Systems and methods for rapid neural network-based image segmentation and radiopharmaceutical uptake determination
Inventors
Sjöstrand, Karl Vilhelm • Richter, Jens Filip Andreas • Johnsson, Kerstin Elsa Maria • Gjertsson, Erik Konrad
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-10973486-B2
Publication Date
2021-04-13
Expiration Date
Abstract
Presented herein are systems and methods that provide for automated analysis of three-dimensional (3D) medical images of a subject in order to automatically identify specific 3D volumes within the 3D images that correspond to specific organs and/or tissue. In certain embodiments, the accurate identification of one or more such volumes can be used to determine quantitative metrics that measure uptake of radiopharmaceuticals in particular organs and/or tissue regions. These uptake metrics can be used to assess disease state in a subject, determine a prognosis for a subject, and/or determine efficacy of a treatment modality.
Core Innovation
The invention provides an automatic method for processing 3D images to identify 3D volumes corresponding to a prostate of a subject and determining one or more uptake metrics indicative of radiopharmaceutical uptake therein. The method receives a 3D anatomical image obtained using an anatomical imaging modality and a 3D functional image obtained using a functional imaging modality, where the functional image comprises voxels with intensity values representing detected radiation emitted from physical volumes in the subject pelvic region.
Within the method, a first module determines an initial volume of interest within the 3D anatomical image corresponding to tissue within the pelvic region while excluding tissue outside the pelvic region. A second module identifies a prostate volume within the initial volume of interest corresponding to the prostate of the subject, and the method then determines the one or more uptake metrics using the 3D functional image and the prostate volume identified within the initial volume of interest.
The invention further includes determining uptake metrics with respect to a reference volume corresponding to a reference tissue region, and determining diagnostic or prognostic values by comparing an uptake metric to one or more threshold value(s). Additional refinements include identifying bladder volume and correcting for cross-talk from the bladder using intensities of voxels corresponding to the bladder volume, and excluding regions corresponding to a dilated bladder volume when determining the uptake metrics.
The disclosure also describes additional tissue-volume selection and auxiliary merging, interactive rendering of selectable and superimposable layers of the 3D anatomical image and/or the 3D functional image in a graphical user interface, and a two-CNN module structure in which the second CNN module comprises a greater number of convolutional filters than the first CNN module. The claims further include determining diagnostic or prognostic values that estimate a risk for clinically significant prostate cancer in the subject.
Claims Coverage
The provided claim set includes multiple independent claims spanning methods and systems, with a shared core workflow. Across the independent claims, there are 8 main inventive features covering automatic 3D anatomical-to-functional processing with pelvic-region initial VOI and prostate volume identification, determination of radiopharmaceutical uptake metrics using the prostate volume and optionally a reference volume, bladder-based cross-talk correction or dilated-bladder exclusion, auxiliary tissue volume handling, diagnostic/prognostic output via threshold comparison, risk estimation for clinically significant prostate cancer, interactive GUI layer rendering, and a two-CNN filter relationship.
Pelvic-region initial volume of interest and prostate identification from 3D anatomical image
A method/system in which a first module determines an initial volume of interest within the 3D anatomical image corresponding to tissue within the pelvic region of the subject and excluding tissue outside the pelvic region, and a second module identifies a prostate volume within the initial volume of interest corresponding to the prostate of the subject.
Radiopharmaceutical uptake metrics from 3D functional image using prostate volume
A method/system in which the one or more uptake metrics are determined using the 3D functional image and the prostate volume identified within the initial volume of interest of the 3D anatomical image, where the 3D functional image comprises voxels with intensity values representing detected radiation.
Reference volume-based uptake metric determination
A method/system wherein the method further comprises identifying a reference volume within the 3D anatomical image corresponding to a reference tissue region, and determining at least one of the one or more uptake metrics using the 3D functional image and the reference volume identified within the 3D anatomical image.
Bladder cross-talk correction using bladder volume intensities
A method/system wherein the method further comprises identifying a bladder volume within the 3D anatomical image corresponding to a bladder of the subject, and correcting for cross-talk from the bladder using intensities of voxels of the 3D functional image corresponding to the identified bladder volume within the 3D anatomical image.
Dilated bladder volume exclusion using morphological dilation
A method/system wherein the method further comprises identifying a bladder volume within the 3D anatomical image corresponding to a bladder of the subject, determining a dilated bladder volume by applying a morphological dilation operation to the identified bladder volume, and determining the one or more uptake metrics using intensity values of voxels that correspond to the prostate volume identified within the VOI but do not correspond to regions within the dilated bladder volume.
Auxiliary tissue volume identification and merging with base tissue volume
A method/system in which the second module identifies one or more additional tissue volumes within the 3D anatomical image, each corresponding to specific tissue regions selected from the group consisting of a pelvic bone of the subject, a bladder of the subject, a rectum of the subject, and a gluteal muscle of the subject, and further identifies auxiliary tissue volumes using one or more auxiliary modules for each auxiliary tissue volume corresponding to a base tissue volume identified by the second module and merges each auxiliary tissue volume with the corresponding base tissue volume.
Diagnostic or prognostic value determination using uptake metric threshold comparisons
A method/system wherein determining, based on at least a portion of the one or more uptake metrics, one or more diagnostic or prognostic values for the subject is performed by comparing an uptake metric to one or more threshold value(s), and at least one value estimates a risk for clinically significant prostate cancer in the subject.
Interactive GUI with selectable and superimposable overlay layers
A method/system that includes causing display of an interactive graphical user interface for presentation to the user of a visual representation of the 3D anatomical image and/or the 3D functional image, and causing graphical rendering within the GUI of the 3D anatomical image and/or the 3D functional image as selectable and superimposable layers, such that either can be selected for display and rendered separately or both selected for display and rendered together by overlaying the 3D anatomical image with the 3D functional image.
Two CNN modules with prostate module having greater convolutional filters
A method/system wherein the first module is a first CNN module and the second module is a second CNN module, and the second CNN module comprises a greater number of convolutional filters than the first CNN module.
Across the independent claims, the invention consistently combines automatic pelvic-region initial VOI and prostate volume identification from a 3D anatomical image with radiopharmaceutical uptake metric determination from a co-existing 3D functional image. The claims further specify optional reference-volume usage, optional bladder-based cross-talk correction via bladder volume intensities or dilated-bladder exclusion, optional auxiliary tissue region identification with merging, diagnostic/prognostic value determination through uptake-metric threshold comparisons, risk estimation for clinically significant prostate cancer, interactive GUI layer rendering, and a two-CNN module architecture difference.
Stated Advantages
Documented Applications
No documented applications found
Interested in licensing this patent?