Inhaled noble gas magnetic resonance imaging (MRI) has emerged as a research tool for the quantitative evaluation of parenchymal abnormalities in a variety of lung pulmonary diseases including emphysema, bronchopulmonary dysplasia, congenital lobar emphysema and alpha 1 antitrypsin deficiency. Recently, a stretched exponential model (SEM) was proposed for the evaluation of hyperpolarized gas multiple b-value diffusion-weighted MRI¹ as a more straightforward alternative to existing approaches.^2,3 The stretched exponential model is not constrained to a specific range of diffusion times, thus, it may be able to reveal information about the lung microstructure at different length scales. Another major advantage is the possibility of simple and straightforward analyses of multiple b-value data. On the other hand, until now, this simplified approach has not been shown to provide clinically-relevant information about the lung parenchyma microstructure such as mean linear intercept (L_m), which is a drawback. However, we hypothesize that there is a correlation between the diffusion scale (L_D) estimated using SEM and the measurement of L_m estimated using the morphometry approach. Therefore, in this proof-of-concept evaluation, our objective was to evaluate L_D and L_m estimates in a small group of never-smokers (NS) as well as ex-smokers with (ESE) and without (ESnE) emphysema (ESnE).As shown in Table 1, 34 subjects including 14 never-smokers, 12 ex-smokers without emphysema and eight ex-smokers with emphysema provided written informed consent to an ethics-board approved study protocol and underwent spirometry, plethysmography, CT and ³He MRI. Imaging was performed at 3.0T (MR750, GEHC, Waukesha WI) using whole-body gradients (5G/cm maximum) and a commercial, rigid linear human RF coil (Rapid Biomedical, Germany). In a single breath-hold, five interleaved acquisitions (TE=3msec, TR=5.0msec, matrix size=128x128, number of slices=7; slice thickness=30mm, and FOV=40x40cm) with and without diffusion sensitization were acquired for a given line of k-space to ensure that RF depolarization (4^o constant flip angle was used) and T₁ relaxation effects (scan time was 2sec per slice) were minimal. The diffusion-sensitization gradient pulse ramp up/down time=500μs, constant time=460μs and diffusion time (Δ)=1.46ms and this resulted in images acquired at five different b values: 0, 1.6, 3.2, 4.8 and 6.4s/cm². A diffusion time of 1.46 ms was used in order to provide ³He diffusion sensitivity to alveolar length scales as previously described³. As previously described^1,4, maps of L_m, diffusivity (DDC) and heterogeneity index (α) along with two b-value (0 and 1.6 s/cm²) ADC, were computed from diffusion-weighted images on a pixel-by-pixel basis.⁴ Figure 1 shows representative centre slice ADC, L_m, DDC and α maps for a single NS, ESnE and ESE subject each while Table 1 shows mean estimates of L_m, DDC and α. Figure 2 shows the experimental non-linear (the second order polynomial) dependencies between the diffusion scale ([2ΔD_o]^1/2) and the mean linear intercept observed for NS (R=0.90), ESnE (R=0.74) and ESE (R=0.94). Figure 2 also shows the experimental dependencies between DDC and α for NS (R=0.95, linear dependence), ESnE (R²=0.81, single exponential dependence) and ESE (R²=0.95, single exponential dependence). There were strong correlations for estimates of mean L_D and L_m in the three different subject subgroups. These findings suggest that the stretched exponential model may be used to estimate clinically-relevant parenchyma microstructure measurements such as the mean linear intercept. However, the nature of the experimentally-observed dependencies is not well-understood and require further mathematical modeling and theory development. In addition, the strongest correlations for mean L_D and L_m estimates were observed in subjects with a narrow distribution of morphometric parameters and this also requires further investigation. Regardless, the application of diffusion-weighted inhaled gas MRI to provide estimates of subvoxel acinar duct morphology is an important breakthrough in our understanding of the wide variety of parenchyma abnormalities that accompany smoking-related emphysema, congenital emphysema and bronchopulmonary dysplasia. Perhaps the greatest impact can be derived from the longitudinal evaluation of emphysema related to alpha-one antitrypsin deficiency, where lung biomarkers of disease progression and treatment efficacy are lacking. We are developing these diffusion-weighted MRI measurements for a better understanding of the differences in these parenchyma diseases and these proof-of concept results provide a foundation to build these modelling approaches.