THREE-DIMENSIONAL MESH FROM MAGNETIC RESONANCE IMAGING AND MAGNETIC RESONANCE IMAGING-FLUOROSCOPY MERGE
An imaging system can be configured to merge a three-dimensional (“3D”) mesh based on a magnetic resonance imaging (“MRI”) image of an anatomical feature to a fluoroscopy image of the anatomical feature. The imaging system can include a computer platform configured to perform operations. The operations can include obtaining the 3D mesh based on the MRI image. The operations can further include obtaining the fluoroscopy image. The operations can further include determining a registration seed. The operations can further include registering the 3D mesh based on the MRI image to the fluoroscopy image using the registration seed.
1 . An imaging system configured to merge a three-dimensional (“3D”) mesh based on a magnetic resonance imaging (“MRI”) image of an anatomical feature to a fluoroscopy image of the anatomical feature, the imaging system comprising a computer platform configured to perform operations comprising:
obtaining the 3D mesh based on the MRI image;
obtaining the fluoroscopy image;
determining a registration seed; and
registering the 3D mesh based on the MRI image to the fluoroscopy image using the registration seed.
2 . The imaging system of claim 1 , wherein obtaining the 3D mesh based on the MRI image comprises:
receiving an MRI image of the anatomical feature;
obtaining a nearest neighbor computerized topography (“CT”) template relative to the MRI image; and
generating the 3D mesh of the anatomical feature based on the MRI image and the nearest neighbor CT template.
3 . The imaging system of claim 1 , wherein the fluoroscopy image comprises a tracked biplanar fluoroscopy image.
4 . The imaging system of claim 1 , wherein determining the registration seed comprises determining the registration seed based on user input.
5 . The imaging system of claim 1 , wherein determining the registration seed comprises:
determining an estimated 3D model based on the fluoroscopy image; and
determining the registration seed by registering the 3D mesh based on the MRI with the estimated 3D model based on the fluoroscopy image.
6 . The imaging system of claim 1 , wherein determining the registration seed comprises estimating the registration seed using a neural network with the 3D mesh based on the MRI and the fluoroscopy image as input.
7 . The imaging system of claim 1 , wherein registering the 3D mesh based on the MRI image to the fluoroscopy image comprises:
determining a digitally reconstructed radiograph based on the 3D mesh based on the MRI image; and
registering the digitally reconstructed radiograph to the fluoroscopy image.
8 . The imaging system of claim 1 , the operations further comprising:
transmitting information associated with a registration between the 3D mesh and the fluoroscopy image to a surgical navigation system.
9 . The imaging system of claim 1 , wherein the anatomical feature comprises a spine.
10 . A method of operating an imaging system to merge a three-dimensional (“3D”) mesh based on a magnetic resonance imaging (“MRI”) image of an anatomical feature to a fluoroscopy image of the anatomical feature, the method comprising:
obtaining the 3D mesh based on the MRI image;
obtaining the fluoroscopy image;
determining a registration seed; and
registering the 3D mesh based on the MRI image to the fluoroscopy image using the registration seed.
11 . The method of claim 10 , wherein obtaining the 3D mesh based on the MRI image comprises:
receiving an MRI image of the anatomical feature;
obtaining a nearest neighbor computerized topography (“CT”) template relative to the MRI image; and
generating the 3D mesh of the anatomical feature based on the MRI image and the nearest neighbor CT template.
12 . The method of claim 10 , wherein the fluoroscopy image comprises a tracked biplanar fluoroscopy image.
13 . The method of claim 10 , wherein determining the registration seed comprises determining the registration seed based on user input.
14 . The method of claim 10 , wherein determining the registration seed comprises:
determining an estimated 3D model based on the fluoroscopy image; and
determining the registration seed by registering the 3D mesh based on the MRI with the estimated 3D model based on the fluoroscopy image.
15 . The method of claim 10 , wherein determining the registration seed comprises estimating the registration seed using a neural network with the 3D mesh based on the MRI and the fluoroscopy image as input.
16 . The method of claim 1 , wherein registering the 3D mesh based on the MRI image to the fluoroscopy image comprises:
determining a digitally reconstructed radiograph based on the 3D mesh based on the MRI image; and
registering the digitally reconstructed radiograph to the fluoroscopy image.
17 . The method of claim 1 , further comprising:
transmitting information associated with a registration between the 3D mesh and the fluoroscopy image to a surgical navigation system.
18 . The method of claim 1 , wherein the anatomical feature comprises a spine.
19 . A computer program product comprising:
a non-transitory computer readable medium storing instructions executable by a computer platform of an imaging system to merge a three-dimensional (“3D”) mesh based on a magnetic resonance imaging (“MRI”) image of an anatomical feature to a fluoroscopy image of the anatomical feature, the computer platform when executing the instructions causes the imaging system to perform operations comprising:
obtaining the 3D mesh based on the MRI image;
obtaining the fluoroscopy image;
determining a registration seed; and
registering the 3D mesh based on the MRI image to the fluoroscopy image using the registration seed.
20 . The computer program of claim 19 , wherein obtaining the 3D mesh based on the MRI image comprises:
receiving an MRI image of the anatomical feature;
obtaining a nearest neighbor computerized topography (“CT”) template relative to the MRI image; and
generating the 3D mesh of the anatomical feature based on the MRI image and the nearest neighbor CT template.