Separation of Electrodermal Activity during Short Interstimulus Interval based on EMD Algorithm

The Electro-dermal activity (EDA) is a property of human skin that was discovered by Fere in 1988. It is a manifestation of the activities in eccrine sweat gland that are innervated by the sympathetic branch of the autonomic nervous system and especially by the sudomotor nerves. When the interstimulus interval (ISI) is considered short, compared to the duration of the first response, the two recorded EDAs unfortunately overlap. The overlapping phenomena present several drawbacks for the investigation of these responses. Consequently, robust and accurate methodologies and algorithms to decompose and analyze the EDA responses are required. Subsequently, the EDA can be broadly divided into skin potential activity (SPA) and skin conductance activity (SCA). This paper presents a new method to separate the EDA overlapping during a short ISI. The proposed method is based on the Empirical Mode Decomposition (EMD) algorithm. We propose to consider only the first extracted intrinsic mode functions (IMF) because they present the highest frequency band and they are rich in information. The other IMFs are omitted and they are considered as useless.