AUTOMATED LUNG SEGMENTATION IN DIGITAL LATERAL CHEST RADIOGRAPHS

We are developing a fully automated computerized scheme for segmenting the lung fields in digital lateral chest radiographs. Existing comput er-aided diagnostic (CAD) schemes and automated lung segmentation meth ods have focused exclusively on the posteroanterior view, despite the diagnostic importance of the lateral view. Information from the latera l radiograph is routinely incorporated by radiologists in their decisi on-making process, and thus computer analysis of lateral images may po tentially add another dimension to current CAD schemes. Automated anal ysis of the lung fields in lateral images will necessarily require acc urate segmentation. Our scheme employs an initial procedure to elimina te external and subcutaneous pixels. Global gray-level histogram analy sis then allows for the identification of a range of gray-level thresh olds. An iterative gray-level thresholding scheme is implemented using this range of thresholds to construct a series of binary images in wh ich contiguous regions are identified and geometrically analyzed. Regi ons determined to be outside the lungs are prevented from contributing to binary images at later iterations. Adaptive local gray-level thres holding is applied along the initial contour that results from the glo bal thresholding procedure to extend the contour closer to the true lu ng borders. This local thresholding method uses regions of interest of various dimensions, depending on the enclosed anatomy. Smoothing and polynomial curve fitting complete the segmentation. A database of 100 normal and 100 abnormal lateral images was analyzed. Quantitative comp arison of computer-segmented lung regions with lung regions manually d elineated by two radiologists indicated that 83% and 84% of normal and abnormal images, respectively, displayed segmentation contours within three standard deviations of the mean inter-radiologist contour degre e-of-overlap value. (C) 1998 American Association of Physicists in Med icine.