Introduction

During my 2023 summer internship at NIMHANS, I explored computational analysis of sleep EEG. The objective was to investigate the trend of aperiodic brain activity across different sleep stages in meditators using the FOOOF algorithm.

Background:

Neural power spectra of sleep electrophysiological signals consist of two major components: an aperiodic component represented by a 1/f like slope and a periodic component represented by oscillations appearing as bumps above the 1/f like slope. Brain activity corresponding to the ‘1/f’ slope has been deemed unimportant for a long time and often removed from analyses to emphasise brain oscillations. Emerging evidence suggests that aperiodic brain activity contributes actively to brain functioning.

Project Objectives:

In this study, extracted aperiodic parameters were analysed for sleep stage sensitivity as well as brain region interaction. The study investigates changes in aperiodic parameters in meditators during sleep, providing valuable insights into the interactions between meditation and sleep in shaping brain electrodynamics. The study also investigates whether aperiodic spectral parameters derived from FOOOF analysis of EEG data hold promise as objective measures for characterising sleep states.

Methodology:

To achieve the project goals, I employed the FOOOF (Fitting Oscillations and One-Over F) algorithm, a powerful computational tool developed to model the power spectrum as a combination of an aperiodic component (1/f slope) and periodic component (peaks over the 1/f slope). The algorithm was applied to a set of annotated whole-night sleep EEG recordings of 45 subjects (24 meditating subjects and 21 non-meditating subjects). All data processing and analyses were carried out using Python programming language, leveraging its rich ecosystem of libraries for scientific computing and data visualization.

Results:

The computational analyses revealed distinct features of aperiodic brain activity across different sleep stages, enhancing our understanding of brain dynamics during sleep. My work suggests that aperiodic spectral parameters derived from FOOOF analysis of EEG data hold promise as objective measures for characterizing sleep states. Subtle yet significant differences in aperiodic measures were observed between meditators and controls during sleep. Further research could explore the potential influence of meditation practices on sleep physiology and aid in the development of targeted interventions.

Conclusion:

My internship experience at NIMHANS provided me with invaluable exposure to sleep research, computational neuroscience, and data analysis. The exploration of aperiodic brain activity during sleep has the potential to enrich our understanding of the brain’s complex organization and may contribute to future advancements in sleep medicine and cognitive neuroscience.

Acknowledgments:

I express my sincere gratitude to my mentors and researchers at NIMHANS for their guidance, encouragement, and support throughout this internship. Their expertise and insights have been instrumental in shaping the course of my work.