Koninklijke Philips Corporation N.v.

Guy A Lavi

Jonathan Lessick

Michael Grass

Claas Bontus

Udo Van Stevendaal

Gilad Shechter

Yoav Bar

Carsten Oliver Schirra

Asher Gringaus

A method for visualizing a region of interest within a viewport includes presenting tomographic image data as an image in a viewport using a set of global image processing operations, presenting a viewing window in the main viewport; and using a second, different set of image processing operations that is local to the viewing window, wherein the second, different set of image processing operations reduces image artifact with respect to a desired structure of interest in the viewing window.

Image viewing window

The invention relates to an imaging system for imaging a region of interest comprising a moving object, which moves less in small motion phases than in large motion phases. Detection values are provided and a small motion determination unit () determines the motion of the object in the region of interest in the small motion phases from the 5 detection values. A large motion determination unit () determines the motion of the object in the large motion phases from the determined motion of the object in the small motion phases. A reconstruction unit () reconstructs an image of the region of interest from the detection values, wherein the reconstruction unit () is adapted for performing a motion compensation using the determined motions in the small and large motion phases.

Imaging system for imaging a region of interest in different phases

Projection data () acquired during a tomographic examination of a periodically moving object is used to reconstruct a plurality of image layers (). The image layers () are combined to generate image data at a desired phase of motion. To generate a weighting function used to combine the image layers, a reference weighting function () is generated at the desired phase. The image layers () are weighted to approximate the first weighting function (). The number of image layers and the size of a sub-region of interest are advantageously selected so that the various image layers can be stored in a relatively high speed memory portion of a computer.

Dynamic computed tomography imaging

A coronary arteries tree is approximated by a base sphere () which is best fitted to vessels centerlines (). The base surface () is gridded to define pixels (). The base sphere () is mapped to fit the centerlines () such that a true form surface () is determined. A wall thickness to the true form surface () is defined, preferably, by a user. A normal of each pixel () is searched for grayscale values of voxels. Each pixel () is assigned a maximum of grayscale values of voxels within the defined wall thickness intersected by the corresponding normal. The resulting true form surface is undistorted mode of visualization revealing the arteries tree in its context running on the true surface drawn through the vessels. Mapping the assigned grayscale values onto the base sphere () visualizes arteries tree on a globe surface () which might be rotatably inspected as a globe. Mapping the assigned grayscale values into a flat surface visualizes arteries tree on a two-dimensional map.

Mapping the coronary arteries on a sphere

A diagnostic imaging apparatus generates a three-dimensional diagnostic image representation. The image representation is converted into an image representation of the patient's heart. A left lower posterior point of the heart image representation is selected as a first approximation of a left ventricle apex and a line of preselected orientation is drawn through the first approximation apex as a first approximation ventricle axis. The ventricle axis is redefined by generating short axis slices across the approximated ventricle axis, isolating a selected one of the ventricles, determining a centroid of each short axis slice of the selected ventricle, and fitting the axis to the apex approximation and the centroids. The apex location is redefined by looking for the lowermost short axis slice orthogonal to the redefined axis which intersects the ventricle volume and selecting the intersection point as the next approximation of the apex. The axis and apex redefining are iteratively repeated to determine the ventricle axis and apex automatically.

Growing community of inventors

Nordia, Israel

Guy A Lavi