Generation of continuous mathematical model for common features of a subject group

• US 7,136,787 B2
• Filed: 12/19/2001
• Issued: 11/14/2006
• Est. Priority Date: 12/19/2001
• Status: Expired due to Term

• Current Assignee: Evidera, Inc. (United kingdom) (Thermo Fisher Scientific Luxembourg I SARL)
• Original Assignee: Archimedes, Inc. (Thermo Fisher Scientific Luxembourg I SARL)
• Inventors: Schlessinger, Leonard, Eddy, David
• Primary Examiner(s): Shah, Kamini
• Assistant Examiner(s): DAY, HERNG DER
• Application Number:
  US10/025,964
• Publication Number:
  US 20050288910A1
• Time in Patent Office: 1,791 Days
• Field of Search: 703/2, 703/11, 707/1, 707/6, 707/7, 382/132
• US Class Current: 703/2
• CPC Class Codes: G06F 17/10 Complex mathematical operations function generation by table look-up G06F1/03; evaluation of elementary functions by calculation G06F7/544
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders
• CPC Class Codes: Y10S 707/99931 Database or file accessing
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99937 Sorting

Abstract

A method for generating a continuous mathematical model of a feature common to subjects in a subject group includes selecting a sample data set from each subject in the subject group, selecting a set of expansion functions to be used in the representation of the sample data set, mathematically expanding each member of the sample data set in the form of a summation of results of multiplying each the expansion function in the set of expansion functions by a different mathematical parameter wherein the expanding determines a value for each of the different mathematical parameters, deriving a corresponding distribution function for each of the mathematical parameters, and generating the continuous mathematical model of the feature from the derived distribution functions and the expansion functions. In this way, the model is continuous in time, incorporates dependencies between various parameters, and allows for creation of simulated subjects having pertinent features occurring in real subjects.

Claims

1. A computer-implemented method for generating a continuous mathematical model of a feature common to subjects in a subject group, wherein the subjects are biological subjects and the feature is a biological feature, the method comprising:

determining a plurality of sample data sets corresponding to at least one physiological characteristic of the subjects in the subject group;

determining, from the sample data sets, a plurality of values for one or more mathematical parameters corresponding to one or more basis functions for the continuous mathematical model;

determining, from the values for the one or more mathematical parameters, one or more distribution-function parameters for specifying one or more distribution functions for the one or more mathematical parameters, wherein a summation of the one or more basis functions multiplied by sampled values of the one or more distribution functions provides the continuous mathematical model of the feature;

simulating the feature by generating sampled values of the one or more distribution functions by computer;

calculating at least one statistical property of the simulated feature for characterizing the at least one physiological characteristic of the subjects in the subject group; and

saving the at least one statistical property of the simulated feature.

2. A method according to claim 1, wherein determining the sample data sets includes receiving the sample data sets from an external data source and storing the sample data sets in a computer memory.

3. A method according to claim 2, wherein determining the sample data sets includes measuring values for the feature and storing the values for the feature in the external data source.

4. A method according to claim 3, wherein measuring the values for the feature includes measuring values of blood pressure for the subjects in the subject group.

5. A method according to claim 1, wherein determining the values for the one or more mathematical parameters includes:

determining initial values for the one or more mathematical parameters according to an optimization criterion;

separating the initial values into bins with corresponding bin ranges.

6. A method according to claim 1, wherein determining the values for the one or more mathematical parameters includes calculating the values for the one or more mathematical parameters from the sample data sets according to an optimization criterion.

7. A method according to claim 1, wherein determining the one or more distribution-function parameters includes calculating the one or more distribution-function parameters from the values for the one or more mathematical parameters according to an optimization criterion.

8. A method according to claim 1, further comprising:

displaying the at least one statistical property of the simulated feature.

9. A method according to claim 1, wherein determining the values for the one or more mathematical parameters corresponding to the one or more basis functions includes:

selecting a plurality of initial basis functions;

determining a plurality of values for a plurality of mathematical parameters corresponding to the initial basis functions;

determining, from the values for the mathematical parameters corresponding to the initial basis functions, a correlation matrix for the initial basis functions; and

determining, from the correlation matrix, the one or more basis functions according to a de-correlation criterion.

10. A method according to claim 1, wherein the feature is a first feature selected from a plurality of features that includes a second feature, and the method further comprises:

determining, from the sample data sets, a plurality of values for one or more second mathematical parameters corresponding to one or more second basis functions for a continuous mathematical model of the second feature; and

determining, from the values for the one or more second mathematical parameters, one or more second distribution-function parameters, wherein

values for the features other than the second feature and values for the one or more second distribution-function parameters specify the one or more second distribution functions for the one or more second mathematical parameters, and

a summation of the one or more second basis functions multiplied by sampled values of the one or more second distribution functions provides the continuous mathematical model of the second feature.

11. A method according to claim 10, further comprising:

simulating the second feature by computer, for given values of the features other than the second feature, by generating sampled values of the one or more second distribution functions;

calculating at least one statistical property of the simulated second feature for characterizing the at least one physiological characteristic of the subjects in the subject group; and

saving the at least one statistical property of the simulated second feature.

12. A method according to claim 1, wherein the one or more basis functions include a single basis function and the summation includes a single term.

13. A method according to claim 1, wherein the one or more basis functions include a plurality of orthogonal functions over a continuous interval.

14. A method according to claim 1, wherein the one or more basis functions include one or more hybrid functions that characterize features common to the subject group over a continuous interval.

15. A method according to claim 14, wherein

the continuous mathematical model of the feature includes a model for occlusion of a coronary artery over the continuous interval, and

the one or more hybrid functions include a first function for blood pressure and a second function for cholesterol level.

16. A method according to claim 15, wherein the one or more hybrid functions include a third function for a product of the blood pressure and the cholesterol level.

17. A method according to claim 1, wherein a computer determines the values for the one or more mathematical parameters and the one or more distribution-function parameters.

18. A method according to claim 1, wherein feature is selected from the group consisting of blood pressure, cholesterol levels, bone mineral density, patency of a coronary artery, heart electrical potentials, contractility of myocardium, cardiac output, visual acuity, serum potassium level, observations for a rash, diameter of a coronary artery, and cancer spread measurements.

19. A method according to claim 1, wherein the one or more distribution functions include at least one normal distribution function and the one or more distribution-function parameters include at least one corresponding standard-deviation parameter.

20. An apparatus for generating a continuous mathematical model of a feature common to subjects in a subject group, wherein the subjects are biological subjects and the feature is a biological feature, the apparatus comprising a computer for executing computer instructions, wherein the computer includes computer instructions for:

determining a plurality of sample data sets corresponding to at least one physiological characteristic of the subjects in the subject group;

determining, from the sample data sets, a plurality of values for one or more mathematical parameters corresponding to one or more basis functions for the continuous mathematical model;

determining, from the values for the one or more mathematical parameters, one or more distribution-function parameters for specifying one or more distribution functions for the one or more mathematical parameters, wherein a summation of the one or more basis functions multiplied by sampled values of the one or more distribution functions provides the continuous mathematical model of the feature;

simulating the feature by generating sampled values of the one or more distribution functions by computer;

calculating at least one statistical property of the simulated feature for characterizing the at least one physiological characteristic of the subjects in the subject group; and

saving the at least one statistical property of the simulated feature.

21. An apparatus according to claim 20, wherein determining the sample data sets includes receiving the sample data sets from an external data source and storing the sample data sets in a computer memory.

22. An apparatus according to claim 21, wherein the sample data sets include values of blood pressure for the subjects in the subject group.

23. An apparatus according to claim 20, wherein determining the values for the one or more mathematical parameters includes:

determining initial values for the one or more mathematical parameters according to an optimization criterion; and

separating the initial values into bins with corresponding bin ranges.

24. An apparatus according to claim 20, wherein determining the values for the one or more mathematical parameters includes calculating the values for the one or more mathematical parameters from the sample data sets according to an optimization criterion.

25. An apparatus according to claim 20, wherein determining the one or more distribution-function parameters includes calculating the one or more distribution-function parameters from the values for the one or more mathematical parameters according to an optimization criterion.

26. An apparatus according to claim 20, wherein the computer further includes computer instructions for:

displaying the at least one statistical property of the simulated feature.

27. An apparatus according to claim 20, wherein determining the values for the one or more mathematical parameters corresponding to the one or more basis functions includes:

selecting a plurality of initial basis functions;

determining a plurality of values for a plurality of mathematical parameters corresponding to the initial basis functions;

determining, from the values for the mathematical parameters corresponding to the initial basis functions, a correlation matrix for the initial basis functions; and

determining, from the correlation matrix, the one or more basis functions according to a de-correlation criterion.

28. An apparatus according to claim 20, wherein the feature is a first feature selected from a plurality of features that includes a second feature, and the computer further includes computer instructions for:

determining, from the sample data sets, a plurality of values for one or more second mathematical parameters corresponding to one or more second basis functions for a continuous mathematical model of the second feature; and

determining, from the values for the one or more second mathematical parameters, one or more second distribution-function parameters, wherein

values for the features other than the second feature and values for the one or more second distribution-function parameters specify the one or more second distribution functions for the one or more second mathematical parameters, and

a summation of the one or more second basis functions multiplied by sampled values of the one or more second distribution functions provides the continuous mathematical model of the second feature.

29. An apparatus according to claim 28, wherein the computer further includes computer instructions for:

simulating the second feature by computer, for given values of the features other than the second feature, by generating sampled values of the one or more second distribution functions;

calculating at least one statistical property of the simulated second feature for characterizing the at least one physiological characteristic of the subjects in the subject group; and

saving the at least one statistical property of the simulated second feature.

30. An apparatus according to claim 20, wherein the one or more basis functions include a single basis function and the summation includes a single term.

31. An apparatus according to claim 20, wherein the one or more basis functions include a plurality of orthogonal functions over a continuous interval.

32. An apparatus according to claim 20, wherein the one or more basis functions include one or more hybrid functions that characterize features common to the subject group over a continuous interval.

33. An apparatus according to claim 32, wherein

the continuous mathematical model of the feature includes a model for occlusion of a coronary artery over the continuous interval, and

the one or more hybrid functions include a first function for blood pressure and a second function for cholesterol level.

34. An apparatus according to claim 33, wherein the one or more hybrid functions include a third function for a product of the blood pressure and the cholesterol level.

35. An apparatus according to claim 20, wherein the feature is selected from the group consisting of blood pressure, cholesterol levels, bone mineral density, patency of a coronary artery, heart electrical potentials, contractility of myocardium, cardiac output, visual acuity, serum potassium level, observations for a rash, diameter of a coronary artery, and cancer spread measurements.

36. An apparatus according to claim 20, wherein the one or more distribution functions include at least one normal distribution function and the one or more distribution-function parameters include at least one corresponding standard-deviation parameter.

37. A computer-readable medium that stores a computer program for generating a continuous mathematical model of a feature common to subjects in a subject group, wherein the subjects are biological subjects and the feature is a biological feature, the computer program comprising instructions for:

determining a plurality of sample data sets corresponding to at least one physiological characteristic of the subjects in the subject group;

determining, from the sample data sets, a plurality of values for one or more mathematical parameters corresponding to one or more basis functions for the continuous mathematical model;

determining, from the values for the one or more mathematical parameters, one or more distribution-function parameters for specifying one or more distribution functions for the one or more mathematical parameters, wherein a summation of the one or more basis functions multiplied by sampled values of the one or more distribution functions provides the continuous mathematical model of the feature;

simulating the feature by generating sampled values of the one or more distribution functions by computer;

calculating at least one statistical property of the simulated feature for characterizing the at least one physiological characteristic of the subjects in the subject group; and

saving the at least one statistical property of the simulated feature.

38. A computer-readable medium according to claim 37, wherein determining the sample data sets includes receiving the sample data sets from an external data source and storing the sample data sets in a computer memory.

39. A computer-readable medium according to claim 38, wherein the sample data sets include values of blood pressure for the subjects in the subject group.

40. A computer-readable medium according to claim 37, wherein determining the values for the one or more mathematical parameters includes:

determining initial values for the one or more mathematical parameters according to an optimization criterion; and

separating the initial values into bins with corresponding bin ranges.

41. A computer-readable medium according to claim 37, wherein determining the values for the one or more mathematical parameters includes calculating the values for the one or more mathematical parameters from the sample data sets according to an optimization criterion.

42. A computer-readable medium according to claim 37, wherein determining the one or more distribution-function parameters includes calculating the one or more distribution-function parameters from the values for the one or more mathematical parameters according to an optimization criterion.

43. A computer-readable medium according to claim 37, wherein the computer program further comprises instructions for:

displaying the at least one statistical property of the simulated feature.

44. A computer-readable medium according to claim 37, wherein determining the values for the one or more mathematical parameters corresponding to the one or more basis functions includes:

selecting a plurality of initial basis functions;

determining a plurality of values for a plurality of mathematical parameters corresponding to the initial basis functions;

determining, from the values for the mathematical parameters corresponding to the initial basis functions, a correlation matrix for the initial basis functions; and

determining, from the correlation matrix, the one or more basis functions according to a de-correlation criterion.

45. A computer-readable medium according to claim 37, wherein the feature is a first feature selected from a plurality of features that includes a second feature, and the computer program further comprises instructions for:

determining, from the sample data sets, a plurality of values for one or more second mathematical parameters corresponding to one or more second basis functions for a continuous mathematical model of the second feature; and

determining, from the values for the one or more second mathematical parameters, one or more second distribution-function parameters, wherein

values for the features other than the second feature and values for the one or more second distribution-function parameters specify the one or more second distribution functions for the one or more second mathematical parameters, and

a summation of the one or more second basis functions multiplied by sampled values of the one or more second distribution functions provides the continuous mathematical model of the second feature.

46. A computer-readable medium according to claim 45, wherein the computer program further comprises instructions for:

simulating the second feature by computer, for given values of the features other than the second feature, by generating sampled values of the one or more second distribution functions;

calculating at least one statistical property of the simulated second feature for characterizing the at least one physiological characteristic of the subjects in the subject group; and

saving the at least one statistical property of the simulated second feature.

47. A computer-readable medium according to claim 37, wherein the one or more basis functions include a single basis function and the summation includes a single term.

48. A computer-readable medium according to claim 37, wherein the one or more basis functions include a plurality of orthogonal functions over a continuous interval.

49. A computer-readable medium according to claim 37, wherein the one or more basis functions include one or more hybrid functions that characterize features common to the subject group over a continuous interval.

50. A computer-readable medium according to claim 49, wherein

the continuous mathematical model of the feature includes a model for occlusion of a coronary artery over the continuous interval, and

the one or more hybrid functions include a first function for blood pressure and a second function for cholesterol level.

51. A computer-readable medium according to claim 50, wherein the one or more hybrid functions include a third function for a product of the blood pressure and the cholesterol level.

52. A computer-readable medium according to claim 37, wherein the feature is selected from the group consisting of blood pressure, cholesterol levels, bone mineral density, patency of a coronary artery, heart electrical potentials, contractility of myocardium, cardiac output, visual acuity, serum potassium level, observations for a rash, diameter of a coronary artery, and cancer spread measurements.

53. A computer-readable medium according to claim 37, wherein the one or more distribution functions include at least one normal distribution function and the one or more distribution-function parameters include at least one corresponding standard-deviation parameter.