US-12620085-B2 - Imaging system and method to estimate contour of a scanned object

US12620085B2US 12620085 B2US12620085 B2US 12620085B2US-12620085-B2

Abstract

Described herein are an imaging system and method. Specifically, the imaging system includes a positioning image acquisition unit configured to acquire positioning images of a scanned object from a plurality of angles, a contour estimation unit configured to estimate a contour of the object in each positioning image in a scanning direction when truncation is present in at least one positioning image, and a display field of view determination unit configured to select a maximum value of an estimated contour as a display field of view of the image. A contour of a scanned object in each positioning image in a scanning direction is estimated when truncation is present in at least one positioning image, thereby determining a suitable display field of view. An appropriate display field of view can be set, so that a reconstructed image can cover the entire contour of the object and have a higher resolution.

Inventors

Ximiao CAO
Bingjie ZHAO
Xueli Wang
Jiang Hsieh

Assignees

GE Precision Healthcare LLC

Dates

Publication Date: 20260505
Application Date: 20220907
Priority Date: 20210908

Claims (9)

1 . An imaging system, comprising: a positioning image acquisition unit, configured to acquire positioning images of a scanned object from a plurality of angles; a truncation determination unit configured to detect truncation in a positioning image by converting the positioning image to a binary image using a threshold and determining that edge pixels consecutively above the threshold indicate truncation; a contour estimation unit, configured to estimate a contour of the scanned object in each positioning image in a scanning direction when truncation is present in at least one of the positioning images, wherein the contour estimation unit: acquires, from each positioning image, a contour of a non-truncated portion at each position in the scanning direction; estimates the contour points at all positions in the scanning direction in the positioning images to estimate a complete contour of the scanned object; and performs polynomial fitting beyond the truncated edge and locates outer contour points as the points where the fitted curve crosses a threshold line; and a display field of view determination unit, configured to select a maximum value of the estimated contour across the plurality of angles and apply the display field of view as a reconstruction field of view in a diagnostic scan.
2 . The imaging system according to claim 1 , wherein the contour estimation unit inputs the acquired positioning images into a trained neural network, thereby estimating the contour of the scanned object in each positioning image.
3 . The imaging system according to claim 1 , wherein the contour estimation unit comprises a preliminary reconstruction unit and a scanned object contour determination unit, the preliminary reconstruction unit reconstructs a plurality of slice images of the scanned object based on an extended display field of view, and the scanned object contour determination unit searches for a contour of the scanned object in each of the slice images.
4 . An imaging method, comprising: acquiring, via a positioning image acquisition unit, positioning images of a scanned object from a plurality of angles; detecting, via a truncation unit, truncation in a positioning image by converting the positioning image to a binary image using a threshold and determining that edge pixels consecutively above the threshold indicate truncation; estimating, via a contour estimation unit, a contour of the scanned object when truncation is present in at least one of the positioning images, wherein estimating the contour includes: acquiring, from each positioning image, a contour of a non-truncated portion at each position in the scanning direction; estimating the contour points at all positions in the scanning direction in the positioning images to estimate a complete contour of the scanned object; and performing polynomial fitting beyond the truncated edge and locates outer contour points as the points where the fitted curve crosses a threshold line; and selecting, via a display field of view determination unit, a maximum value of the estimated contour across the plurality of angles and apply the display field of view as a reconstruction field of view in a diagnostic scan.
5 . The imaging method according to claim 4 , wherein the estimating the contour of the scanned object comprises assuming the scanned object corresponding to a truncated portion in the acquired positioning image as a homogeneous body mold, thereby estimating the contour of the scanned object in each positioning image.
6 . The imaging method according to claim 4 , wherein the estimating the contour of the scanned object comprises inputting the acquired positioning images into a trained neural network, thereby estimating the contour of the scanned object in each positioning image.
7 . The imaging method according to claim 4 , wherein the estimating the contour of the scanned object comprises performing a preliminary reconstruction of a slice image at a specific position in a scanning direction based on an extended display field of view, and searching for a contour of the scanned object in each of the slice images.
8 . A system, comprising a processor configured to perform the imaging method according to claim 4 .
9 . A computer program product, comprising a computer program, wherein the computer program, when executed by a processor, implements the steps of the method according to claim 4 .

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to Chinese Patent Application No. 202111049401.2, filed on Sep. 8, 2021, the disclosure of which is incorporated herein by reference in its entirety. TECHNICAL FIELD The present invention relates to medical imaging, and in particular to an imaging system and a method for determining a display field of view. BACKGROUND A medical imaging system is usually used to acquire slice images of a scanned object (i.e., diagnostic images) to provide information needed for diagnosis or treatment. In the case of computed tomography (CT), for example, X-rays are emitted by a tube and then pass through the scanned object. The attenuated rays are detected by an X-ray detector and the projection signal collected and converted by a data acquisition system (DAS) is reconstructed to acquire the slice images. When performing a CT scan, it is usually necessary to configure a scan field of view (SFOV) and a display field of view (DFOV). The SFOV is a scanning range set by the CT apparatus itself (e.g., 50 cm). For different examination regions, an appropriate size of SFOV may be selected to ensure that the SFOV covers the examination region and that the resolution of the displayed image can be improved. The DFOV is usually within the range of the SFOV, and is a display range of a reconstructed image through configuration prior to the examination. When scanning, the scanned object is usually placed within the SFOV. For example, by means of aligning the center of the scanned object with the center of the SFOV, the SFOV covers the scanned object, and thus the reconstructed slice image reflects the complete information of the scanned object. When the scanned object is obese, or the center of the scanned object is shifted, or the scanned object is in a special pose, a portion of the scanned object will be outside of the SFOV and the reconstructed image will be truncated in this case. In this case, image reconstruction with a DFOV that is larger than the SFOV is usually required to provide complete information of the scanned object. However, an excessively large DFOV may cause the resolution of the image to decrease. Therefore, a novel imaging system and method are needed that can accurately set an appropriate DFOV, so that a reconstructed image can cover the entire contour of a scanned object and have a higher resolution. SUMMARY With respect to the above technical problems, provided in embodiments of the present invention is an imaging system, comprising: a positioning image acquisition unit, configured to acquire positioning images of a scanned object from a plurality of angles; a contour estimation unit, configured to estimate a contour of the scanned object in each positioning image in a scanning direction when truncation is present in at least one positioning image; and a display field of view determination unit, configured to select a maximum value of an estimated contour as a display field of view of the image. In an aspect of the present invention, the imaging system further comprises a truncation determination unit, configured to determine whether truncation is present in the positioning images. In an aspect of the present invention, the contour estimation unit acquires, from each positioning image, a contour of a non-truncated portion at each position in the scanning direction and performs polynomial fitting to acquire the contour of the scanned object in each positioning image. In an aspect of the present invention, the contour estimation unit assumes the scanned object corresponding to a truncated portion in the acquired positioning image as a homogeneous body mold, thereby estimating the contour of the scanned object in each positioning image. In an aspect of the present invention, the contour estimation unit inputs the acquired positioning images into a trained neural network, thereby estimating the contour of the scanned object in each positioning image. In an aspect of the present invention, the contour estimation unit comprises a preliminary reconstruction unit and a scanned object contour determination unit, the preliminary reconstruction unit roughly reconstructs a slice image of the scanned object based on an extended display field of view, and the scanned object contour determination unit searches for a contour of the scanned object in each of the slice images. Further provided in embodiments of the present invention is an imaging method, comprising acquiring positioning images of a scanned object from a plurality of angles, estimating a contour of the scanned object when truncation is present in at least one positioning image, and selecting a maximum value of the estimated contour as a display field of view of the image. In an aspect of the present invention, the imaging method further comprising determining whether truncation is present in the positioning image after acquiring the positioning images of the scanned object. In an aspect of the