Automatic recognition of known patterns in physiological measurement data

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.

Description

RELATED APPLICATIONS[0001]The present application is a continuation of and claims priority to co-pending International Application No. PCT / EP2011 / 073084, filed Dec. 16, 2011, which claims priority to European Application No. EP 10 196 379.1, filed Dec. 22, 2010, and U.S. application Ser. No. 12 / 975,654, filed Dec. 22, 2010, the entire disclosures of which being expressly incorporated herein by reference.FIELD OF THE DISCLOSURE[0002]The invention relates to a method and a device for analysing physiological measurement values of a user. Furthermore, the invention relates to a computer program with program code for carrying out a method according to the invention. Such devices, methods and computer programs can be used in general for acquiring and analysing physiological measurement data of a user, for example in long-term monitoring of human or animal users within the scope of so-called home monitoring or else during hospital stays. The method, the device and the computer program can ...

AI Technical Summary

Benefits of technology

[0043]By way of example, data reduction can easily be achieved by modifying a temporal grid, for example by only assigning measurement values to the reduced measurement data record at specific times, for example at specific time intervals. However, within the scope of the present invention, an indexing method is particularly preferred; it will still be explained in more detail in an exemplary fashion below. Within the scope of the present invention, the reduced measurement data record can be stored, in particular in addition to the non-reduced measurement data record. Thus, for example, as explained above, at least one reduced comparison pattern can be generated from the comparison pattern as a result of the data reduction step, for example by a reduced current pattern being generated from the current pattern and / or by at least one reduced pattern of interest being generated from the pattern of interest as a result of the data reduction step. This reduced comparison pattern can be stored in addition to the comparison pattern and can, for example, firstly be used for a coarse search for candidate patterns before, if suitable candidate patterns are identified, the non-reduced comparison pattern can then be used for a refined comparison.

[0046]There can preferably be a data reduction that, in the following text, is also referred to as indexing. In this type of data reduction that can be used as an alternative to other forms of data reduction, or in addition thereto, a plurality of measurement value levels are prescribed in the data reduction step. By way of example, in the case of a blood-glucose measurement these can be concentration levels that can usually occur, for example a blood-glucose mesh in steps of 10 mg / dl or 20 mg / dl. In principle, other types of meshes are also possible. This makes it possible to subject the data in the measurement data record to a temporal grid and / or a measurement-value level grid.

[0093]The proposed method, the computer program and the device have a number of advantages over known methods, computer programs and devices of the aforementioned type. In particular, the invention provides the option of automatic, e.g. retrospective, consideration of measured measurement values, e.g. current glucose measurement values or glucose measurement values of interest, and historical data and / or other data in a measurement data record. Here, particular attention can be given to the earlier profile of this measurement value, for example in order to find a historical situation of the user that is as identical as possible to the current situation of the user, in order for example to give the user the option of reacting in an ideal fashion to the current situation. To this end, use can be made not only of the measurement values and their profiles, but also of other values and / or boundary conditions such as e.g. the insulin bolus, the time of day, stress levels or other events such as e.g. meals, physical activities or physical sensitivities, in order to be able to find and display the ideal historical measurement time. The optionally proposed method of data reduction in particular allows subjecting the measurement data to a temporal grid, which offers a quick option for comparing measurement profiles to one another. Accordingly, it is also easy to implement the proposed method online and preferably as a real-time method, in small hand-held instruments. Furthermore, the method generally does not require any user interaction, in contrast to the described methods from the prior art discussed above. The method can in particular run in real-time, online and, preferably, completely in the background.

[0097]Furthermore, additional information can be provided to the user as to how he can influence possible future developments. This method can assist with reacting in a current situation as a result of a fast reaction time, and can alternatively retrospectively support the evaluation of historical data.

Problems solved by technology

These systems also can rely on lancing and manipulation of the fingers or alternate blood draw sites, which can be extremely painful and inconvenient, particularly for children.

However, a technical challenge consists of the fact that the measurement data record reaches a technical time resolution that is confronted with a huge data volume and hence requires novel methods of data preparation, data aggregation and data reuse.

Otherwise, the increase in the data volume can even lead to a reduction in the user-friendliness of the methods and devices for the user, and to a lacking overview for the treating medical practitioner.

Thus, the known methods in principle are very time-consuming and are possibly too complicated for potential users, in particular for children, elderly patients or patients with dementia.

For example, the assignment of names to particular events, such as naming a specific meal, or a qualification and quantification of certain user events by the user himself is subject to very subjective criteria, and so, possibly, finding a corresponding pattern may not be possible or may even be misleading as a result of initial naming or storing that was not thought through by inexperienced or overburdened patients.

A manual method is not efficient, particularly for large data stocks, as are already expected for a measurement period spanning a couple of days.

Moreover, many methods presuppose smoothing of the generally noisy measurement value profiles.

Furthermore, there are technical challenges in quantifying the similarity of portions.

Moreover, manual methods in principle are time-consuming and generally inefficient.

Furthermore, there has not yet been a satisfactory solution to technical challenges that occur in the processing of found patterns, particularly if a number of possible patterns have been identified.

Images