Computed tomography of the thorax in the diagnosis and management of malignant disease

Detection of occult cardiac invasion by two dimensional echocardiography in patients with bronchial carcinoma

Cardiac invasion by bronchial carcinoma may prevent successful resection but may be undetected before operation. In a retrospective analysis of 100 consecutive thoracotomies nine patients had unsuspected cardiac invasion by tumour. A prospective study of preoperative two dimensional echocardiography was therefore undertaken in patients with bronchial carcinoma who had no clinical evidence of cardiac tumour. Comparison with anatomical findings was possible in 65 patients in whom an echocardiogram of suitable quality had been obtained. There was one false negative among 55 negative echocardiograms and three false positives among 10 positive echocardiograms; non-malignant pericardial disease accounted for the echocardiographic finding in one of the latter. The predictive value of a negative test was 98%, and the predictive value of a positive test was highest (80%) if the echocardiogram suggested atrial invasion.

Magnetic resonance imaging of the chest: a diagnostic comparison with computed tomography and hilar tomography

Computed tomography (with and without contrast enhancement) provides excellent diagnostic accuracy for the evaluation of the chest. Oblique (55 degrees) and anteroposterior hilar tomography is accurate for the evaluation of hilar nodes and masses. Magnetic resonance techniques provide excellent differentiation of vascular and nonvascular structures and therefore should be useful in the hilum and mediastinum. Magnetic resonance imaging was used in 55 patients with known pathologic conditions in the mediastinum, hilum, and lungs to determine the accuracy and efficacy of this technique compared with computed and hilar tomography. The pathologic conditions included primary and metastatic neoplasms, benign masses, vascular abnormalities, and pulmonary nodules and infiltrates. Spatial resolution with magnetic resonance imaging is less than with computed tomography with our instrument (0.15 T resistive magnet). However, in the hilum and mediastinum, magnetic resonance imaging provided diagnostic information equal to that of computed tomography with contrast in 90% of patients. Vascular and nonvascular structures were more easily differentiated than with hilar tomography. Computed tomography was far superior in the evaluation of multiple pulmonary nodules. Lesions of the chest wall were better seen with magnetic resonance imaging because of the improved soft tissue contrast.