Sleep spindles - one of the major types of brain oscillations during sleep - were first described by a physicist and polymath Alfred Loomis nearly a hundred years ago, and ever since remain a topic of great interest and intense investigation. However, while numerous roles have been proposed for sleep spindles – from memory consolidation to sensory disconnection, functional and mechanistic insights provided thus far are surprisingly modest, and often remain speculative. One likely explanation is that the traditional approach of defining and detecting spindles in the EEG is suboptimal, and conventional analyses do not allow us to take into account some essential elements of spindle-activity. Our work provides a novel and original solution to this fundamental problem.
At the core of our approach is an algorithm that detects oscillatory events in brain signals based on their damping, a measurement used to parameterize oscillatory strength. All previous studies focused merely on quantitative measurements of spindle activity (eg detections based on a fixed amplitude threshold to determine incidence) but did not take into account the oscillatory strength of individual spindle events. We took inspiration from engineering and physics, where the level of damping in oscillatory systems is parameterized in terms of a so-called Quality-Factor. Analogous to this metric, we defined an index derived from the damping level (ie oscillatory strength) of spindle oscillations, which we refer to as oscillatory-Quality (oQ).
We then went on to comprehensively characterize spindle activity using this new parameter across a variety of experimental approaches and settings – from its detailed brain topography and its intracortical dynamics in space and time, to transgenic mouse models relevant to impaired synaptic plasticity in schizophrenia and real-time sensory stimulation.
The accepted general approach to define and detect sleep oscillations is deficient, as it ignores the key physical property of any oscillatory activity - its oscillatory strength. The key far-reaching conclusion we make from our study is that it is spindle quality, or strength, that matters for understanding the origin and function of spindles, rather than their quantity, and oscillatory-Quality of sleep spindles provides the necessary metric for that. We were struck by how sensitive and informative our new metric is – from reflecting local and global synchronisation of spindle events to predicting the probability of arousal in response to sensory stimulation or preceding sleep-wake history. Using this approach, we do not only make an important advance towards clarifying the molecular and network mechanisms underlying the generation and propagation of spindles, but also offer new insights into the causes of spindle deficits in neuropsychiatric disorders.
Vladyslav Vyazovskiy comments, ‘This work was made possible by combining multidisciplinary expertise in systems and molecular neuroscience, sensory neurophysiology, neurobiology of sleep, physics and time series analysis. It was led by a talented graduate student Cristina Blanco-Duque, in collaboration with Eckehard Olbrich from Max Planck Institute for Mathematics in the Sciences, Peter Achermann from the University of Zurich, and my long-term collaborator from Experimental Psychology David Bannerman.’
Read the paper here