Method and system for analyzing multi-dimensional data

• US 6,823,286 B2
• Filed: 06/26/2003
• Issued: 11/23/2004
• Est. Priority Date: 06/29/2000
• Status: Expired due to Fees

• Current Assignee: National Institutes of Health (NIH) (Government of the United States of America)
• Original Assignee: Trustees Of Columbia University In The City Of New York (Columbia University)
• Inventors: Yuste, Rafael, Kumar, Vikram S., Froemke, Robert C.
• Primary Examiner(s): Hoff, Marc S.
• Assistant Examiner(s): Raymond, Edward
• Application Number:
  US10/297,160
• Publication Number:
  US 20040015310A1
• Time in Patent Office: 516 Days
• Field of Search: 702/179, 702/180, 702/181, 702/182, 702/187-188, 702/194, 702/196, 717/125, 717/127, 717/132, 717/134, 717/155
• US Class Current: 702/179
• CPC Class Codes: G06F 18/295 Markov models or related models, e.g. semi-Markov models; Markov random fields; Networks embedding Markov models
• CPC Class Codes: G06T 2207/30072 Microarray; Biochip, DNA array; Well plate
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G16B 25/00 ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
• CPC Class Codes: G16B 40/00 ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
• CPC Class Codes: G16B 40/20 Supervised data analysis

Abstract

Disclosed is a method for analyzing a sequence of data arrays. A selection of at least one type of region of interest and at least one region of interest from said data arrays is made. The sequence of data arrays are then transformed into a simplified data array. Events of interest in the selected regions of interest are then detected and stored in a second simplified data array. The data is then analyzed to determine relationships between the detected events of interest

Claims

1. A method for analyzing a sequence of data arrays representative of tissue slice images or gene expression information comprising the steps of:

(a) selecting at least one type of region of interest of said data arrays;

(b) selecting at least one region of interest for each of said at least one selected type of region of interest selected in step (a) from said data arrays;

(c) transforming said sequence of data arrays into a first simplified data array having a first dimension equal to the number of said at least one region of interest selected in step (b), a second dimension equal to the number of data arrays in said sequence, and a third dimension equal to the number of said selected types of regions of interest;

(d) detecting events of interest in said at least one selected region of interest;

(e) storing said detected events of interest in a second simplified data array of binary data, having the same dimensions as said first simplified data array; and

(f) analyzing data in one of the data arrays selected from the group consisting of said first simplified data array and said second simplified data array, to determine relationships between said detected events of interest.

2. The method of claim 1, wherein said sequence of data arrays comprises a series of two dimensional arrays corresponding to a time series of two dimensional observations of an object.

3. The method of claim 1, wherein said object is neuronal cell slice containing a plurality of neurons and wherein said step of selecting a region of interest comprises selecting one or more neurons in said series of two dimensional observations.

4. The method of claim 2, wherein one type of region of interest is selected in step (a) and wherein said step of selecting at least one region of interest comprises selecting at least one portion of said sequence of data arrays that corresponds to a particular spatial region of said two dimensional observations of said object.

5. The method of claim 4, wherein said transforming step comprises extracting data from each array in said sequence of data arrays that corresponds to said particular spatial region of interest and storing said data in said first simplified array, each data entry having an associated first dimensional index and an associated second dimensional index, wherein said first dimensional index is between one and the number of spatial regions corresponding to said selected regions of interest of said sequence of data arrays, and said second dimensional index is between one and the number of two dimensional observations of said object in said sequence of data arrays.

6. The method of claim 5, wherein said step of detecting events of interest comprises

calculating a statistical mean and statistical standard deviation from a data population consisting of all entries in said first simplified array having identical first dimensional indexes, for each of said first dimensional indexes;

determining for each entry in said first simplified array whether said entry exceeds, by a predetermined number of said standard deviation associated with said entry, the mean associated with said entry and denominating such a data entry an event.

7. The method of claim 6, wherein entries in said second simplified data array have the same associated first and second dimensional indexes as corresponding entries in said first simplified data array and wherein said storing said detected events of interests comprises storing a one in said second simplified array when the data entry with the corresponding first and second dimensional indexes in said first simplified array is denominated an event and storing a zero in said second simplified array otherwise.

8. The method of claim 5, wherein said detecting events of interest comprises determining whether a first data entry in said first simplified array exceeds, by a threshold amount, a second data entry in said first simplified array wherein said second data entry has an identical first dimensional index as said first data entry and a second dimensional index corresponding to an earlier point in time than said first data entry and denominating said second data entry an event.

9. The method of claim 8, wherein entries in said second simplified data array have the same associated first and second dimensional indexes as corresponding entries in said first simplified data array and wherein said storing said detected events of interests comprises storing a one in said second simplified array when the data entry with the corresponding first and second dimensional indexes in said first simplified array is denominated an event and storing a zero in said second simplified array otherwise.

10. The method of claim 1, wherein said step of analyzing comprises plotting at least a portion of said data in said first simplified data for visual analysis.

11. The method of claim 1, wherein said step of analyzing comprises detecting said events of interest that are coactive and determining whether the number of coactive events is statistically significant.

12. The method of claim 11, wherein said step of detecting events of interest that are coactive comprises detecting instances where said events of interest are detected in two or more of said regions of interest simultaneously.

13. The method of claim 11, wherein said step of detecting events of interest that are coactive comprises detecting instances were events of interest are detected in two of said regions of interest simultaneously at a plurality of locations along said second dimension of said second simplified data array.

14. The method of claim 1, wherein said step of analyzing comprises calculating a strength of correlation between at least two regions of interest based on the number of coactive events of interest occurring in said at least two regions of interest and displaying a correlation map illustrating the strength of correlation between said regions of interest by lines connecting said regions of interest wherein the thickness of each of the lines is proportional to said calculated strength of correlation between respective regions of interest connected by the line.

15. The method of claim 1, wherein said step of analyzing comprises displaying a cross-correlogram between events of interest occurring in at least one of said regions of interest.

16. The method of claim 1, wherein said step of analyzing comprises detecting at least one hidden Markov state sequence from said second simplified data array.

17. The method of claim 16, wherein said step of analyzing further comprises displaying a cross-correlogram between events of interest occurring in one of said regions of interest while said region of interest is in one of said detected hidden Markov states.

18. The method of claim 1, further comprising:

before said step of detecting, performing a singular valued decomposition on said data in said first simplified data array to calculate a predetermined number of eigenmodes;

modifying said data in said first simplified data array by removing the data that corresponds to the first of said predetermined number of eigenmodes; and

storing said modified data into said first simplified data array.

19. A method for analyzing a sequence of data arrays representative of tissue slice images or gene expression information comprising the steps of:

(a) selecting at least one type of region of interest of said data arrays;

(b) selecting at least one region of interest for each of said at least one type of region of interest selected in step

(b) from said data arrays;

(c) transforming said sequence of data arrays into a first simplified data array having a first dimension equal to the number of said selected regions of interest, a second dimension equal to the number of data arrays in said sequence, and a third dimension equal to the number of said selected types of regions of interest; and

(d) performing a singular valued decomposition on said first simplified data array to determine relationships between said regions of interest.

20. A system for analyzing a sequence of data arrays representative of tissue slice images or gene expression information comprising in a data processor:

a region type selector for selecting at least one type of region of interest of said data arrays;

a region selector for selecting at least one region of interest for each of said selected type of region of interest from said data arrays;

a first data transformer for transforming said sequence of data arrays into a first simplified data array having a first dimension equal to the number of said selected regions of interest, a second dimension equal to the number of data arrays in said sequence, and a third dimension equal to the number of said selected types of regions of interest;

an event detector for detecting events of interest in said regions of interest;

a second data transformer for storing said detected events of interest into a second simplified data array of binary data, having the same dimensions as said first simplified data array; and

a data analyzer for analyzing data in one of the data arrays selected from the group consisting of said first simplified data array and said second simplified data array, to determine relationships between said detected events of interest.

21. The system of claim 20, wherein said sequence of data arrays comprises a series of two dimensional arrays corresponding to a time series of two dimensional observations of an object.

22. The system of claim 21, wherein said object is neuronal cell slice containing a plurality of neurons and wherein said step of selecting a region of interest comprises selecting one or more neurons in said series of two dimensional observations.

23. The system of claim 21, wherein said region type selector selects one type of region of interest and wherein said region selector selects at least one portion of said sequence of data arrays that corresponds to a particular spatial region of said two dimensional observations of said object.

24. The system of claim 23, wherein said region selector receives coordinates corresponding to said particular spatial region from a user.

25. The system of claim 23, wherein said first data transformer extracts data from each array in said sequence of data arrays that corresponds to said particular spatial region of interest and stores said data in said first simplified array, wherein each data entry in said first simplified array has an associated first dimensional index and an associated second dimensional index, wherein said first dimensional index is between one and the number of spatial regions corresponding to selected regions of interest of said sequence of data arrays and said second dimensional index is between one and the number of two dimensional observations of said object in said sequence of data arrays.

26. The system of claim 25, wherein said event detector further comprises:

a statistical calculator for calculating a statistical mean and statistical standard deviation from a data population consisting of all entries in said first simplified data array having identical first dimensional indexes, for each of said first dimensional indexes; and

a comparator for determining for each entry in said first simplified array whether said entry exceeds, by a predetermined number of said standard deviations associated with said entry the mean associated with said entry and denominating such a data entry an event.

27. The system of claim 26, wherein said second data transformer stores entries in said second simplified data array having the same associated first and second dimensional index as corresponding entries in said first simplified data array and wherein said second data transformer stores a one in said second simplified array when the data entry with corresponding first and second dimensional indexes in said first simplified array is denominated an event and stores a zero in said second simplified data array otherwise.

28. The system of claim 25, wherein said event detector determines whether a first data entry in said first simplified array exceeds, by a threshold amount, a second data entry in said first simplified data array, wherein said second data entry has an identical first dimensional index as said first data entry and a second dimensional index corresponding to an earlier point in time than said first data entry, and denominates said second data entry an event.

29. The system of claim 28, wherein said second data transformer stores entries in said second simplified data array having the same associated first and second dimensional index as corresponding entries in said first simplified data array and wherein said second data transformer stores a one in said second simplified array when the data entry with corresponding first and second dimensional indexes in said first simplified array is denominated an event and stores a zero in said second simplified data array otherwise.

30. The system of claim 20, wherein said data analyzer plots at least a portion of said data in said first simplified array for visual analysis.

31. The system of claim 20, wherein said data analyzer detects said events of interest that are coactive and determines whether the number of coactive events is statistically significant.

32. The system of claim 31, wherein said data analyzer detects said events of interest that are coactive comprises operability to detect instances where said events of interest are detected in two or more of said regions of interest simultaneously.

33. The system of claim 31, wherein said data analyzer detects instances where events of interest are detected in two of said regions of interest simultaneously at a plurality of locations along said second dimension of said second simplified data array.

34. The system of claim 20, wherein said data analyzer calculates a strength of correlation between at least two regions of interest based on the number of coactive events of interest occurring in said at least two regions of interest and displays a correlation map illustrating the strength of correlation between said regions of interest by lines connecting said regions of interest wherein the thickness of each of the lines is proportional to said calculated strength of correlation between respective regions of interest connected by the line.

35. The system of claim 20, wherein said data analyzer displays a cross-correlogram between events of interest occurring in at least one of said regions of interest.

36. The system of claim 20, wherein said data analyzer detects at least one hidden Markov state sequence from said second simplified data array.

37. The system of claim 36, wherein said data analyzer displays a cross-correlogram between events of interest occurring in one of said regions of interest while said region of interest is in one of said detected hidden Markov states.

38. The system of claim 20, wherein said system further comprises:

a decomposer to perform a singular valued decomposition on said data in said first simplified data array and calculating a predetermined number of eigenmodes;

a data modifier for modifying said data said first simplified data array by removing the data that corresponds to the first of said predetermined number of eigenmodes; and

data storage for storing said modified data into said first simplified data array.

39. A system for analyzing a sequence of data arrays representative of tissue slice images or gene expression information comprising in a data processor:

a region type selector for selecting at least one type of region of interest of said data arrays;

a region selector for selecting at least one region of interest for each of said selected type of region of interest from said data arrays;

a first data transformer for transforming said sequence of data arrays into a first simplified data array having a first dimension equal to the number of said selected regions of interest, a second dimension equal to the number of data arrays in said sequence, and a third dimension equal to the number of said selected types of regions of interest;

a decomposer for performing a singular value decomposition on said first simplified data array to determine relationship between said regions of interest.