Systems and methods for automated image analysis
An image analysis system including at least one processor and at least one memory is provided. The image analysis system is configured to receive image data associated with a brain of a patient, the image data including a first three-dimensional (3D) diffusion weighted imaging (DWI) image acquired using a magnetic resonance imaging (MRI) system and a second 3D DWI image, concurrently provide the first 3D DWI image to a first channel of a trained model and the second 3D DWI image to a second channel of the trained model, receive an indicator associated with the first 3D DWI image and the second 3D DWI image from the model, generate a report based on the indicator, and cause the report to be output to at least one of a memory or a display.
1. A image analysis system comprising at least one processor and at least one memory, the image analysis system configured to:
receive image data associated with a brain of a patient, wherein the image data comprises a first three-dimensional (3D) diffusion weighted imaging (DWI) image acquired using a magnetic resonance imaging (MRI) system and a second 3D DWI image;
concurrently provide the first 3D DWI image to a first channel of a trained model and the second 3D DWI image to a second channel of the trained model;
receive an indicator associated with the first 3D DWI image and the second 3D DWI image from the model;
generate a report based on the indicator; and
cause the report to be output to at least one of a memory or a display.
2. The system of claim 1 , wherein the first 3D DWI image is generated using a first b value, and the second 3D DWI image is generated using a second b value.
3. The system of claim 1 , wherein the image data comprises a third 3D DWI image and a fourth 3D DWI image, at least a portion of the third 3D DWI image and the fourth 3D DWI image being associated with a different spatial location in the brain than at least a portion of the first 3D DWI image and the second 3D DWI, the system further configured to:
provide the third 3D DWI image and the fourth 3D DWI image to the trained model;
receive an indicator associated with the third 3D DWI image and the fourth 3D DWI image from the trained model; and
generate the report further based on the indicator associated with the third 3D DWI image and the fourth 3D DWI image.
4. The system of claim 3 , wherein a subset of pixels included in the third 3D DWI image and the fourth 3D DWI image and a subset of pixels included in the first 3D DWI image and the second 3D DWI image are associated with a common spatial area, and wherein the system is further configured to determine the presence of a lesion in the common spatial area based on the indicator associated with the third 3D DWI image and the fourth 3D DWI image and the indicator associated with the first 3D DWI image and the second 3D DWI image.
5. The system of claim 1 , wherein the first 3D DWI image comprises a portion of a larger 3D DWI image of the brain of the patient.
6. The system of claim 1 , wherein the first 3D DWI image is generated using a first set of imaging parameters, and the second 3D DWI image is generated using a second set of imaging parameters.
7. The system of claim 1 , wherein the indicator includes a probability map indicative of locations in the brain of the patient where lesions have occurred.
8. The system of claim 1 , wherein the model comprises a convolutional neural network comprising a number of inception modules.
9. The system of claim 8 , wherein the model comprises a second convolutional neural network comprising a number of inception modules.
10. The system of claim 1 , wherein the report comprises at least one of a segmentation map indicating locations of lesions in the brain of the patient or a volume of lesions in the brain of the patient.
11. The system of claim 1 , wherein the report comprises a predicted occurrence of an ischemic stroke.
12. An image analysis method comprising:
receiving image data associated with a brain of a patient, the image data comprising a first three-dimensional (3D) diffusion weighted image (DWI) image acquired using a magnetic resonance imaging (MRI) system and a second 3D DWI image;
concurrently providing the first 3D DWI image to a first channel of a trained model and the second 3D DWI image to a second channel of the trained model;
receiving an indicator associated with the first 3D DWI image and the second 3D DWI image from the model;
generating a report based on the indicator; and
causing the report to be output to at least one of a memory or a display.
13. The method of claim 12 , wherein the first 3D DWI image is generated using a first b value, and the second 3D DWI image is generated using a second b value.
14. The method of claim 12 , wherein the image data comprises a third 3D DWI image and a fourth 3D DWI image, at least a portion of the third 3D DWI image and the fourth 3D DWI image being associated with a different spatial location in the brain than at least a portion of the first 3D DWI image and the second 3D DWI, the method further comprising:
providing the third 3D DWI image and the fourth 3D DWI image to the trained model;
receiving an indicator associated with the third 3D DWI image and the fourth 3D DWI image from the trained model; and
generating the report further based on the indicator associated with the third 3D DWI image and the fourth 3D DWI image.
15. The method of claim 14 , wherein a subset of pixels included in the third 3D DWI image and the fourth 3D DWI image and a subset of pixels included in the first 3D DWI image and the second 3D DWI image are associated with a common spatial area, and wherein the method further comprises determining the presence of a lesion in the common spatial area based on the indicator associated with the third 3D DWI image and the fourth 3D DWI image and the indicator associated with the first 3D DWI image and the second 3D DWI image.
16. The method of claim 12 , wherein the first 3D DWI image comprises a portion of a larger 3D DWI image of the brain of the patient.
17. The method of claim 12 , wherein the first 3D DWI image is generated using a first set of imaging parameters, and the second 3D DWI image is generated using a second set of imaging parameters.
18. The method of claim 12 , wherein the indicator includes a probability map indicative of locations in the brain of the patient where lesions have occurred.
19. The method of claim 12 , wherein the model comprises a convolutional neural network comprising a number of inception modules.
20. The method of claim 19 , wherein the model comprises a second convolutional neural network comprising a number of inception modules.
21. The method of claim 12 , wherein the report comprises at least one of a segmentation map indicating locations of lesions in the brain of the patient or a volume of lesions in the brain of the patient.
22. The method of claim 12 , wherein the report comprises a predicted occurrence of an ischemic stroke.
23. An ischemic stroke analysis system comprising at least one processor and at least one memory, the ischemic stroke analysis system configured to:
receive image data associated with a brain of a patient, the image data comprising a first three-dimensional (3D) diffusion weighted imaging (DWI) image and a second 3D DWI image acquired using a magnetic resonance imaging (MRI) system;
provide the image data to a trained model;
receive an indicator associated with the first 3D DWI image and the second 3D DWI image from the model;
generate a report comprising a segmentation map of lesions in the brain of the patient based on the indicator; and
cause the report to be output to at least one of a memory or a display.