Automated detection of lung nodules in CT scans: Preliminary results

We have developed a fully automated computerized method for the detection o f lung nodules in helical computed tomography (CT) scans of the thorax. Thi s method is based on two-dimensional and three-dimensional analyses of the image data acquired during diagnostic CT scans. Lung segmentation proceeds on a section-by-section basis to construct a segmented lung volume within w hich further analysis is performed. Multiple gray-level thresholds are appl ied to the segmented lung volume to create a series of thresholded lung vol umes. An 18-point connectivity scheme is used to identify contiguous three- dimensional structures within each thresholded lung volume, and those struc tures that satisfy a volume criterion are selected as initial lung nodule c andidates. Morphological and gray-level features are computed for each nodu le candidate, After a rule-based approach is applied to greatly reduce the number of nodule candidates that corresponds to non-nodules, the features o f remaining candidates are merged through linear discriminant analysis. The automated method was applied to a database of 43 diagnostic thoracic CT sc ans. Receiver operating characteristic (ROC) analysis was used to evaluate the ability of the classifier to differentiate nodule candidates that corre spond to actual nodules from false-positive candidates. The area under the ROC curve for this categorization task attained a value of 0.90 during leav e-one-out-by-case evaluation. The automated method yielded an overall nodul e detection sensitivity of 70% with an average of 1.5 false-positive detect ions per section when applied to the complete 43-case database. A correspon ding nodule, detection sensitivity of 89% with an average of 1.3 false-posi tive detections per section was achieved with a subset of 20 cases that con tained only one or two nodules per case. (C) 2001 American Association of P hysicists in Medicine.