Systems and methods for analyzing pathologies utilizing quantitative imaging

Patent metadata

Field	Value
Publication number	US-10176408-B2
Application number	US-201514959732-A
Country	US
Kind code	B2
Filing date	Dec 4, 2015
Priority date	Aug 14, 2015
Publication date	Jan 8, 2019
Grant date	Jan 8, 2019

Abstract

Systems and methods for analyzing pathologies utilizing quantitative imaging are presented herein. Advantageously, the systems and methods of the present disclosure utilize a hierarchical analytics framework that identifies and quantify biological properties/analytes from imaging data and then identifies and characterizes one or more pathologies based on the quantified biological properties/analytes. This hierarchical approach of using imaging to examine underlying biology as an intermediary to assessing pathology provides many analytic and processing advantages over systems and methods that are configured to directly determine and characterize pathology from underlying imaging data.

First claim

The invention claimed is:

1. A system comprising a processor and a non-transient storage medium including processor executable instructions implementing an analyzer module including a hierarchical analytics framework configured to:
utilize a first set of machine learned algorithms to identify and quantify a set of biological properties utilizing medical imaging data; and
utilize a second set of machine learned algorithms independent from the first machine learned algorithm to identify and characterize one or more medical conditions based on the quantified biological properties, wherein the characterization of the one or more medical conditions is indicative of therapeutic/treatment options or further diagnostics;
wherein the algorithms in each of the first and second sets of algorithms are independently derived utilizing machine learning; and
wherein the first set of algorithms is distinctly trained from the second set of algorithms.

2. The system of claim 1 , wherein the analytics framework implements an algorithm for identifying and characterizing the one or more medical conditions based on the quantified biological properties wherein a training set from one or more non-radiological or non-imaging data sources was used in training the algorithm.

3. The system of claim 1 , wherein the analytics framework implements an algorithm for identifying and quantifying the biological properties utilizing radiological imaging data, wherein a training set from one or more non-radiological data sources was used training the algorithm.

4. The system of claim 1 , wherein data from a plurality of same or different types of data sources is incorporated into the process of identifying and characterizing the one or more medical conditions.

5. The system of claim 4 , wherein data from one or more non-imaging data sources is used in conjunction with the imaging data such that the set of biological properties includes one or more biological properties identified or quantified based at least in part on the data from one or more non-imaging data sources.

6. The system of claim 5 , wherein the data from non-imaging sources includes one or more of (i) demographics, (ii) results from cultures or other lab tests, (iii) genomic, proteomic or metabolomic expression profiles, or (iv) diagnostic observations.

7. The system of claim 4 , wherein data from one or more non-radiological data sources is used in conjunction with radiological imaging data such that the set of biological properties includes one or more biological properties identified or quantified based at least in part on the data from one or more non-radiological data sources.

8. The system of claim 1 , wherein the system is configured to simultaneously provide a user with information on the one or more medical conditions as well as the underlying biological properties used in the identification or characterization of the one or more medical conditions.

9. The system of claim 1 , wherein the system is configured to determine at least one of (i) which of the biological parameters in the set have the greatest amount of uncertainty regarding the identification or quantification thereof or (ii) which of the biological parameters in the set are most deterministic of the identification or characterization of the one or more medical conditions.

10. The system of claim 1 , wherein the identifying and quantifying the set of biological properties utilizing the imaging data includes receiving patient data including the image data and parsing the received data into a set of empirical parameters including one or more imaging features of an imaged target.

11. The system of claim 10 , wherein the parsing the received data includes pre-processing image data including performing one or more of: (i) intensity vector analysis, (ii) image registration and transformation analysis or (iii) anatomic region analysis.

12. The system of claim 10 , wherein the imaging features are derived based on one or more of: (i) temporal operators, (ii) fractal analysis, (iii) spatial operators or (iv) or an augmented Markov analysis.

13. The system of claim 1 , wherein an imaged target is a lesion and wherein the biological properties include (i) a size of the lesion, (ii) a shape of the lesion, (iii) a characterization of the margin of the lesion, (iv) a solidity of the lesion, (v) a heterogeneity of the lesion, (vi) a measure of the lesion's invasive extent or potential extent, (vii) a compositional measure of calcification related to the lesion and (viii) a measure of cell metabolism with respect to the lesion.

14. The system of claim 1 , wherein at least one or the biological properties is quantified by assessing differences between a plurality of targets.

15. The system of claim 1 , wherein an imaged target is a blood vessel and wherein the biological properties include (i) an indication of plaque coverage of the vessel wall, (ii) an indication of stenosis of the vessel wall, (iii) an indication of dilation of the vessel wall, and (iv) an indication of vessel wall thickness.

16. The system of claim 1 , wherein an imaged target is a vascular tissue and wherein the biological properties include (i) an indication of a lipid core of the vascular or related tissue, (ii) a measure of fibrosis of the vascular or related tissue, (iii) a measure of calcification of the vascular or related tissue, (iv) an indication of any hemorrhage in the vascular or related tissue, (v) a measure of permeability of the vascular or related tissue, (vi) an indication of thrombosis of the vascular or related tissue, and (vii) an indication of ulceration of the vascular or related tissue.

17. The system of claim 1 , wherein set of biological properties includes one or more anatomical, morphological, structural, compositional, functional, chemical, biochemical, physiological, histological or genetic characteristics.

18. The system of claim 1 , wherein the characterization of the one or more medical conditions includes phenotyping the medical conditions.

19. The system of claim 18 , wherein the characterization of the one or more medical conditions further includes determining predictive outcomes for the medical conditions.

20. The system of claim 19 , wherein the one or more predictive outcomes are predicated on a predetermined causality rating between phenotypes and the predictive outcomes.

21. The system of claim 1 , wherein the storage medium further includes processor executable instructions implementing a trainer module, for training one or more algorithms implemented by the hierarchical analytics framework.

22. The system of claim 1 , wherein the storage medium further includes processor executable instructions implementing a cohort module for enabling a user to define one or more cohort groupings of individuals for further analysis.

23. The system of claim 1 , wherein the analyzer module includes algorithms for calculating imaging features from the imaging data, wherein some of the imaging features are computed on a per-pixel basis, while other imaging features are computed on a region-of-interest basis.

24. The system of claim 1 , wherein the algorithms in each of the first and second sets of algorithms are independently characterized by one or more of neural nets, SVMs, partial least squares, principle components analysis or random forests.

25. The system of claim 1 , wherein the analyzer module is configured to enable delineating of a field for the imag

Assignees

• Paik David S
• Buckler Andrew J
• Johnson Kjell
• Ma Xiaonan
• Moulton Keith A
• Elucid Bioimaging Inc

Inventors

• Paik David S
• Buckler Andrew J
• Johnson Kjell
• Ma Xiaonan
• Moulton Keith A

Classifications

• G06V10/84

  using probabilistic graphical models from image or video features, e.g. Markov models or Bayesian networks · CPC title

• G06F18/29

  Graphical models, e.g. Bayesian networks · CPC title

• G06V20/698 Primary

  Matching; Classification · CPC title

• G06K9/6296 Primary

  Physics · mapped topic

• G06K9/00147

  Physics · mapped topic