Imaging of lung ventilation has been addressed by means of hyperpolarized gas ventilation. Its widespread use is limited by the requirements of expensive isotopes and polarizers. As alternative, the inhalation of inert flourinated (19F) gases has been proposed. Here gases with a high number of 19F atoms, like sulfurhexafluorid (SF₆) or hexafluorethan (C₂F₆), which are inexpensive and nontoxic, have been used. In recent pre-clinical works, intubation have been used to ensure sufficient gas concentration in the lungs [1].

In this work we have investigated the use of SF₆ in combination with ultra-short echo time (UTE) imaging techniques for imaging of the wash-in and wash-out of the gas during free breathing as well as to image the steady-state gas distribution in mice lungs. SF₆ is characterized by ultrashort T1 and T2* values well below 2 ms [2]. UTE imaging techniques in combination with short repetition times (TR) and high flip angles appear promising for yielding good SNR in the final images.

Three two-years-old female BalbC mice were investigated. These old animals were used as a model for age-related decline of lung functions. They were imaged with an 11.7 T small animal system (BioSpec 117/16, Bruker, Ettlingen, Germany) using a 1H/19F Tx/Rx volume coil. Mice were anesthetized with an intraperoneal injection of ketamin (2 mg/30 g) and xylazine (0.4 mg/30 g). Gas was delivered through a commercially available mask positioned outside the coil.

Under air ventilation, a 1H 3D UTE scan was acquired for lung anatomy (TR = 4 ms, TE= 0.008 ms, FA= 3°, pixels 128x128x128, FOV 40x40x40 mm, acquisition time 3min25s). For 19F imaging, the delivered gas was switched to a mixture of 70% SF₆ and 30% O₂. For imaging the inflow characteristics, a 19F 2D UTE acquisition was performed (1 cm-thick coronal slice, FOV 40x40 mm, pixel =128x128, TE =0.15 ms, TR =4.2 ms, FA = 90°, 300-fold oversampling, acquisition time 8min26s, golden angle ordering). For 3D assessment of the SF₆ distribution in the lungs, a 19F 3D UTE scan was subsequently measured during SF₆/O₂ ventilation (TR =2 ms, TE =0.046 ms, FA =60°, pixels 128x128x128, FOV 40x40x40 mm, 4-fold oversampling, acquisition time 6min50s). For assessment of the washout, a second 19F 2D UTE scan was performed after the delivered gas was switched back to air (same acquisition parameters, but 120-fold oversampling, acquisition time 3min25s).

All images were reconstructed with in-house developed software implemented in Matlab (The MathWorks, Natick, MA-US). 19F images were then filtered applying a median filter (2D UTE: 3x3 pixel, 3D UTE: 3x3x3 pixel). Temporal resolved images were reconstructed from 19F 2D UTE data with a sliding window protocol of 500 spokes, an overlap of 250 spokes between subsequent windows and a temporal resolution of 2.1 s.

Region of Interest (ROIs) where manually defined over the lungs to determine the variation of signal intensity over time.

Figure 1 shows the resulting image quality achieved with 19F 2D UTE. A ventilation defect is clearly visible in the left lung. The respective anatomic shows abnormal high intensity in the lung parenchyma of the right lung, probably responsible for the ventilation deficit.

Figure 2 shows the resulting temporal resolved images of the same dataset of figure 1 and the wash-in curves for both lungs. Uptake of gas in the right lung is fast, taking about 6 second to reach to a plateau, while the damaged left lung shows a slower uptake. Wash-out curves and images for the same animal and ROIs are shown in figure 3. Wash-out of the right lung is slightly faster than in the left lung.

Static 19F 3D UTE images of the same animal are shown in figure 4 next to the corresponding 1H 3D UTE. Pattern of ventilation deficits and abnormal high intensity in lung tissue from proton images shows a good correspondence. Both lungs appears to be compromised, with the exception of the posterior right lung.

In this work we have shown the feasibility of implementing a simple and fast protocol to visualize wash-in, wash-out and 3D distribution of SF₆ gas in mice during free breathing. No special equipment, intubation or complex acquisition methods were used.

Further studies will be necessary for validation, but our results suggest that this protocol may allow non-invasive quantification of lung function and detection of ventilation deficits in animal models of common lung diseases.