System, method, and computer program product for determining a needle injection site
Inventors
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Assignees
Member
Carnegie Mellon UniversityCarnegie Mellon University is a global research institution based in Pittsburgh, Pennsylvania, recognized for interdisciplinary education, research, and innovation in science, engineering, arts, technology, and social sciences. The university leads advancements in artificial intelligence, robotics, digital health, and performing arts. Located in a technology-driven and culturally rich city, CMU powers real-world impact through research centers, industry engagement, workforce training, and initiatives that shape regional and global communities.
Carnegie Mellon University is a global research institution based in Pittsburgh, Pennsylvania, recognized for interdisciplinary education, research, and innovation in science, engineering, arts, technology, and social sciences. The university leads advancements in artificial intelligence, robotics, digital health, and performing arts. Located in a technology-driven and culturally rich city, CMU powers real-world impact through research centers, industry engagement, workforce training, and initiatives that shape regional and global communities.
Publication Number
US-12579740-B2
Publication Date
2026-03-17
Expiration Date
Abstract
Provided is a system, method, and computer program product for determining a needle injection site. The method includes segmenting, with at least one computing device, an image of a sequence of images into at least one object based on a machine-learning model configured to estimate its uncertainty for each segmentation, generating, with the at least one computing device, a 3D model of the at least one object, and determining, with the at least one computing device, an insertion location of the at least one object based at least partially on an output of the machine-learning model.
Core Innovation
Not explicitly described in patent.
Not explicitly described in patent.
Claims Coverage
The provided independent claims are directed to a method, a system, and a computer program product, with a common core workflow of uncertainty-estimating segmentation of an image sequence, 3D model generation by filtering uncertain segmentation results using segmentation maps and uncertainty maps, and insertion-location determination based on the machine-learning model output. The inventive features focus on uncertainty estimation per segmentation, uncertainty-aware filtering for 3D reconstruction, and insertion-location determination using specific anatomical detections and scoring along an object.
Uncertainty-estimating segmentation of an image sequence
Segment an image of a sequence of images into at least one object based on a machine-learning model configured to estimate its uncertainty for each segmentation.
3D model generation by filtering uncertain segmentation results using segmentation and uncertainty maps
Generate a 3D model of the at least one object by filtering uncertain segmentation results based on segmentation maps and uncertainty maps output by the machine-learning model, including calculating an average uncertainty value for each segmented object and filtering by class based on an uncertainty threshold determined based on a predictive mean and a predictive variance output by the machine-learning model.
Insertion location determination based on machine-learning model output
Determine an insertion location of the at least one object based at least partially on an output of the machine-learning model.
Insertion location determination using bifurcation point and caudal end of a segmented ligament with total site scores
Determine an insertion location on the object based on a plurality of total site scores generated for different points along the object, including detecting a bifurcation point in the object in the 3D model, detecting a caudal end of a segmented ligament, and determining an insertion region between the bifurcation point and the caudal end of the segmented ligament.
Overall claim coverage ties uncertainty-estimating segmentation to uncertainty-aware 3D model generation via segmentation maps, uncertainty maps, average uncertainty, and class-based thresholds from predictive mean and predictive variance, and uses the machine-learning model output for insertion-location determination, further narrowed in the system claim to bifurcation-point and caudal-end detection of a segmented ligament with insertion-region definition and insertion selection based on total site scores along the object.
Stated Advantages
Documented Applications
No documented applications found
Interested in licensing this patent?