Correlation between pathological conditions and biometric data

Early Warning Systems (EWS) based on biometric data are used to alert health-care providers and authorities about emerging health problems on the individual level.

Background

Early Warning Systems (EWS) based on biometric data are used to alert health-care providers and authorities about emerging health problems on the individual level. Systems based on several different types of input parameters have recently entered the market.

Goal

Investigations and documentation of the correspondence between physiological processes related to diseases and biometric measurements. Detection of deviations in non-uniformly sampled data with different noise sources.

Problems

• A clinical study needs to be performed in order to establish a correspondence between biometric data and pathological conditions. The first project will look into correlations between blood glucose and infections.
• To be able to distinguish deviations caused by infections from natural changes caused by e.g. diet, a carefully planned design of experiments must be performed. From a statistical point of view, it would be preferable with frequently sampling. This clearly contradicts with the preference for most patients. A reasonable trade-off must therefore be made in the design of the experiment.

Methods

• A randomized control study in collaboration with UNN.
• Time series methods, nonparametric smoothing, scale-space methods will be applied. The problem is related to statistical quality control so many of the existing methods in that area will be utilized.

Project manager

Fred Godtliebsen, UiT

Project members from partners

Fred Godtliebsen (UiT), Vedad Hadziavdic (NST)

Researchers

1 Post.doc researcher (NST?) and 1 PhD student (NST?)

International collaborators

Prof. Lasse Holmström, Department of Mathematics and Statistics, University of Oulo, Finland,
Dr. Jörg Polzehl, Weierstrass Institut für Angewandte Mathematik und Statistikk, Berlin, Germany
Prof. Probal Chaudhuri, Indian Statistical Institute, Calcutta, India
Prof. James Stephen Marron, University of North Carolina at Chapel Hill, Chapel Hill, USA.

Project start/stop

1.2.07 – 31.12.10

Contribution to health care

Detection and predictions of diseases on individual level, better insulin dosage prediction.

Contribution to new industry

Products based on the above-mentioned technology and glucose measurement devices. There is a potential for both new hardware product and new services that can be offered by both health institutions and hardware vendors.

Contact info

e-mail: Fred Godtliebsen, phone: +47 776 44 019

E-mail this

Print this