Methods of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification and for T-Cell therapy initiation and tracking and related precision medical treatment predictive modeling
A method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification comprises providing a multilayer convolutional network for neurological tumor identification configured for segmenting data sets of full neurologic scans into resolution voxels; supervised learning and validation of the platform by classification of tissue within classification voxels of a specific given training and validation data sets by the multilayer convolutional network for neurological tumor identification with each classification voxel of the training and validation data sets having a predetermined ground truth; and implementing the platform by classification of tissue within classification voxels of a specific given patient data sets by the multilayer convolutional network for neurological tumor identification with each classification voxel of each data set assigned a label. The platform may be used for T-cell therapy initiation and tracking. An artificial intelligence based neuroradiology platform implemented according to the method is disclosed.
1. A method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification comprising the steps of:
Providing a multilayer convolutional network for neurological tumor identification configured for segmenting data sets of full neurologic scans into resolution voxels, wherein the resolution voxels are representative of uniform volumetric regions and define the smallest volumetric region within which the platform can segment data sets;
Defining classification voxels which are representative of uniform volumetric regions wherein each classification voxel is made up of one or more resolution voxels;
Supervised learning of the platform by classification of tissue within said classification voxels of a specific given training data set by the multilayer convolutional network for neurological tumor identification with each classification voxel of the training data set having a predetermined ground truth;
Validating the classification of tissue within said classification voxels of a specific given validation data set by the multilayer convolutional network for neurological tumor identification with each classification voxel of the validation data set having a predetermined ground truth;
Implementing the platform by classification of tissue within classification voxels of a specific given patient data sets by the multilayer convolutional network for neurological tumor identification with each classification voxel of each data set assigned a label.
2. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 1 wherein the neurological tumors are tumors of the spine.
3. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 1 wherein the neurological tumors are brain tumors.
4. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 3 wherein the full neurologic scans include MRI scans.
5. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the classification voxels are at least 20 times larger than the resolution voxels.
6. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the classification voxels are at least 10 times larger than the resolution voxels.
7. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the classification voxels are at least 5 times larger than the resolution voxels.
8. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 7 resolution voxels are cubes of 1 mm.
9. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the specific given training data set includes at least 40 full scans.
10. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the specific given training data set includes at least 60 full scans.
11. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the specific given training data set includes at least 100 full scans.
12. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the validation data set includes less full scans than the training data set.
13. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the validating step results in a standard measuring metric.
14. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 13 wherein the standard measuring metric includes one of a Sorensen-Dice coefficient or a Hausdorff distance.
15. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 13 wherein the standard measuring metric includes a mean Sorensen-Dice coefficient of at least 0.60.
16. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 13 wherein the standard measuring metric includes a mean Hausdorff distance less than −1.50.
17. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 13 further including the step of resetting the size of the classification voxel following supervised learning, wherein the size of the classification voxel used in the validating step is smaller than the size of the classification voxel used in supervised learning.
18. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 13 further including the step of resetting the size of the classification voxel following validation, wherein size of the classification voxel used in the validating step is equal to or smaller than the classification voxel size used by the platform in implementation.
19. The method of implementing an artificial intelligence based neuroradiology platform for neurological tumor identification according to claim 4 wherein the platform identifies false markers of tumor growth as an indication to introduce T-cell therapy.
20. An artificial intelligence based neuroradiology platform for neurological tumor identification implemented according to the method of claim 1 .