Vous devez pour visualiser cette page disposer de la version 8 de Macromedia Flash Player, vous pouvez la récupérer en cliquant ici

language

Patent 2 : analyzing heart rate variability

Method and device for evaluating pain and stress in a living being

The device and the method for processing an RR series are both mobilized to analyze heart rate variability and to use resulting data in order to assess the degree of pain or of stress in a living being.

What patent n°2 brings

Lille CHRU’s patent n°2 provides an automatic method for processing a RR series that solves all the problems that have just been described beside. It allows to calculate quantitative information (the parameter) that reflects the activity of the nervous system with very good sensitivity..

Most previous methods for measuring and analyzing heart rate variability used Fourier spectral analysis. However, spectral analysis has many drawbacks: firstly, an RR series is not stationary. Thus, applying a fast Fourier transform to this type of series yields inacurrate or even erroneous results, which cannot be interpreted with absolute certainty. Besides, calculating the spectral density curve using a fast Fourier transform requires a great deal of time and computing power. It is thus difficult to implement this method in real time, particularly in embedded systems.

Also, to obtain an acceptable frequential resolution, the fast Fourier transform must be calculated in relatively large timeframes (256 seconds for example), which means resorting to a great number of RR series samples. As a consequence, this method, which is based on spectral analysis, is hampered by a “memory effect” that delays the detection of a change in the RR series.

That is why most scientists currently consider that this method is not satisactory for calculating a stress parameter, and that it cannot successfully be applied to a non-stationary signal (such as the tachogram).