Diffusion-weighted MR-imaging for the detection of pulmonary nodules at 1.5 Tesla: intraindividual comparison with...

To investigate the feasibility of diffusion-weighted imaging (DWI) MRI for detecting pulmonary nodules at 1.5 Tesla in comparison with standard multidetector computed tomography (MDCT).

Twenty patients with disseminated cancer disease in which MDCT had assured the presence of at least one pulmonary nodule were examined using a respiratory-gated DWI MR-sequence. Grey scale inverted source images and coronal maximum intensity projection (MIP) images were consensually analysed by two experienced radiologists. Size and location of any nodule detected were assessed. Additionally, the readers evaluated each hemithorax for the presence of at least one nodule and applied a four-point conspicuity scale (1-hemithorax definitely affected; 4-hemithorax definitely not affected). MDCT data served as reference.

At MDCT, a total of 71 pulmonary noduIes was found (size 3-5mm, n=16; 6-9mm, n=22; ≥10mm, n=33). For the DWI MR-sequence, a sensitivity of 86.4% was calculated for nodules ranging 6-9mm and 97% for nodules ≥10mm. In contrast, only 43.8% of lesions ≤5mm was detected. The separate analysis of each hemithorax for the presence of at least one pulmonary nodule revealed a specificity rate, PPV and NPV of DWI-MR of 92.3%, 96% and 80%, respectively.

The presented study is the first to confirm the diagnostic potential of DWI-MR in the detection of solid lung nodules. This technique allows for the detection of nodules ≥6mm with reasonably high sensitivity rates (>86%). The observation of false positive findings decreases the accuracy of this approach compared with MDCT.