Abstract: A method for representing digitized scanned image data obtained by a non-intrusive cross-sectional image generating device. By representing an original set of user-input data points in polar coordinates, such that all of the points lie in the upper two quadrants of the coordinate system, a cubic spline algorithm can be applied once to generate a series of polynomial functions that describe a curve connecting all of the input data points. The method also uses the calculated cubic spline coefficients for the initial curve to determine intermediate data points at each of which a uniform length perpendicular is constructed. By traversing each perpendicular a desired length, new data points equidistant from the initial curve are calculated, resulting in additional curves equidistant from the initial curve. Any of the curve representations generated can then be used to select pixels from the scanned image data to display a reformatted view of the original data.

Abstract: A method for generating synthesized cross-sectional images perpendicular to a line or curve that approximates the shape of a structure scanned by a non-intrusive cross-sectional image generating device, such as a CT, MR, or ultrasonic scanner. The inventive method involves three basic processes: generation of a set of perpendicular lines along an initial curve that approximates the curvature of the scanned structure; selection from each original scanned image of pixels corresponding to discrete points on each of the generated perpendicular lines; and interpolation between adjacent lines of pixels collected in the second step to generate a complete perpendicular cross-sectional image. The invention permits making a single pass through all axial image data, during which selected pixels for a perpendicular cross-sectional view are collected in a temporary data file for subsequent interpolation for display purposes.