Automated organ risk segmentation machine learning methods and systems

1. An image data processing system comprising a processor to:

detect an anatomy in an image and to remove items not included in the anatomy from the image;

generate a bounding box around a region of interest in the anatomy;

classify image data within the bounding box at the voxel level to identify an object in the region of interest; and

output an indication of the object identified in the region of interest segmented in the image, wherein the indication is generated based on both a first fully connected neural network and a gradient boosting machine, wherein the first fully connected neural network and the gradient boosting machine are built using a plurality of image features created using image intensity values and at least one additional neural network, wherein the indication further comprises radiation planning information comprising radiation location information and radiation dosage information for the object, and wherein the object is an organ identified from the image.

2. The system of claim 1 , wherein the processor is configured to combine a plurality of models to classify the object in the region of interest, and wherein the processor is configured to employ at least two models to classify the image data within the bounding box, and combine the at least two models to classify the object and determine at least one feature related to the object.

3. The system of claim 2 , wherein the at least two models employed include a first deep learning model to classify a target voxel and neighboring voxels and a second deep learning model to classify the target voxel and neighboring patches.

4. The system of claim 1 , wherein the processor is to use the at least one additional neural network comprising a second fully connected neural network and a convolutional neural network to identify the image features and produce the first fully connected neural network and the gradient boosting machine.

5. The system of claim 1 , wherein the image includes a computed tomography image.

6. The system of claim 4 , wherein the image is a three dimensional computed tomography image processed slice-by-slice to identify the object.

7. The system of claim 1 , wherein the processor is to generate a first bounding box in an axial image slice, a second bounding box in a coronal image slice, and a third bounding box in a sagittal image slice and to generate the bounding box around the region of interest in a three-dimensional image by combining the first bounding box, the second bounding box, and the third bounding box.

8. The system of claim 7 , wherein the processor is to be trained using a deep learning network and a database of positive and negative classification results including at least one rotated image slice, at least one translated image slice, and at least one duplicated image slice.

9. A non-transitory computer-readable storage medium including instructions which, when executed, cause a processor to at least implement a method of image processing comprising:

detecting an anatomy in an image;

removing items not included in the anatomy from the image;

generating a bounding box around a region of interest in the anatomy;

classifying image data within the bounding box at the voxel level to identify an object in the region of interest; and

outputting an indication of the object identified in the region of interest segmented in the image, wherein the indication is generated based on both a first fully connected neural network and a gradient boosting machine, wherein the first fully connected neural network and the gradient boosting machine are built using a plurality of image features created using image intensity values and at least one additional neural network, wherein the indication further comprises radiation planning information comprising radiation location information and radiation dosage information for the object, and wherein the object is an organ identified from the image.

10. The non-transitory computer-readable storage medium of claim 9 , further including combining a plurality of models to classify the object in the region of interest, wherein at least two models are employed to classify the image data within the bounding box, the at least two models combined to classify the object and determine at least one feature related to the object.

11. The non-transitory computer-readable storage medium of claim 10 , wherein the at least two models include a first deep learning model to classify a target voxel and neighboring voxels and a second deep learning model to classify the target voxel and neighboring patches.

12. The non-transitory computer-readable storage medium of claim 10 , wherein the at least one additional neural network comprising a second fully connected neural network, and a convolutional neural network are used to identify the image features and produce a first fully connected neural network and the gradient boosting machine to be provided via an output.

13. The non-transitory computer-readable storage medium of claim 9 , wherein the image includes a computed tomography image.

14. The non-transitory computer-readable storage medium of claim 13 , wherein the computed tomography image is a three-dimensional computed tomography image processed slice-by-slice to identify the object.

15. The non-transitory computer-readable storage medium of claim 9 , wherein generating a bounding box around a region of interest in the anatomy further includes:

generating a first bounding box in an axial image slice, a second bounding box in a coronal image slice, and a third bounding box in a sagittal image slice; and generating the bounding box around the region of interest in a three-dimensional image by combining the first bounding box, the second bounding box, and the third bounding box.

16. The non-transitory computer-readable storage medium of claim 15 , further including training at least one model using a deep learning network and a database of positive and negative classification results including at least one rotated image slice, at least one translated image slice, and at least one duplicated image slice.

17. A computer-implemented method of image processing comprising:

detecting, using at least one processor, an anatomy in an image;

removing, using the at least one processor, items not included in the anatomy from the image; generating, using the at least one processor, a bounding box around a region of interest in the anatomy;

classifying, using the at least one processor, image data within the bounding box at the voxel level to identify an object in the region of interest; and

outputting, using the at least one processor, an indication of the object identified in the region of interest segmented in the image, wherein the indication is generated based on both a first fully connected neural network and a gradient boosting machine, wherein the first fully connected neural network and the gradient boosting machine are built using a plurality of image features created using image intensity values and at least one additional neural network, wherein the indication further comprises radiation planning information comprising radiation location information and radiation dosage information for the object, and wherein the object is an organ identified from the image.

18. The method of claim 17 , further including combining a plurality of models to classify the object in the region of interest, wherein at least two models are employed to classify the image data within the bounding box, wherein the at least two models are combined to classify the object and determine at least one feature related to the object.