Subvolume identification for prediction of treatment outcome
View Patent ↗Physiological imaging-defined subvolumes of tissues/disease are identified to yield spatially-defined prognostic and/or predictive indicators and/or focal therapy targets within such tissues, in particular tumors, for evaluation over time, for example, prior to and after a therapy treatment. Medical image data is analyzed to delineate subvolumes of tissue based upon multiple physiological, metabolic, and biological imaging properties, where those subvolumes are extracted and analyzed in a probabilistic manner to associate with one or more abnormal or disease phenotype conditions.
1. A method of analyzing medical image data of a region of interest in a sample tissue, the method comprising:
obtaining, at a computer system, the medical image data of the region of interest, the medical image data containing a plurality of image segments;
identifying, at the computer system, one or more candidate physiological, metabolic, molecular and/or biologic parameters that may indicate an abnormal or disease phenotype condition within the region of interest;
analyzing, in a subvolume analysis engine, each of the image segments using an algorithm to determine, for each image segment, a probability function for each of the identified one or more candidate physiological, metabolic, and/or biologic parameters; and
modeling the image segments, in the subvolume analysis engine, to form a probability map from the analysis of each of the image segments, the probability map providing an indication of heterogeneous distributions of the one or more candidate physiological, metabolic, and/or biological parameters for each of the image segments of the medical image data and analyzing the resulting model to identify the abnormal or disease phenotype condition within the region of interest by determining, for the medical image data, an abnormal or disease phenotype metric from the model and comparing the abnormal or disease phenotype metric with other obtained medical image data, where the identification produces a diagnostic determination of the region of interest, a prognostic determination of the target tissue, or a predictive determination of the target tissue.
2. The method of claim 1 , further comprising automatically applying the diagnostic determination, prognostic determination, or predictive determination to a target tissue treatment.
3. The method of claim 2 , wherein the target tissue is a tumor and wherein the target tissue treatment is radiation therapy.
4. The method of claim 1 , wherein the algorithm is a fuzzy logic algorithm, a genetic algorithm or Gaussian mixture model.
5. The method of claim 1 , wherein the image segments are voxels or groups of voxels.
6. The method of claim 1 , wherein the one or more physiological, metabolic, molecular and/or biologic parameters include but not limit to regional cerebral blood volume (rCBV), a volume transfer coefficient (K trans ), diffusion coefficient and/or fluorodeoxyglucose (FDG) uptake.
7. The method of claim 1 , wherein the medical image data is magnetic resonance image (MRI) data, positron emission tomography (PET) image data or single photon emission computed tomography (SPECT) image data.
8. The method of claim 1 , wherein analyzing each of the image segments comprises applying a probability density function for each of the candidate parameters for each image segment.
9. The method of claim 8 , further comprising determining the probability density function for each of the candidate parameters for an image segment collected prior to treatment of the target tissue and for an image segment collected after starting treatment of the target tissue.
10. The method of claim 8 , wherein the analysis comprises modeling the probability density functions using a fuzzy-c-means (FCM) clustering analysis, Gaussian mixture model and/or other discrimination analyses.
11. The method of claim 1 , further comprising assigning each of the image segments to the one or more of the candidate parameters based on the analysis of the resulting model.
12. Apparatus for analyzing medical image data of a region of interest in a sample tissue, the apparatus comprising:
a computer system having a processor executing instructions that, when executed, (i) obtain the medical image data of the region of interest, the medical image data containing image segments, and (ii) identify one or more candidate physiological, metabolic, molecular and/or biologic parameters or features that may indicate an abnormal or disease phenotype condition within the region of interest; and
the computer system further comprising a subvolume analysis engine to (i) analyze each of the image segments using an algorithm to determine, for each image segment, a probability function for each of the identified one or more candidate physiological, metabolic, and/or biologic parameters, (ii) model the image segments, in the subvolume analysis engine to form a probability map from the analysis of each of the image segments, the probability map providing an indication of heterogeneous distributions of the one or more candidate physiological, metabolic, and/or biological parameters for each of the image segments of the medical image data, and (iii) analyze the resulting model to identify the abnormal or disease phenotype condition within the region of interest by determining, for the medical image data, an abnormal or disease phenotype metric from the model and comparing the abnormal or disease phenotype metric with other obtained medical image data, where the identification produces a diagnostic determination of the region of interest, a prognostic determination of the target tissue, or a predictive determination of the target tissue.
13. The apparatus of claim 12 , wherein the subvolume analysis engine is configured to further automatically apply the diagnostic determination, prognostic determination, or predictive determination to a target tissue treatment.
14. The apparatus of claim 12 , wherein the algorithm is a fuzzy logic algorithm, a genetic algorithm or Gaussian mixture model.
15. The apparatus of claim 12 , wherein the image segments are voxels or groups of voxels.
16. The apparatus of claim 12 , wherein the one or more physiological, metabolic, molecular and/or biologic parameters include but not limit to regional cerebral blood volume (rCBV), a volume transfer coefficient (K trans ), diffusion coefficient and/or fluorodeoxyglucose (FDG) uptake.
17. The apparatus of claim 12 , wherein the medical image data is magnetic resonance image (MRI) data, positron emission tomography (PET) image data or single photon emission computed tomography (SPECT) image data.
18. The apparatus of claim 12 , wherein the subvolume analysis engine is configured to apply a probability density function for each of the candidate parameters for each image segment.
19. The apparatus of claim 18 , the subvolume analysis engine is configured to further determine the probability density function for each of the candidate parameters for an image segment collected prior to treatment of the target tissue and for an image segment collected after starting treatment of the target tissue.
20. The apparatus of claim 18 , wherein the analysis comprises modeling the probability density functions using a fuzzy-c-means (FCM) clustering analysis, Gaussian mixture model and/or other discrimination analyses.