Both the detection and interpretation of focal abnormalities on chest X-ray (CXR) are challenging tasks. CXR accuracy depends on the view (e.g. the supine view has limited sensitivity) and technological equipment. The detection of small focal abnormalities (e.g. lung nodules) varies between anatomical regions according to the presence of dense anatomic structures, such as the bones and the hila. The interpretation of focal abnormalities on CXR is paramount within the whole clinical assessment, because CXR findings can guide the patient’s management, or warrant further investigations, such as computed tomography. Focal lung abnormalities on CXR are still a cornerstone of diagnostic algorithms; however, the radiological approach has progressively changed in the last decade because of the progressive development of both hardware and software applications that enable sensitive detection and accurate characterization.
Feature based analyses of lung nodules from computed tomography (CT) images
