ROC method for early detection of disease

• US 6,030,341 A
• Filed: 08/24/1998
• Issued: 02/29/2000
• Est. Priority Date: 09/12/1995
• Status: Expired due to Term

• Current Assignee: Queen Mary And Westfield College
• Original Assignee: The General Hospital Corporation (Mass General Brigham, Inc.)
• Inventors: Knapp, Robert, Jacobs, Ian, Skates, Steven J.
• Primary Examiner(s): Jaworski, Francis J.
• Application Number:
  US09/139,088
• Time in Patent Office: 554 Days
• Field of Search: 600/300, 705/2-3, 395/22, 128/920-924
• 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 5/7275 Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
• CPC Class Codes: G01N 33/57449 of ovaries
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment
• CPC Class Codes: Y10S 128/923 by comparison of patient data to other data

Abstract

A method of assessing risk of an individual for a disease is disclosed. A level of a marker for the disease is detected, and the risk of disease is computed from a statistical analysis of the marker level distributions for tested normal and diseased populations, using multivariate distributions. The computed risk is compared to thresholds to triage the individual into a normal, borderline, or elevated risk group, and a course of action based on the risk group is determined. Memory storing a program for execution on a data processing system, as well as a computer system comprising said memory is also disclosed, the memory comprising a receiver for signals representative of levels of a marker for the disease; a central processing unit for computing risk of disease from one or more levels of the marker, and for comparing the computed risk to thresholds to triage the individual; and storage medium for the marker levels for subsequent computations of risk of disease.

Claims

1. Memory storing a program for execution on a data processing system comprising:

means for receiving signals representative of levels of a marker for disease from an individual;

means for computing a risk of disease from one or more levels of a marker;

means for comparing the computed risk to thresholds to triage the individual into one of normal, borderline and elevated risk groups; and

means for storing the marker levels for subsequent computations of risk of disease.

2. Memory as claimed in claim 1 further comprising means for outputting a test interval value which is inversely related to the computed risk, to identify the interval at which an individual should again be test to determine the level of the marker to disease.

3. Memory as claimed in claim 2 wherein the risk of disease is again computed using a marker level obtained at the test interval.

4. Memory as claimed in claim 1 further comprising means for creating a bar graph display indicating the computed risk, the display being green when the risk is less than the normal threshold, the display being yellow when the risk is between the borderline thresholds, and the display being red when the risk is greater than the elevated threshold.

5. Memory as claimed in claim 1 wherein risk of disease is computed form a statistical analysis of marker level distributions for tested normal and diseased populations.

6. Memory as claimed in claim 5 wherein risk of disease is computed from multivariate distributions.

7. Memory as claimed in claim 6 wherein the marker level changes exponentially over time for diseased individuals and the multivariate distributions are determined from the slope and intercept of the log of detected marker levels over time for individuals within normal and diseased populations.

8. Memory as claimed in claim 7 wherein random draws are used to compute average densities from the normal and disease populations given the marker level for the individual.

9. Memory as claimed in claim 8 wherein risk of disease is computed as a product of an odds factor and an estimate of prior odds, wherein the odds factor is the ratio of the average density from the disease population to the average density from the normal population.

10. Memory as claimed in claim 8 wherein a time correlation component is included in a linear regression model of marker levels to compute risk of disease.

11. A computer comprising:

an input device for receiving marker levels;

a display for displaying risk of disease and courses of action, based upon the risk of disease; and

memory comprising means for receiving signals representative of levels of a marker for disease;

means for computing risk of disease from one or more levels of the marker;

means for comparing the computed risk to thresholds to triage the individual into one of normal, borderline and elevated risk groups; and

means for storing the marker levels for subsequent computations of risk of disease.

12. A computer system as claimed in claim 11, wherein the display is a bar graph display indicating the computed risk, the bar graph display being green when the risk is less than the normal threshold, the bar graph display being yellow when the risk is between the borderline thresholds, and the bar graph display being red when the risk is greater than the elevated threshold.

13. A computer system as claimed in claim 11 wherein the risk of disease is again computed using a marker level obtained at the test interval.

14. A computer system as claimed in claim 11 wherein risk of disease is computed from a statistical analysis of marker level distributions for tested normal and diseased populations.

15. A computer system as claimed in claim 14 wherein risk of disease is computed from multivariate distributions.

16. A computer system as claimed in claim 15 wherein the marker level changes exponentially over time for diseased individuals and the multivariate distributions are determined from the slope and intercept of the log of detected marker levels over time for individuals within normal and diseased populations.

17. A computer system as claimed in claim 16 wherein random draws are used to compute average densities from the normal and disease populations given the marker level for the individual.

18. A computer system as claimed in claim 17 wherein risk of disease is computed as a product of an odds factor and an estimate of prior odds, wherein the odds factor is the ratio of the average density from the disease population to the average density from the normal population.

19. A computer system as claimed in claim 17 wherein a time correlation component is included in a linear regression model of marker levels to compute risk of disease.