Generating reformatted views of a three-dimensional anatomy scan using deep-learning estimated scan prescription masks
Techniques are described for generating reformatted views of a three-dimensional (3D) anatomy scan using deep-learning estimated scan prescription masks. According to an embodiment, a system is provided that comprises a memory that stores computer executable components, and a processor that executes the computer executable components stored in the memory. The computer executable components comprise a mask generation component that employs a pre-trained neural network model to generate masks for different anatomical landmarks depicted in one or more calibration images captured of an anatomical region of a patient. The computer executable components further comprise a reformatting component that reformats 3D image data captured of the anatomical region of the patient using the masks to generate different representations of the 3D image data that correspond to the different anatomical landmarks.
1. A system, comprising:
a memory that stores computer executable components; and
a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise:
a mask generation component that employs a pre-trained neural network model to generate masks for different anatomical landmarks depicted in one or more calibration images captured of an anatomical region of a patient; and
a reformatting component that reformats three-dimensional image data captured of the anatomical region of the patient using the masks to generate different representations of the three-dimensional image data that correspond to the different anatomical landmarks, wherein the different representations comprise at least one of synthetic two-dimensional images generated from the three-dimensional image data or synthetic three-dimensional images generated from the three-dimensional image data.
2. The system of claim 1 , wherein the masks respectively provide an anatomical frame of reference for the different anatomical landmarks as depicted in the three-dimensional image data.
3. The system of claim 1 , wherein the masks include an anatomy coverage mask defining a three-dimensional region of interest for the anatomical region and wherein the three-dimensional image data was captured within the three-dimensional region of interest using the anatomy coverage mask.
4. The system of claim 1 , wherein the three-dimensional image data was captured along a first scan plane, wherein different anatomical landmarks comprise one or more second scan planes that are different than the first scan plan, and wherein the different representations provide perspectives of the three-dimensional image data generated from the one or more second scan planes.
5. The system of claim 1 , wherein the different anatomical landmarks comprise one or more anatomical structures and wherein the different representations provide perspectives of the three-dimensional image data relative to the one or more anatomical structures.
6. The system of claim 1 , wherein the different anatomical landmarks comprise a non-planar anatomical structure.
7. The system of claim 6 , wherein the masks comprise a non-planar mask corresponding to the non-planar anatomical structure and wherein the reformatting component generates a planar representation of the non-planar anatomical structure using the three-dimensional image data and the non-planar mask.
8. The system of claim 1 , wherein the one or more calibration images have a lower resolution relative to the three-dimensional image data.
9. The system of claim 1 , wherein the one or more calibration images comprise one or more localizer images or coil sensitivity images.
10. The system of claim 1 , further comprising:
a rendering component that renders the different representations via a display.
11. The system of claim 1 , wherein the computer executable components further compromise:
a quantification component that quantifies one or more geometric or visual properties associated with one or more of the different anatomical landmarks using the masks and the three-dimensional image data; and
an anomaly detection component that evaluates presence of one or more anomalies associated with the one or more of the different anatomical landmarks based on the one or more geometric and visual properties using machine learning and artificial intelligence techniques.
12. A method comprising:
employing, by a system operatively coupled to a processor, a pre-trained neural network model to generate masks for different anatomical landmarks depicted in one or more calibration images captured of an anatomical region of a patient; and
reformatting, by the system, three-dimensional image data captured of the anatomical region of the patient using the masks to generate different representations of the three-dimensional image data that correspond to the different anatomical landmarks, wherein the different representations comprise at least one of synthetic two-dimensional images generated from the three-dimensional image data or synthetic three-dimensional images generated from the three-dimensional image data.
13. The method of claim 12 , wherein the masks respectively provide an anatomical frame of reference for the different anatomical landmarks as depicted in the three-dimensional image data.
14. The method of claim 12 , wherein the three-dimensional image data was captured along a first scan plane, wherein different anatomical landmarks comprise one or more second scan planes that are different than the first scan plan, and wherein the different representations provide perspectives of the three-dimensional image data generated from the one or more second scan planes.
15. The method of claim 12 , wherein the different anatomical landmarks comprise one or more anatomical structures and wherein the different representations provide perspectives of the three-dimensional image data relative to the one or more anatomical structures.
16. The method of claim 12 , wherein the different anatomical landmarks comprise a non-planar anatomical structure.
17. The method of claim 12 , wherein the masks comprise a non-planar mask corresponding to the non-planar anatomical structure and wherein the reformatting component generates a planar representation of the non-planar anatomical structure using the three-dimensional image data and the non-planar mask.
18. A machine-readable storage medium, comprising executable instructions that, when executed by a processor, facilitate performance of operations, comprising:
employing a pre-trained neural network model to generate masks for different anatomical landmarks depicted in one or more calibration images captured of an anatomical region of a patient; and
reformatting three-dimensional image data captured of the anatomical region of the patient using the masks to generate different representations of the three-dimensional image data that correspond to the different anatomical landmarks, wherein the different representations comprise at least one of synthetic two-dimensional images generated from the three-dimensional image data or synthetic three-dimensional images generated from the three-dimensional image data.
19. The machine-readable storage medium of claim 18 , wherein the three-dimensional image data was captured along a first scan plane, wherein different anatomical landmarks comprise one or more second scan planes that are different than the first scan plan, and wherein the different representations provide perspectives of the three-dimensional image data generated from the one or more second scan planes.
20. The machine-readable storage medium of claim 18 , wherein the masks include an anatomy coverage mask defining a three-dimensional region of interest for the anatomical region and wherein the three-dimensional image data was captured within the three-dimensional region of interest using the anatomy coverage mask.