Singular Spectrum Analysis of pupillometry data. Identification of the sympathetic and parasympathetic activity

Abstract

The pupil shows a fast dynamical adaptation to external light changes, in order to potentiate visual perception. It contracts in high luminosity environments, so as to limit the quantity of photons that reach the retina (miosis) and dilates in environments with little luminosity, thus increasing the number of photons in the retina (mydriasis). Hence, the pupil acts as a normalizer of the quantity of light that enters the optical system. These pupillary movements are controlled by the action of smooth muscles, which are associated with the sympathetic (SNS) and parasympathetic autonomous nervous systems (PsNS). Therefore, the study of the pupillary dynamics can lead to inferences on the functioning of those two antagonist systems and their involvement in pupillary control. In addition, our study may prove clinically helpful in determining a patient's level of consciousness, as well as identifying certain diseases, such as sleep disorders, photophobia, schizophrenia, Adie syndrome, Alzheimer's and Parkinson's disease. With a clear goal in its clinical applications, we focus our research in the identification of the frequency intervals pertaining to the action of the autonomous nervous system, as well as their individual sympathetic and parasympathetic roles. We analyze high-resolution, non-invasive pupillometric signals, both in a basal condition and as a response to visual flashes or to the application of a cold stimulus. We use Singular Spectrum Analysis (SSA) to identify the frequencies of interest related to the actions of the SNS and PsNS. In addition, and as a complementary means of frequency analysis, we also use the more classical wavelets analysis. We find out that SSA is an ideal tool for the identification of the desired frequencies in stake. The agreement between the outcomes of both analyses supports our conclusions regarding the autonomous nervous system. (C) 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).