Anirudh Wodeyar @awodeyar@qoto.org

Just had a public science (my first!) article published in The Hindu in print(!!) on how Giant AIs may be an atlas, but they will never be the territory.

Our direct human observations point to a potential mechanism linking epileptiform spikes to cognitive dysfunction: sleep-activated epileptic spikes inhibit sleep spindles.

Epileptic spikes propagate from the cortex to the thalamus, with spikes tending to propagate to the thalamus more often in patients with sleep activated spikes (SWAS) and most in patients with epileptic encephalopathy (EE-SWAS). (see first figure below)

In patients with severe cognitive dysfunction, thalamic spikes reduced spindles for 30 seconds and decreased overall spindle rate. See below for an example sweep of data showing the reduction of thalamic spindle occurrence from spikes. (see second image below)

Findings show how slow oscillations can facilitate both epileptic spikes and sleep spindles, which can lead to them looking correlated.

Our study utilizes a unique dataset of simultaneous human and cortical recordings to investigate the relationship between epileptic activity and two of the cardinal sleep oscillations – slow oscillations (0.5 – 2 Hz) and #spindles (9-16 Hz) – generated in the thalamus.

Our new paper (https://biorxiv.org/content/10.1101/2023.04.17.537191v1) using thalamic data (!) explores how epileptic spikes block sleep spindle production during non-REM sleep. Discover the impact of epilepsy on sleep-dependent memory consolidation! #tootprint #epilepsy #neuroscience

A fun quick project using mostly what I already had coded up for a research project looking at connected alpha oscillators in the brain. The oscillators are connected by the HCP white matter connections.

Both the spheres and the music are synced to the simulated oscillator data. So to me this is what a resting brain sounds like.

The music is generated using piano notes connected to the oscillator activity and the location influences frequency.

Color is orange when positive and purple when negative. And size changes with intensity.
#generativeart

The big message we wanted to convey is that depending on what is intended for the phase to track (do you want it to act like a clock or do you want it to tell you when the peak/trough is reached?) you might want to consider alternative methods and use uncertainty metrics.

I focused in this #tootprint on one situation when the phase can become quite ambiguous - amplitude modulation, but in the paper we consider other cases as well including non-sinusoidal oscillations which lead to other considerations - do check it out! - https://doi.org/10.1101/2023.01.05.522914

As an aside - "high confidence" moments needn't only be times when the amplitude is high. By applying confidence/credible limits and looking across methods we can help refine phase estimation and better help us understand the role of phase in neural dynamics.

We show how we can use confidence limits on the phase coming from a Hilbert transform and credible intervals from the state space model to define moments of time ("high confidence") when the phase will be correlated across these different methods. Here's an example for AR(2) sim data.

We consider two alternatives: a state space model of rhythms and using zero-crossings to define phase. Each of these suggests a different phase estimate during times when the rhythm power decreases or if the waveform of the rhythm is non-sinusoidal.

But in moments when there's less rhythm power the filter-Hilbert approach has to create a cycle since the instantaneous frequency is bounded by the filter. But this may not be what is desired for the phase estimate. Alternatives will suggest different estimates for the phase.

Classically, I've seen the filter-Hilbert transform-analytic signal approach used to define phase. This will work fine in many high rhythm power scenarios because the phase that comes out using this approach will strongly correlate with that from other approaches.

All this despite the fact that phase is only explicitly defined for a pure sinusoid, or a narrowband oscillation. If the data is anything else, we are constructing one potential phase estimate of many.

Phase is *the* exciting characteristic of brain rhythms, amplitude frequently seems to act like a barometer of rhythm presence while phase does all the grunt work in expectations about the function of rhythms.

For any #neuroscientists (particularly folks working on intracranial data) in #Boston, I'm excited to be presenting at the first annual MGH-MIT inBrain symposium at the Ether Dome, MGH tomorrow 12/9 at 3 pm! Come by if you're around :)