Purpose: An experimental study using porcine lung explants and a dedicated chest phantom to evaluate the signal intensity of artificial alveolar infiltrates with T 1 - and T 2 -weighted MRI sequences.
Material and methods: 10 porcine lung explants were intubated, transferred into the cavity of a MRI-compatible chest phantom and inflated by continuous evacuation of the artificial pleural space. All lungs were examined with MRI at 1.5 T before and after intra-tracheal instillation of either 100 or 200 ml gelatine-stabilised liquid to simulate alveolar infiltrates. MR-examination comprised gradient echo (2D- and 3D-GRE) and fast spin echo sequences (T 2 -TSE and T 2 -HASTE). The signal intensity of lung parenchyma was evaluated at representative cross sections using a standardised scheme. Control studies were acquired with helical CT.
Results: The instilled liquid caused patchy confluent alveolar infiltrates resembling the findings in patients with pneumonia or ARDS. CT revealed typical ground-glass opacities. Before the application of the liquid, only T 2 -HASTE and T 2 -TSE displayed lung parenchyma signals with a signal/noise ratio of 3.62 and 1.39, respectively. After application of the liquid, both T 2 -weighted sequences showed clearly visible infiltrates with an increase in signal intensity of approx. 30 % at 100 ml (p < 0.01) and 60 % at 200 ml (p < 0.01). With 2D- and 3D-GRE the infiltrates were not visible, although the lung parenchyma signal increase was statistically significant. On 2D-GRE the increase in signal intensity reached 0.74 % (p = 0.32) after 100 ml and 5.6 % (p < 0.01) after 200 ml (for 3D-GRE: 2.2 % [p = 0.02] at 100 ml and 4.4 % at 200 ml [p < 0.01]). The CT controls revealed a significant increase of lung density of 17 H.E. at 100 ml (p = 0.02) and 75 H.E. at 200 ml (p < 0.01).
Conclusions: MRI with T 2 -weighted sequences detects artificial alveolar infiltrates with high signal intensity and may be a highly sensitive tool to detect pneumonia in patients.