AUTOMATIC RECOGNITION OF KNOWN PATTERNS IN PHYSIOLOGICAL MEASUREMENT DATA

• US 20130338453A1
• Filed: 06/20/2013
• Published: 12/19/2013
• Est. Priority Date: 12/22/2010
• Status: Active Grant

• Current Assignee: Roche Diagnostics GmbH (Roche Holding AG)
• Original Assignee: Roche Diagnostics GmbH (Roche Holding AG)
• Inventors: DUKE, David, SONI, Abhishek S., STEIGER, Bernd, RASCH-MENGES, Jrgen, BROSSART, Michael
• Granted Patent: US 10,575,791 B2
• Time in Patent Office: Days
• US Class Current: 600/309
• CPC Class Codes: A61B 5/14532 for measuring glucose, e.g. by tissue impedance measurement
• CPC Class Codes: A61B 5/486 Bio-feedback A61B5/375 takes precedence
• CPC Class Codes: A61B 5/7246 using correlation, e.g. template matching or determination of similarity
• CPC Class Codes: A61B 5/7282 Event detection, e.g. detecting unique waveforms indicative of a medical condition A61B5/4094, A61B5/4818, A61B5/349, A61B5/0823 take precedence
• CPC Class Codes: A61B 5/742 using visual displays A61B5/339 takes precedence
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

A method for analysing physiological measurement values of a user is proposed. The method comprises at least one data acquisition step, wherein, during the data acquisition step, physiological measurement values of the user are acquired at different measurement times and stored in a measurement data record; at least one pattern selection step, wherein, during the pattern selection step, measurement values acquired during one comparison time interval are selected as at least one comparison pattern; and at least one pattern recognition step, wherein, during the pattern recognition step, patterns corresponding to the comparison pattern are sought after in the measurement data record.

Claims

1. A method for analysing physiological measurement values of a user, comprising:

a) at least one data acquisition step, wherein, during the data acquisition step, physiological measurement values of the user are acquired at different measurement times and stored in a measurement data record;

b) at least one pattern selection step, wherein, during the pattern selection step, measurement values acquired during at least one comparison time interval are selected as at least one comparison pattern;

c) at least one pattern recognition step, wherein, during the pattern recognition step, patterns corresponding to the comparison pattern are sought after in the measurement data record; and

d) at least one data reduction step generating a reduced measurement data record from the measurement data record, wherein the reduced measurement data record is used in at least one of;

the pattern selection step b) and the pattern recognition step c).

2. The method according to claim 1, wherein:

the comparison time interval comprises a current time interval, and

the comparison pattern comprises a current pattern, wherein, during the pattern recognition step, patterns, acquired before the comparison time interval and corresponding to the current pattern, are sought after.

3. The method according to claim 1, wherein:

the comparison time interval comprises a selectable time interval, and

the comparison pattern comprises a pattern of interest.

4. The method according to claim 1, wherein:

a plurality of measurement value levels are prescribed in the data reduction step, wherein it is identified when a measurement value level is crossed between two measurement values adjacent in time, and wherein the measurement value of the adjacent measurement values that is closer to the measurement value level is assigned to the reduced measurement data record.

5. The method according to claim 4, wherein the comparison pattern is reduced to a reduced comparison pattern in the data reduction step and at least part of the pattern recognition step is carried out using the reduced comparison pattern.

6. The method according to claim 1, wherein one or more boundary conditions are stored in addition to the measurement values, the boundary conditions characterizing events that have a potential physiological influence on a body of the user, and wherein the boundary conditions are taken into account during the pattern recognition step.

7. The method according to claim 1, wherein, during the pattern selection step, at least one measurement value acquired at an anchor time is specified as an anchor value.

8. The method according to claim 7, wherein candidate values recorded at candidate times are sought after in the measurement data record during the pattern recognition step, wherein the candidate values correspond to the anchor value.

9. The method according to claim 1, wherein:

a correspondence is quantified if at least one pattern that corresponds to the comparison pattern is found,

the quantification of the correspondence relates to the candidate values and the measurement data record including a bias being substantially constant over time, or

the quantification of the correspondence relates to a dynamic representation of the candidate values and a dynamic representation of the measurement data record excluding a bias, the bias representing the absolute level of the candidate values and the measurement data record.

10. The method according to claim 1, further comprising:

e) at least one interaction step, wherein, during the interaction step, at least one result of the pattern recognition step is one of processed, flagged, and both processed and flagged, to the user.

11. The method according to claim 10, wherein, during the interaction step, a possible future profile of the measurement values is one of established, displayed, and both established and displayed, by means of at least one display element.

12. The method according to claim 10, wherein, during the interaction step, one or more patterns, identified during the pattern recognition step and corresponding to the comparison pattern, are subjected to an analysis.

13. The method according to claim 10, wherein, during the interaction step, at least one instruction for the user is generated.

14. The method according to claim 1, wherein steps b) and c) are carried out repeatedly, wherein a plurality of comparison patterns of the measurement data record are selected and wherein at least one pattern recognition step is respectively carried out using the selected comparison patterns.

15. The method according to claim 14, wherein corresponding patterns established during the pattern recognition steps are combined into groups of respectively corresponding patterns.

16. A computer program with program code for carrying out the method according to claim 1 when the program is executed on a device selected from the group consisting of a computer, a computer network and a processor.

18. The method according to claim 1, wherein the method is used to identify the presence of a physiological body state of the user.

19. The method according to claim 7, wherein the at least one measurement value is selected from the group consisting of a current measurement value and a measurement value of interest.

20. The method according to claim 7, wherein the anchor time is selected from the group consisting of a current measurement time and a measurement time of interest.

21. The method according to claim 8, wherein at least one time interval, which, at an established candidate time, behaves like the comparison time interval at the anchor time, is determined as candidate time interval.

22. The method according to claim 21, wherein the measurement values acquired during the candidate time interval are identified as a candidate pattern.

23. The method according to claim 22, wherein the candidate pattern is compared to the comparison pattern.

24. The method according to claim 9, wherein a correspondence is quantified by means of at least one correlation.

25. The method according to claim 10, wherein at least one pattern, identified during the pattern recognition step and corresponding to the comparison pattern, is displayed by means of at least one display element during the interaction step.

26. The method according to claim 12, wherein the analysis is a statistical analysis.

27. The method according to claim 26, wherein the statistical analysis is performed under formation of an averaged pattern or an error corridor.

28. The method according to claim 10, wherein method steps b), c) and d) are carried out repeatedly.

29. The method according to claim 14, wherein a plurality of comparison patterns of the measurement data record are selected in succession.

30. The method according to claim 15, wherein the groups of corresponding patterns are subjected to at least one analysis.

31. The method according to claim 30, wherein the groups of corresponding patterns are subjected to a statistical analysis.

32. The method according to claim 30, wherein in the statistical analysis is performed by comparing the found groups.

17. A device for analyzing physiological measurement values of a user, comprising:

a) at least one data acquisition device for acquiring physiological measurement values of the user and storing the values in a measurement data record;

b) at least one pattern selection device, wherein the pattern selection device is configured to select measurement values acquired during at least one comparison time interval as at least one comparison pattern; and

c) at least one pattern recognition device, wherein the pattern recognition device is configured to search for patterns corresponding to the comparison pattern in the measurement data record;

wherein the device for analyzing physiological measurement values of a user is adapted for carrying out at least one data reduction step generating a reduced measurement data record from the measurement data record, and wherein the reduced measurement data record is used by at least one of the at least one pattern selection device and the at least one pattern recognition device.