Intuitive display for rotator cuff tear diagnostics
Inventors
Geiger, Bernhard • Schwier, Michael • Grbic, Sasa • Raithel, Esther • Lin, Dana • Chabin, Guillaume
Assignees
Siemens Healthineers AG • New York University NYU
Publication Number
US-12076158-B2
Publication Date
2024-09-03
Expiration Date
2041-02-05
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
Systems and methods for an intuitive display of one or more anatomical objects are provided. One or more 3D medical images of one or more anatomical objects of a patient are received. Correspondences between the one or more 3D medical images and points on a 2D map representing the one or more anatomical objects are determined. The 2D map is updated with patient information extracted from the one or more 3D medical images. The updated 2D map with the determined correspondences is output.
Core Innovation
The invention provides systems and methods for an intuitive display of anatomical objects, with a particular focus on rotator cuff tear diagnostics. The process begins by receiving one or more 3D medical images of anatomical objects, such as the muscles and bones related to the rotator cuff. These images are processed to determine correspondences between points in the 3D medical images and points on a 2D map that represents the anatomical objects in a symbolic or unfolded state. The 2D map is then updated with patient-specific information, such as anatomical labels, muscle quality, and tear characteristics, which may be automatically determined from the 3D images using trained machine learning networks.
The problem addressed is the complexity and time consumption of current MRI image interpretation for rotator cuff tears, which typically involves manual scrolling through numerous images and measuring tears on different imaging planes. While artificial intelligence methods can automate some diagnostic tasks, their results often require time-consuming manual verification by radiologists or clinicians. The present invention aims to make this process more efficient and intuitive by integrating and visually representing findings on an interactive 2D map directly linked to the corresponding 3D image data.
Through methods such as user input, machine learning identification of landmarks, or atlas-based techniques, the system establishes accurate correspondences between the 3D medical data and the 2D representation. Users can interact with the 2D map, selecting points and receiving corresponding 2D slices or specific image regions, thus bridging the gap between machine learning findings and clinical verification. This approach is designed to reduce verification time and efficiently present relevant diagnostic information related to the rotator cuff.
Claims Coverage
There are three forms of independent claims in this patent: a method, an apparatus, and a non-transitory computer readable medium, each with inventive features related to intuitive anatomical display and rotator cuff diagnostics.
Method for generating intuitive 2D maps from 3D medical images with patient-specific information
This feature involves receiving 3D medical images of anatomical objects, determining correspondences between 2D slices and points on a 2D map by: - Annotating an atlas of the anatomical objects with features that correspond to those on the 2D map. - Segmenting anatomical structures in both the 3D images and annotated atlas using a trained machine learning-based segmentation network. - Registering the annotated atlas with the 3D images to establish correspondences based on segmented structures. The method includes updating the 2D map with patient information from the 3D images and presenting the updated map to users for interaction, such that selecting points on the map displays the corresponding 2D image slices.
Apparatus for mapping and interactive display of anatomical information
This feature describes an apparatus with means to: - Receive 3D medical images of anatomical objects. - Determine correspondences between 2D slices and points on a 2D map by annotating an atlas and registering it to 3D images. - Update the 2D map with extracted patient data. - Present the map for user interaction via a display, where selecting points on the map displays corresponding image slices.
Non-transitory computer readable medium for executing the anatomical mapping and display method
This feature covers a non-transitory computer readable medium storing instructions that, when executed by a processor, perform: - Receiving 3D medical images. - Determining correspondences between 2D slices and map points via atlas annotation and registration. - Updating the 2D map with patient-specific information. - Presenting the map to users for interactive selection and display of corresponding image slices.
In summary, the patent claims cover methods, apparatuses, and computer-readable media for transforming 3D medical images into interactive 2D maps with established correspondences and patient-specific information, specifically enhancing the workflow for rotator cuff diagnostics.
Stated Advantages
Significantly reduces the time for a user to verify automatically determined findings of machine learning based networks.
Efficiently presents findings and provides information of interest that may reduce errors in diagnostics.
Documented Applications
Assessment and verification of rotator cuff tears using intuitive 2D and 3D anatomical displays updated with patient information.
Interested in licensing this patent?