Analysis of skin conductance response

With the present project we want to develop a new method to analyze skin conductance data, which accounts for its nature as a composite of tonic and phasic activation. Skin conductance data are known to show a typical pattern of partially overlapping phasic responses, set on a slowly changing background. The phasic responses follow a characteristic profile. Given a good estimate of the typical template of the phasic response it should be possible to decompose the raw skin conductance data so as to describe them as a superposition of several overlapping phasic responses plus some background. This analysis can be seen as a type of data reduction, whereby the activity in a given time window can be described by a few numbers, namely the onset times and the amplitudes of the phasic responses that occurred within this time window. The course of a single skin conductance response is well described by a multi-exponential function (e.g., Bateman function) deduced from a multi-compartment model. A decomposition method based on these functions shall be tested and optimized for reliably extracting precise activation information out of raw EDA data. The power of the method shall be evaluated systematically using computer-simulations and real data with varying stimulus intensity and inter-stimulus interval. It shall be validated with different ANS measures and compared to commonly used analysis methods. Finally, a powerful software for decomposition of EDA data shall be provided for scientific purposes.

With the present project we want to develop a new method to analyze skin conductance data, which accounts for its nature as a composite of tonic and phasic activation. Skin conductance data are known to show a typical pattern of partially overlapping phasic responses, set on a slowly changing background. The phasic responses follow a characteristic profile. Given a good estimate of the typical template of the phasic response it should be possible to decompose the raw skin conductance data so as to describe them as a superposition of several overlapping phasic responses plus some background. This analysis can be seen as a type of data reduction, whereby the activity in a given time window can be described by a few numbers, namely the onset times and the amplitudes of the phasic responses that occurred within this time window. The course of a single skin conductance response is well described by a multi-exponential function (e.g., Bateman function) deduced from a multi-compartment model. A decomposition method based on these functions shall be tested and optimized for reliably extracting precise activation information out of raw EDA data. The power of the method shall be evaluated systematically using computer-simulations and real data with varying stimulus intensity and inter-stimulus interval. It shall be validated with different ANS measures and compared to commonly used analysis methods. Finally, a powerful software for decomposition of EDA data shall be provided for scientific purposes.