Conventional methods for the segmentation of lung fields from thorax CT scans are based on thresholding. They rely on a large grey value contrast between the lung parenchyma and surrounding tissues. In the presence of consolidations or other high density pathologies, these methods fail. For the segmentation of such scans, a lung shape should be induced without relying solely on grey level information. We present a segmentation-by-registration approach to segment the lung fields from several thin-slice CT scans (slice-thickness 1 mm) containing high density pathologies. A scan of a normal subject is elastically registered to each of the abnormal scans. Applying the found deformations to a lung mask created for the normal subject, a segmentation of the abnormal lungs is found. We implemented a conventional lung field segmentation method and compared it to the one using non-rigid registration techniques. The results of the algorithms were evaluated against manual segmentations in several slices of each scan. It is shown that the segmentation-by-registration approach can successfully identify the lung regions where the conventional method fails.