This project is an extension of previous work involving the creation, development and testing of image-processing techniques designed to improve diagnostic performance. Current work has centered on methods relevant to the processes of radiologic image subtraction and tomosynthesis. Work continues in the area of automatic manipulation to facilitate registration. Particular emphasis has been placed on: 1) methods for automatically segmenting radiographic images into regions of diagnostic interest, in anticipation of the automated detection of associated lesions, 2) methods of quantifying the apparent size of lesions from these images, 3) methods for increasing the efficiency of complex, spatial-frequency-dependent manipulations essential for optimization of diagnostic performance of specific tasks. The recognition and delineation of areas showing trabecular bone was set as a primary target because of its importance in the diagnosis and monitoring of periodontal and other lytic bone diseases. Efforts continue in the use of second-order gray-level statistics for that purpose. Recent findings indicate that fractal geometry may provide a realistic model to characterize and segment trabecular bone. Other work involves the use """"""""shaded aperture"""""""" sampling techniques to eliminate ringing artifacts associated with high-pass filtered tomosynthetic reconstructions. Practical results show that optimum weighting of 25 projections permits substantial suppression of artifacts caused by structures lying in other planes. Significant progress has been achieved in methods to quantitate lesion size. An automated procedure has been implemented to estimate bone mass loss that may be either focal or distributed over some area. The use of Fisher's linear discriminant for detection threshold setting, followed by morphological image processing attains a precision 0.5 mg in dental and orthopedic applications.
