Radiological diagnosis in lung disease: factoring treatment options into the choice of diagnostic modality.

BACKGROUND Chest X-ray, computed tomography (CT), and magnetic resonance imaging (MRI) each have characteristic advantages and disadvantages that need to be considered in clinical decision-making. This point is discussed in reference to the main types of lung disease that are encountered in practice. METHOD A selective literature search was performed in the PubMed and Google Scholar databases. Existing clinical guidelines on the main types of lung disease and studies concerning radiological diagnosis were also con - sidered in this review. RESULTS There have been no more than a few large-scale, controlled comparative trials of different radiological techniques. Chest X-ray provides general orientation as an initial diagnostic study and is especially useful in the diagnosis of pneumonia, cancer, and chronic obstructive pulmonary disease (COPD). Multi-detector CT affords nearly isotropic spatial resolution at a radiation dose of only 0.2-5 mSv, much lower than before. Its main indications, according to current guidelines, are tumors, acute pulmonary embolism, pulmonary hypertension, pulmonary fibrosis, advanced COPD, and pneumonia in a high-risk patient. MRI is used in the diagnosis of cystic fibrosis, pulmonary embolism, pulmonary hypertension, and bronchial carcinoma. The positive predictive value (PPV) of a chest X-ray in outpatients with pneumonia is only 27% (gold standard, CT); in contrast, an initial, non-randomized trial of MRI in nosocomial pneumonia revealed a PPV of 95%. For the staging of mediastinal lymph nodes in bronchial carcinoma, MRI has a PPV of 88% and positron emission tomography with CT (PET/CT) has a PPV of 79%, while CT alone has a PPV of 41% (gold standard, histology). CONCLUSION The choice of radiologicalal technique for the detection, staging, follow-up, and quantification of lung disease should be based on the individual clinical options, so that appropriate treatment can be provided without excessive use of diagnostic testing.