Method for a brain region location and shape prediction
A volumetric segmentation method is disclosed for brain region analysis, in particular but not limited to, regions of the basal ganglia such as the subthalamic nucleus (STN). This serves for visualization and localization within the sub-cortical region of the basal ganglia, as an example of prediction of a region of interest for deep brain stimulation procedures. A statistical shape model is applied for variation modes of the STN, or the corresponding regions of interest, and its predictors on high-quality training sets obtained from high-field, e.g., 7T, MR imaging. The partial least squares regression (PLSR) method is applied to induce the spatial relationship between the region to be predicted, e.g., STN, and its predictors. The prediction accuracy for validating the invention is evaluated by measuring the shape similarity and the errors in position, size, and orientation between manually segmented STN and its predicted one.
1. A brain image pipeline method for operating an electronic device, comprising:
receiving a patient's brain image, wherein the patient's brain image has a predictor region and a region of interest;
accessing a database including a plurality of brain images different than the patient's brain image;
retrieving from the database a training set of one or more brain images, wherein each brain image of the training set has a predictor region associated with a region of interest that is anatomically different than the predictor region, and wherein the predictor region of each brain image of the training set corresponds to the predictor region in the patient's brain image and the region of interest of each brain image of the training set corresponds to the region of interest in the patient's brain image;
processing each brain image of the training set to generate predictor information representative of a predicted region of interest having one or more of a predicted shape, location, size or orientation based on a relationship between a shape, location, size or orientation of the region of interest of each brain image in the training set with respect to an associated shape, location, size or orientation of the predictor region of each brain image in the training set; and
processing the patient's brain image using the predictor information to incorporate the predicted region of interest having the one or more of the predicted shape, location, size or orientation into the patient's brain image with respect to the predictor region in the patient's brain image, to produce a patient-specific atlas.
2. The brain image pipeline method of claim 1 wherein accessing the database including the plurality of brain images includes accessing a database including one or both of (1) two or more image types or (2) two or more imaging modalities.
3. The brain image pipeline method of claim 2 wherein:
receiving the patient's brain image includes receiving a magnetic resonance brain image taken with a field strength of 3 Tesla or lower;
accessing the database including the plurality of brain images includes accessing the database including magnetic resonance brain images taken with a field strength of 7 Tesla or higher; and
retrieving from the database the training set of brain images includes retrieving at least one of the magnetic resonance brain images taken with the field strength of 7 Tesla or higher.
4. The brain image pipeline method of claim 3 wherein:
accessing the database including the plurality of brain images includes accessing the database including computed tomography brain images; and
retrieving from the database the training set of brain images includes retrieving at least one of the computed tomography brain images.
5. The brain image pipeline method of claim 2 wherein:
accessing the database including the plurality of brain images includes accessing the database including magnetic resonance brain images and computed tomography brain images; and
retrieving from the database the training set of brain images includes retrieving at least one magnetic resonance brain image and at least one computed tomography brain image.
6. The brain image pipeline method of claim 2 wherein accessing the database including the plurality of brain images includes accessing the database including both of (1) two or more image types and (2) two or more imaging modalities.
7. The brain image pipeline method of claim 1 wherein:
processing each brain image of the training set includes processing each brain image of the training set to generate the predictor information representative of the predicted region of interest having a predicted shape and location; and
processing the patient's brain image using the predictor information includes processing the patient's brain image using the predictor information to incorporate the predicted region of interest having the predicted shape and location into the patient's brain image.
8. The brain image pipeline method of claim 1 wherein:
processing each brain image of the training set of brain images includes processing each brain image of the training set of brain images to generate the predictor information representative of the predicted region of interest having a predicted shape and pose; and
processing the patient's brain image using the predictor information includes processing the patient's brain image using the predictor information to incorporate the predicted region of interest having the predicted shape and pose into the patient's brain image.
9. The brain image pipeline method of claim 1 wherein:
processing each brain image of the training set of brain images to generate the predictor information includes generating a statistical shape model representative of the predicted region of interest; and
processing the patient's brain image to incorporate the predicted region of interest includes processing the patient's brain image using the statistical shape model.
10. The brain image pipeline method of claim 1 wherein the region of interest includes the subthalamic nucleus.
11. The brain image pipeline method of claim 10 wherein the predictor region includes the red nucleus.
12. The brain image pipeline method of claim 11 and further including:
receiving a post-implantation image of the brain of the patient after insertion of an implant; and
merging the post-implantation image with the patient-specific atlas to form a post-implantation composite image.
13. The brain image pipeline method of claim 12 wherein:
receiving the post-implantation image includes receiving an image of the brain of the patient after the insertion of an electrode; and
the method further includes displaying the post-implantation composite image including the electrode as an aid in programming the electrode.
14. The brain image pipeline method of claim 1 wherein the method further includes:
incorporating into the patient-specific atlas one or both of a target or a path for a surgical procedure to follow to arrive at the target; and
displaying the patient-specific atlas with the incorporated one or both of the target or the path.
15. The brain image pipeline method of claim 14 wherein:
processing the patient's brain image to produce a patient-specific atlas includes producing the patient-specific atlas including a subthalamic nucleus; and
incorporating the one or both of the target or the path into the patient-specific atlas includes incorporating one or both of the target or the path into the patient-specific atlas including the subthalamic nucleus.