.

To relate these measurements to microscopic tissue parameters, we used the standard "single sphere approximation" [1], where only one alveolus is considered in a locally uniform assembly of alveoli (see Fig. 1). In this approximation it suffices to consider the dephasing process of the local magnetization between the spherically shaped alveolus with radius $$$R$$$ and a surrounding spherical shell with radius $$$R_D$$$, where $$$R_D$$$ depends on the local air volume fraction $$$\eta = R^3/R_D^3$$$. The susceptibility difference $$$\Delta\chi$$$ in a static field $$$B_0$$$ between an air-filled alveolus and the surrounding tissue generates a local dipole field with strength $$$\delta\omega \propto \gamma B_0 \Delta\chi$$$ and the diffusion process of spin-bearing particles is described by the diffusion coefficient $$$D$$$ and the correlation time $$$\tau \propto R^2/D$$$.

Both effects are mathematically combined within the Bloch-Torrey-equation [2]. Careful mathematical analysis yields an analytical expression for the line shape $$$p(\omega)$$$ in dependence of the microscopic tissue parameters such as the alveolar radius $$$R$$$, diffusion coefficient $$$D$$$ and air volume fraction $$$\eta$$$.

A mathematical analysis of the line shape suggests a natural classification into two different dephasing regimes: in the diffusion- regime that represents a strong influence of diffusion- related effects, the line shape follows a Lorentzian form. However, for only small influences of diffusion effects (in the so-called strong-dephasing regime), the line shape is asymmetric. In this latter regime, the line shape can be approximated as a convolution of the static dephasing line shape $$$p_0(\omega)$$$ as obtained by Cheng et al. [3] and a Lorentzian line shape, whose width is determined by the diffusion process:

$$p(\omega) \approx \frac{\tau\delta\omega}{\pi} \int \limits_{-\infty}^{+\infty} \frac{p_0(\omega-\omega')}{1+\tau^2\omega'^2} d \omega'.$$

The measured line shapes for a PRESS-voxel are compared with theoretical predictions in Fig. 3: as expected, the width of the line shape increases for increasing magnetic field strength $$$B_0$$$ and constant diffusion effects, since this corresponds to an increase in local dipole field strength $$$\delta\omega$$$. Furthermore, the asymmetry of the measured line shape depends on air volume fractions $$$\eta$$$ that vary in inspiration and expiration.

In this work, the local line shape in human lung tissue is linked to microscopic tissue parameters, such as alveolar radius, diffusion coefficient and susceptibility difference. In vivo measurements in human lung tissue at 1.5T and 3T agree very well with theoretical predictions. Thus, the local alveolar radius as well as local air volume fraction $$$\eta$$$ can be estimated [4]. The presented evaluation of microscopic tissue parameters has several advantages: the measurements can be performed with standard sequences and the determination of microscopic tissue parameters is robust. Furthermore the measurements do not need contrast agents. Thus, the presented method is of clinical interest, since it improves the quantitative diagnosis of early pulmonary fibrosis and emphysema.