System, devices, and methods for detecting occlusions in a biological subject including spectral learning

• US 20090292195A1
• Filed: 04/30/2009
• Published: 11/26/2009
• Est. Priority Date: 12/18/2007
• Status: Active Grant

• Current Assignee: Gearbox LLC (Intellectual Ventures LLC)
• Original Assignee: Searete LLC (Intellectual Ventures LLC)
• Inventors: Boyden, Edward S., Leuthardt, Eric C.
• Granted Patent: US 9,672,471 B2
• Time in Patent Office: Days
• US Class Current: 600/407
• CPC Class Codes: A61B 5/02007 Evaluating blood vessel condition, e.g. elasticity, compliance
• CPC Class Codes: A61B 5/021 Measuring pressure in heart or blood vessels A61B5/0205 takes precedence
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

Systems, devices, and methods are described for detecting an embolus, thrombus, or a deep vein thrombus in a biological subject.

Claims

1. A hemodynamics monitoring method, comprising:

obtaining a first spectral information from a biological subject while varying at least one of a wavelength or a frequency associated with an interrogation optical excitation energy source;

partitioning the spectral information into one or more information subsets; and

comparing at least one parameter associated with a second spectral information from a biological subject associated to at least one parameter associated with at least one of the one or more information subsets.

2. The hemodynamics monitoring method of claim 1, wherein partitioning the spectral information into the one or more information subsets includes grouping the spectral information into one or more information subsets using a clustering protocol.

3. The hemodynamics monitoring method of claim 1, wherein partitioning the spectral information into the one or more information subsets includes grouping the spectral information into one or more information subsets using at least one of a Spectral Clustering protocol or a Spectral Learning protocol.

4. The hemodynamics monitoring method of claim 1, wherein partitioning the spectral information into the one or more information subsets includes grouping the spectral information into one or more information subsets using at least one of a Fuzzy C-Means Clustering protocol, a Graph-Theoretic protocol, a Hierarchical Clustering protocol, a K-Means Clustering protocol, a Locality-Sensitive Hashing protocol, a Mixture of Gaussians protocol, a Model-Based Clustering protocol, a Cluster-Weighted Modeling protocol, an Expectations-Maximization protocol, a Principal Components Analysis protocol, or a Partitional protocol.

5. The hemodynamics monitoring method of claim 1, wherein partitioning the spectral information into the one or more information subsets includes partitioning a detected spectrum into one or more information subsets with at least one of a prism, a monochromator, a diffraction grating, or a bypass filter.

6. The hemodynamics monitoring method of claim 1, further comprising:

generating a response based on the comparison of the at least one parameter associated with the second spectral information to the at least one parameter associated with at least one of the one or more information subsets.

7. The hemodynamics monitoring method of claim 6, wherein generating the response based on the comparison includes generating a response based on the comparison of the at least one parameter associated with the second spectral information to a threshold diameter of at least one cluster associated with a set of reference cluster information associated with the biological subject.

8. The hemodynamics monitoring method of claim 6, wherein generating the response based on the comparison includes generating a response based on the comparison of the at least one parameter associated with the second spectral information to a threshold diameter of at least one cluster associated with a set of reference cluster information from the biological subject.

9. The hemodynamics monitoring method of claim 6, wherein generating the response based on the comparison includes generating a response based on the comparison of the second spectral information to an average squared distance of at least one cluster centroid associated with a reference information data.

10. The hemodynamics monitoring method of claim 6, wherein generating the response based on the comparison includes generating a response based on the comparison of the second spectral information to an inverse of a distance to at least one cluster centroid associated with a reference information data.

11. The hemodynamics monitoring method of claim 6, wherein generating the response includes generating an occlusion aggregation rate.

12. The hemodynamics monitoring method of claim 6, wherein generating the response includes generating a response based on a voxel intensity.

13. The hemodynamics monitoring method of claim 6, wherein generating the response includes generating an occlusion aggregation rate based on a voxel intensity.

14. The hemodynamics monitoring method of claim 6, wherein generating the response includes generating at least one parameter associated with a degree of belonging to at least one cluster centroid associated with a reference information data.

15. The hemodynamics monitoring method of claim 6, wherein generating the response includes generating a rate of deviation from a threshold value.

16. The hemodynamics monitoring method of claim 6, further comprising:

performing a real-time update of at least one parameter associated with a spectral blood clotting model associated with the biological subject.

17. A computer program product, comprising:

one or more signal-bearing media containing computer instructions which, when run on a computing device, cause the computing device to implement a method including

obtaining a first spectral information from a biological subject while varying at least one of a wavelength or a frequency associated with an interrogation optical excitation energy source;

partitioning the spectral information into one or more information subsets; and

comparing at least one parameter associated with a second spectral information from a biological subject associated to at least one parameter associated with at least one of the one or more information subsets.

18. The computer program product of claim 17, wherein the one or more signal-bearing media containing computer instructions further comprises:

computer instructions which, when run on a computing device, cause the computing device to

generating a response based on the comparison of the at least one parameter associated with the second spectral information to the at least one parameter associated with at least one of the one or more information subsets.

19. The computer program product of claim 17, wherein partitioning the spectral information into the one or more information subsets includes grouping the spectral information into one or more information subsets using a clustering protocol.

20. The computer program product of claim 17, wherein partitioning the spectral information into the one or more information subsets includes grouping the spectral information into one or more information subsets using at least one of a Spectral Clustering protocol or a Spectral Learning protocol.

21. The computer program product of claim 17, wherein partitioning the spectral information into the one or more information subsets includes grouping the spectral information into one or more information subsets using at least one of a Fuzzy C-Means Clustering protocol, a Graph-Theoretic protocol, a Hierarchical Clustering protocol, a K-Means Clustering protocol, a Locality-Sensitive Hashing protocol, a Mixture of Gaussians protocol, a Model-Based Clustering protocol, a Cluster-Weighted Modeling protocol, an Expectations-Maximization protocol, a Principal Components Analysis protocol, or a Partitional protocol.

22. The computer program product of claim 17, wherein the one or more signal-bearing media containing computer instructions further includes:

computer instructions which, when run on a computing device, cause the computing device to

performing a real-time update of at least one parameter associated with a spectral blood vessel occlusion model associated with the biological subject.

23. A computer program product, comprising:

one or more signal-bearing media containing computer instructions which, when run on a computing device, cause the computing device to implement a method including

obtaining a first spectral information from a biological subject while varying at least one of a wavelength or a frequency associated with an interrogation optical excitation energy source;

partitioning the spectral information into one or more information subsets; and

comparing at least one parameter associated with a second spectral information from a biological subject to at least one parameter associated with at least one of the one or more information subsets.

24. The computer program product of claim 23, wherein partitioning the spectral information into the one or more information subsets includes automatically generating one or more data clusters using at least one of a Spectral Clustering protocol or a Spectral Learning protocol.

25. The computer program product of claim 23, wherein partitioning the spectral information into the one or more information subsets automatically generating one or more data clusters using at least one of a Fuzzy C-Means Clustering protocol, a Graph-Theoretic protocol, a Hierarchical Clustering protocol, a K-Means Clustering protocol, a Locality-Sensitive Hashing protocol, a Mixture of Gaussians protocol, a Model-Based Clustering protocol, a Cluster-Weighted Modeling protocol, an Expectations-Maximization protocol, a Principal Components Analysis protocol, or a Partitional protocol.

26. The computer program product of claim 23, wherein the one or more signal-bearing media containing computer instructions further comprises:

computer instructions which, when run on a computing device, cause the computing device to

generating a response based at least in part on the comparison of the at least one parameter associated with the second spectral information to the at least one parameter associated with at least one of the one or more information subsets.

27. The computer program product of claim 23, wherein generating the response includes generating at least one of information associated with a statistical probability, a local cluster density, a deviation from a target cluster distance, a distance from a cluster centroid, an euclidina distance, or a probability density.

28. An apparatus, comprising:

circuitry for obtaining spectral information from a biological subject while varying at least one of a wavelength or a frequency associated with an interrogation optical excitation energy source; and

circuitry for partitional clustering the obtained spectral information into one or more information subsets.

29. The apparatus of claim 28, further comprising:

circuitry for real-time comparing at least one parameter associated with an obtained spectral information from a biological subject to at least one parameter associated with at least one of the one or more information subsets.

30. An occlusion monitoring method, comprising:

obtaining spectral information from a biological subject while varying at least one of a wavelength or a frequency associated with an interrogation optical excitation energy source;

comparing at least one parameter associated with the obtained spectral information to one or more information subsets derived from partitioning spectral information associated with the biological subject; and

generating a response based on the comparison of the at least one parameter associated with the obtained spectral information to the one or more information subsets derived from partitioning spectral information associated with the biological subject.

31. The method of claim 30, wherein obtaining the spectral information includes concurrently detecting an excitation radiation and an emission radiation to generate a spectrum.

32. The method of claim 30, wherein comparing the at least one parameter associated with the obtained spectral information to the one or more information subsets derived from partitioning the spectral information includes comparing the at least one parameter associated with the obtained spectral information to one or more information subsets derived from grouping the spectral information into one or more information subsets using at least one of a Spectral Clustering protocol or a Spectral Learning protocol.

33. The method of claim 30, wherein comparing the at least one parameter associated with the obtained spectral information to the one or more information subsets derived from partitioning spectral information includes comparing the at least one parameter associated with the obtained spectral information to one or more information subsets derived from grouping the spectral information into one or more information subsets using at least one of a Fuzzy C-Means Clustering protocol, a Graph-Theoretic protocol, a Hierarchical Clustering protocol, a K-Means Clustering protocol, a Locality-Sensitive Hashing protocol, a Mixture of Gaussians protocol, a Model-Based Clustering protocol, a Partitional protocol, a Spectral Clustering protocol, a Cluster-Weighted Modeling protocol, an Expectations-Maximization protocol, a Principal Components Analysis protocol, or a Spectral Learning protocol.

34. The method of claim 30, wherein generating the response includes generating at least one of information associated with a statistical probability, a local cluster density, a deviation from a target cluster distance, a distance from a cluster centroid, an euclidina distance, a probability density.

35. The method of claim 30, wherein generating the response includes automatically updating at least one parameter associated with a spectral tissue model.

36. The method of claim 30, wherein generating the response includes performing a real-time update of at least one parameter associated with a spectral blood clotting model associated with the biological subject.