Systems and methods for automated and interactive analysis of bone scan images for detection of metastases
Inventors
Sjöstrand, Karl Vilhelm • Richter, Jens Filip Andreas • Edenbrandt, Lars
Assignees
Exini Diagnostics AB • Progenies Pharmaceuticals Inc • Progenics Pharmaceuticals Inc
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11534125-B2
Publication Date
2022-12-27
Expiration Date
Abstract
Presented herein are systems and methods that provide for improved computer aided display and analysis of nuclear medicine images. In particular, in certain embodiments, the systems and methods described herein provide improvements to several image processing steps used for automated analysis of bone scan images for assessing cancer status of a patient. For example, improved approaches for image segmentation, hotspot detection, automated classification of hotspots as representing metastases, and computation of risk indices such as bone scan index (BSI) values are provided.
Core Innovation
The invention relates to a method for lesion marking and quantitative analysis of nuclear medicine images of a human subject that automatically processes a bone scan image set obtained following administration of an agent. The method automatically segments each image to identify skeletal regions of interest corresponding to anatomical regions of a skeleton, automatically detects an initial set of one or more hotspots as areas of elevated intensity in the annotated set of images, and associates the hotspots with extracted hotspot features to calculate a metastasis likelihood value for each hotspot.
In related embodiments, hotspot detection uses intensities of pixels together with one or more region-dependent threshold values, including values associated with the femur region and/or the humerus region that provide enhanced hotspot detection sensitivity in the femur region and/or the humerus region to compensate for reduced uptake of the agent therein. Skeletal region identification is further constrained by defining a femur portion encompassing at least three quarters of the femur along its length and/or a humerus portion encompassing at least three quarters of the humerus along its length, and atlas images are registered to apply identifications of skeletal regions of interest to the images.
The invention further refines quantitative risk analysis by selecting subsets of hotspots based at least in part on metastasis likelihood values and one or more global hotspot features determined using a plurality of hotspots in an initial set, and rendering a graphical representation of at least a portion of the initial set or selected subset of hotspots within a graphical user interface (GUI). For risk index calculation, the method computes skeletal involvement factors based on ratios of hotspot size to assigned skeletal region size, adjusts the skeletal involvement factors using one or more region-dependent correction factors, and sums adjusted skeletal involvement factors to determine one or more risk index values.
Claims Coverage
The provided excerpts list multiple independent claims that cover a complete lesion-marking and quantitative analysis workflow with anatomical constrained segmentation, hotspot detection with region-dependent thresholds, per-hotspot metastasis likelihood computation, GUI rendering for hotspot visualization and selection, and risk index generation via subset selection and skeletal involvement factor aggregation. The inventive features include atlas-based registration, femur and humerus coverage constraints, global threshold scaling, subset selection using global hotspot features, and risk index calculation with region-dependent correction factors.
Atlas-based registration for skeletal region-of-interest labeling
Comparing each member of the bone scan image set with a corresponding atlas image of an atlas image set, identifying the one or more skeletal regions of interest in the atlas image, and registering the corresponding atlas image with the image of the bone scan image set using the identified knee region and/or the identified elbow region as landmark(s), such that the identifications of the one or more skeletal regions of interest of the atlas image are applied to the image of the bone scan image set.
Femur and humerus coverage constraints for skeletal regions of interest
Automatically segmenting each image to identify skeletal regions of interest wherein the one or more skeletal regions of interest comprise at least one of a femur region corresponding to a portion of a femur encompassing at least three quarters of the femur along its length and/or a humerus region corresponding to a portion of a humerus encompassing at least three quarters of the humerus along its length.
Region-dependent hotspot detection thresholds with femur/humerus sensitivity compensation
Automatically detecting an initial set of one or more hotspots by identifying the one or more hotspots using intensities of pixels in the annotated set of images and using one or more region-dependent threshold values, wherein the one or more region dependent threshold values include one or more values associated with the femur region and/or the humerus region that provide enhanced hotspot detection sensitivity in the femur region and/or the humerus region to compensate for reduced uptake of the agent therein.
Metastasis likelihood from extracted hotspot features
For each hotspot in the initial set of hotspots, extracting a set of hotspot features and calculating a metastasis likelihood value corresponding to a likelihood of the hotspot representing a metastasis based on the set of hotspot features.
Hotspot visualization in a graphical user interface (GUI)
Causing rendering of a graphical representation of at least a portion of the initial set of hotspots for display within a graphical user interface (GUI).
Global threshold scaling factor to adjust preliminary threshold values
Using intensities of pixels in the annotated set of images and a plurality of preliminary threshold values to detect a set of potential hotspots, computing a global threshold scaling factor using the set of potential hotspots, adjusting the plurality of preliminary threshold values using the global threshold scaling factor to obtain a plurality of adjusted threshold values, and using intensities of pixels and the plurality of adjusted threshold values to identify the initial set of hotspots.
Subset selection using metastasis likelihood values and global hotspot features
Selecting a first subset of the initial set of hotspots wherein selection of a particular hotspot for inclusion is based at least in part on the metastasis likelihood value and one or more global hotspot features determined using a plurality of hotspots in the initial set of hotspots.
Risk index calculation via skeletal involvement factors with region-dependent correction factors
Calculating one or more risk index values using at least a portion of the first subset of hotspots by computing, for each particular hotspot of the portion of first subset, a skeletal involvement factor based on a ratio of a size of the particular hotspot to a size of a particular skeletal region to which the particular hotspot is assigned, adjusting the skeletal involvement factors using one or more region-dependent correction factors to obtain one or more adjusted skeletal involvement factors, and summing the adjusted skeletal involvement factors to determine the one or more risk index values.
Across the independent claims, the coverage centers on automatic skeletal region segmentation, region-dependent hotspot detection with femur and humerus sensitivity compensation and atlas-based registration, per-hotspot metastasis likelihood calculation from hotspot features, GUI-based rendering and subset selection, and risk index computation using skeletal involvement factors aggregated with region-dependent correction factors.
Stated Advantages
Enhanced hotspot detection sensitivity in the femur region and/or the humerus region to compensate for reduced uptake of the agent therein.
Improved skeletal region-of-interest labeling by applying atlas identifications through registering atlas images using knee and/or elbow landmark regions.
Potentially improved initial hotspot identification by adjusting preliminary threshold values using a global threshold scaling factor computed from potential hotspots.
Risk index computation that limits errors from assigning and aggregating hotspot-derived skeletal involvement factors by using region-dependent correction factors.
Documented Applications
Automated and interactive lesion marking and quantitative analysis of nuclear medicine bone scan images of a human subject, including rendering hotspots in a GUI and calculating quantitative risk indices based on selected hotspots.
Interested in licensing this patent?