System for predicting future health

• US 6,059,724 A
• Filed: 02/14/1997
• Issued: 05/09/2000
• Est. Priority Date: 02/14/1997
• Status: Expired due to Term

• Current Assignee: Biosignia, Inc.
• Original Assignee: Biosignal Ltd. (One North Entertainment Ltd.)
• Inventors: Helms, Ronald W., Campell, T. Colin, Tomasko, Lisa
• Primary Examiner(s): WINAKUR, ERIC FRANK
• Application Number:
  US08/800,314
• Time in Patent Office: 1,180 Days
• Field of Search: 600/300, 600/301, 128/920, 128/923-925
• US Class Current: 600/300
• CPC Class Codes: A61B 5/00 Measuring for diagnostic purposes radiation diagnosis A61B6/00; diagnosis by ultrasonic, sonic or infrasonic waves A61B8/00; Identification of persons
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass
• CPC Class Codes: Y10S 128/923 by comparison of patient data to other data

Abstract

A computer-based system is disclosed for predicting future health of individuals comprising:

(a) a computer comprising a processor containing a database of longitudinally-acquired biomarker values from individual members of a test population, subpopulation D of aid members being identified as having acquired a specified biological condition within a specified time period or age interval and a subpopulation D being identified as not having acquired the specified biological condition within the specified time period or age interval; and

(b) a computer program that includes steps for:

(1) selecting from said biomarkers a subset of biomarkers for discriminating between members belonging to the subpopulations D and D, wherein the subset of biomarkers is selected based on distributions of the biomarker values of the individual members of the test population; and

(2) using the distributions of the selected biomarkers to develop a statistical procedure that is capable of being used for:

(i) classifying members of the test population as belonging within a subpopulation PD having a prescribed high probability of acquiring the specified biological condition within the specified time period or age interval or as belonging within a subpopulation PD having a prescribed low probability of acquiring the specified biological condition within the specified time period or age interval; or

(ii) estimating quantitatively, for each member of the test population, the probability of acquiring the specified biological condition within the specified time period or age interval.

Claims

1. A computer-based system for predicting future health of individuals comprising:

(a) a computer comprising a processor containing a database of longitudinally-acquired biomarker values from individual members of a test population, subpopulation D of said members being identified as having acquired a specified biological condition within a specified time period or age interval and a subpopulation D being identified as not having acquired the specified biological condition within the specified time period or age interval; and

(b) a computer program that includes steps for;

(1) selecting from said biomarkers a subset of biomarkers for discriminating between members belonging to the subpopulations D and D, wherein the subset of biomarkers is selected based on distributions of the biomarker values of the individual members of the test population; and

(2) using the distributions of the selected biomarkers to develop a statistical procedure that is capable of being used for;

(i) classifying members of the test population as belonging within a subpopulation PD having a prescribed high probability of acquiring the specified biological condition within the specified time period or age interval or as belonging within a subpopulation PD having a prescribed low probability of acquiring the specified biological condition within the specified time period or age interval;

or

(ii) estimating quantitatively, for a member of the test population, the probability of acquiring the specified biological condition within the specified time period or age internal;

wherein changes in said longitudinally-acquired biomarker values are used to develop said statistical procedure.

2. The computer-based system of claim 1 wherein the statistical procedure comprises a discriminant function utilizing estimated mean vectors and estimated covariance matrices of the distributions of biomarker values within the subpopulations D and D.

3. The computer-based system of claim 2 wherein estimates of parameters of the distributions of the selected biomarkers are obtained by fitting a general linear mixed model to the biomarker data from the test population.

4. The computer-based system of claim 2 wherein:

(a) the estimated mean vectors are modeled as vector-valued functions of expected-value parameters or values of covariates;

or

(b) the estimated covariance matrices are modeled as matrix-valued functions of covariance parameters or values of covariates.

5. The computer-based system of claim 4 wherein estimates of parameters of the distributions of the selected biomarkers are obtained by fitting a general linear mixed model to the biomarker data from the test population.

6. The computer-based system of claim 4 wherein an estimated mean vector or probability incorporates an estimate of the realized value of a random subject effect vector for a member being classified or of a member for whom a probability is estimated.

7. The computer-based system of claim 6 wherein estimates of parameters of the distributions of the selected biomarkers are obtained by fitting a general linear mixed model to the biomarker data from the test population.

8. A method of predicting an individual'"'"'s health comprising:

collecting a plurality of biomarker values from an individual, wherein at least one of said biomarker values is obtained by physically measuring the biomarker value; and

applying a statistical procedure to said plurality of biomarker values so as;

(i) to classify said individual as having a prescribed high probability of acquiring a specified biological condition within a specified time period or age interval or as having a prescribed low probability of acquiring the specified biological condition within the specified time period or age interval;

or

(ii) to estimate quantitatively for said individual the probability of acquiring the specified biological condition within the specified time period or age interval;

wherein said statistical procedure is based on;

(1) collecting a database of longitudinally-acquired biomarker values from individual members of a test population, subpopulation D of said members being identified as having acquired the specified biological condition within the specified time period or age interval and a subpopulation D being identified as not having acquired the specified biological condition within the specified time period or age interval;

(2) selecting from said biomarkers a subset of biomarkers for discriminating between members belonging to the subpopulations D and D, wherein the subset of biomarker is selected based on distributions of the biomarker values of the individual members of the test population; and

(3) using the distributions of the selected biomarkers to develop said statistical procedure;

wherein changes in said longitudinally-acquired biomarkers values are used to develop said statistical procedure.

9. The method according to claim 8 wherein at least one of said biomarker values is obtained from a biological sample.

10. The method according to claim 9 wherein said biological sample is a serum or urine sample.

11. A computer-based system for predicting an individual'"'"'s future health comprising:

(a) a computer comprising a processor containing a plurality of biomarker values from an individual; and

(b) a computer program that includes steps for applying a statistical procedure to said plurality of biomarker values so as;

(i) to classify said individual as having a prescribed high probability of acquiring a specified biological condition within a specified time period or age interval or as having a prescribed low probability of acquiring the specified biological condition within the specified time period or age interval;

or

(ii) to estimate quantitatively for said individual the probability of acquiring the specified biological condition within the specified time period or age interval;

wherein said statistical procedure is based on;

(1) collecting a database of longitudinally-acquired biomarker values from individual members of a test population, subpopulation D of said members being identified as having acquired the specified biological condition within the specified time period or age interval and a subpopulation D being identified as not having acquired the specified biological condition within the specified time period or age interval;

(2) selecting from said biomarkers a subset of biomarkers for discriminating between members belonging to the subpopulations D and D, wherein the subset of biomarkers is selected based on distributions of the biomarker values of the individual members of the test population; and

(3) using the distributions of the selected biomarkers to develop said statistical procedure;

wherein changes in said longitudinally-acquired biomarker values are used to develop said statistical procedure.

12. The computer-based system of claim 11 wherein the plurality of biomarker values from said individual includes longitudinally-acquired biomarker values.

13. The computer-based system of claim 11 wherein the specified biological condition is death due to a specified underlying cause of death within the specified time period or age interval.

14. The computer-based system of claim 11 wherein the specified biological condition is a specified morbidity within the specified time period or age interval.

15. The computer-based system of claim 11 wherein the specified time period is a period of at least two years.

16. The computer-based system of claim 11 wherein the specified time period is a period of at least three years.

17. A method for assessing an individual'"'"'s future risk of death from specified underlying causes of death comprising:

collecting a plurality of biomarker values from an individual, wherein at least one of said biomarker values is obtained by physically measuring the biomarker value; and

applying a statistical procedure to said plurality of biomarker values so as to determine whether said individual is classified as having a prescribed high probability of dying, within a specified time period or age interval, from any one of the underlying causes of death that account in the aggregate for at least 60% of all deaths in a test population over the specified time period or age interval.

18. A method for assessing an individual'"'"'s evidence of good health comprising:

collecting a plurality of biomarker values from an individual, wherein at least one of said biomarker values is obtained by physically measuring the biomarker value; and

applying a statistical procedure to said plurality of biomarker values so as to determine whether said individual is classified as having a prescribed high probability of not dying, within a specified time period or age interval, from any one of the underlying causes of death that account in the aggregate for at least 60% of all deaths in a test population over the specified time period or age interval.

19. A computer-based system for assessing an individual'"'"'s future risk of death from a specified underlying cause of death comprising:

(a) a computer comprising a processor containing a plurality of biomarker values from an individual; and

(b) a computer program that includes steps for applying a statistical procedure to said plurality of biomarker values so as to determine whether said individual is classified as having a prescribed high probability of dying, within a specified time period or age interval, from any one of the underlying causes of death that account in the aggregate for at least 60% of all deaths in a test population over the specified time period or age interval.

20. A computer-based system for assessing an individual'"'"'s evidence of good health comprising:

(a) a computer comprising a processor containing a plurality of biomarker values from an individual; and

(b) a computer program that includes steps for applying a statistical procedure to said plurality of biomarker values so as to determine whether said individual is classified as having a prescribed high probability of not dying, within a specified time period or age interval, from any one of the underlying causes of death that account in the aggregate for at least 60% of all deaths in a test population over the specified time period or age interval.

21. An apparatus for assessing an individual'"'"'s risk of future health problems comprising:

(a) a stooge device for storing a plurality of biomarker values from an individual; and

(b) a processor coupled to the storage device and programmed;

1) to receive from the storage device said plurality of biomarker values; and

2) to apply a statistical procedure to said plurality of biomarker values so as;

(i) to classify said individual as belonging within a subpopulation PD having a prescribed high probability of acquiring a specified biological condition within a specified time period or age interval or as belonging within a subpopulation PD having a prescribed low probability of acquiring the specified biological condition within the specified time period or age interval;

or

(ii) to estimate quantitatively the probability for said individual acquiring the specified biological condition within the specified time period or age interval;

wherein said statistical procedure is based on;

(1) collecting a database of longitudinally-acquired biomarker values from individual members of a test population, subpopulation D of said members being identified as having acquired the specified biological condition within the specified time period or age interval and a subpopulation D being identified as not having acquired the specified biological condition within the specified time period or age interval;

(2) selecting from said biomarkers a subset of biomarkers for discriminating between members belonging to the subpopulations D and D, wherein the subset of biomarkers is selected based on distributions of the biomarker values of the individual members of the test population; and

(3) using the distributions of the selected biomarkers to develop said statistical procedure;

wherein changes in said longitudinally-acquired biomarker values are used to develop said statistical procedure.

22. A method of predicting an individual'"'"'s health comprising:

collecting a plurality of biomarker values from an individual, wherein at least one of said biomarker values is obtained by physically measuring the biomarker value; and

applying a multivariate statistical procedure to said plurality of biomarker values so as to provide a statistically-based quantitative probability of said individual acquiring a specified morbidity or suffering a mortality with a specified underlying cause of death within a specified time period or age interval, wherein said multivariate statistical procedure is based on comparing said individual'"'"'s plurality of biomarker values with a longitudinally-acquired database, wherein said database includes biomarker values from individual members of a test population, a subpopulation D of said members being identified as having acquired the specified morbidity or suffering the mortality with the specified underlying cause of death within the specified time period or age interval and a subpopulation D being identified as not having acquired the specified morbidity or suffering the mortality with the specified underlying cause of death within the specified time period or age interval;

wherein changes in the longitudinally-acquired biomarker values to develop said statistical procedure.

23. The method according to claim 22 wherein said method includes the step of selecting a subset of biomarkers having discriminatory power for discriminating between said members identified as having acquired the specified biological condition within the specified time period or age interval and said members identified as not having acquired the specified biological condition within the specified time period or age interval.

24. The method according to claim 23 wherein the biomarkers in said subset provide a statistically significant contribution to the discriminatory power of said subset of biomarkers when used in combination with the other biomarkers in said subset.

25. The method according to claim 24 wherein said statistically significant contribution to the discriminatory power is established based on selecting the biomarkers in said subset by evaluating distributions of said longitudinally-acquired biomarker values.

26. The method according to claim 24 wherein said statistically significant contribution to the discriminatory power is established based on evaluating ratios or longitudinally-observed changes in the ratios of an individual'"'"'s biomarker values.

27. The method according to claim 26 wherein the evaluation step comprises assessing whether longitudinal trends in said longitudinally-acquired biomarkers values are statistically correlated with acquisition of the specified biological condition.

28. The method according to claim 22 wherein said database comprises longitudinally-acquired biomarker values and at least one of said longitudinally-acquired biomarker values is obtained by physically measuring the biomarker value.

29. A computer-based system for predicting the future health of individuals comprising:

(a) a computer comprising a processor containing;

(i) a plurality of biomarker values from an individual; and

(ii) a longitudinally-acquired database, wherein said database includes biomarker values from individual members of a test population, subpopulation D of said members being identified as having acquired a specified morbidity or suffering a mortality with a specified underlying cause of death within a specified time period or age interval and a subpopulation D being identified as not having acquired the specified morbidity or suffering a mortality with the specified underlying cause of death within the specified time period or age interval; and

(b) a computer program that includes steps for applying a multivariate statistical procedure to compare said individual'"'"'s plurality of biomarker values with said longitudinally-acquired database so as to provide a statistically-based quantitative probability for said individual acquiring the specified morbidity or suffering a mortality with the specified underlying cause of death within the specified time period or age interval;

wherein changes in the longitudinally-acquired biomarker values are used to develop said statistical procedure.

30. An apparatus for assessing an individuals'"'"'s an risk off future health problems comprising:

(a) a storage device for storing a plurality of biomarker values from an individual; and

(b) a processor coupled to said storage device, said processor being programmed to receive from said storage device said plurality of biomarker values and apply a multivariate statistical procedure comparing said plurality of biomarker values with a longitudinally-acquired database, wherein said database includes biomarker values from individual members of a test population, subpopulation D of said members being identified as having acquired a specified morbidity or suffering a mortality with a specified underlying cause of death within a specified time period or age interval, and a subpopulation D being identified as not having acquired the specified morbidity or suffering a mortality with the specified underlying cause of death within the specified time period or age interval, so as to provide a statistically-based quantitative probability for said individual acquiring the specified morbidity or suffering a mortality with the specified underlying cause of death within the specified time period or age interval;

wherein changes in the longitudinally-acquired biomarker values are used to develop said statistical procedure.