Tom Donoghue
LightNews LightNews
tomdonoghue.bsky.social
Tom Donoghue
@tomdonoghue.bsky.social
890 followers 480 following 70 posts
Cognitive & Computational Neuro Scientist - studying electrophysiological signals in human brains, mostly by writing Python code.
Lecturer of Cognitive Neuroscience @ University of Manchester.
https://tomdonoghue.github.io/
Posts Replies Media Videos
tomdonoghue.bsky.social
Based on this high level overview of lots of interesting & exciting work, I also tried to integrate the discussion points and findings thus far to create a set of suggested recommendations for future work studying aperiodic activity in clinical populations. Hopefully this is useful for future work!
A copy of Table 3 from the manuscript, overviewing a set of recommendations for future work.
October 13, 2025 at 3:26 PM
tomdonoghue.bsky.social
I also discuss motivations for studying aperiodic activity, including as a potential biomarker & the oft discussed putative interpretation of the exponent as a marker of E/I balance. While this is an exciting possibility, I also urge some caution based on this review & other recent empirical work:
October 13, 2025 at 3:21 PM
tomdonoghue.bsky.social
It turns out there is a lot of current research on this, across quite a range of disorders with lots of reported links between clinical disorders and aperiodic activity - but also quite a bit of variability in the findings and overlapping discussion points that this review explores and discusses.
Copy of Table 1 from the paper, showing a summary of the most studied disorders and an overview of the findings
October 13, 2025 at 3:17 PM
tomdonoghue.bsky.social
I have now started at the University of Manchester as a lecturer
(equivalent to assistant professor)!

I’m looking forward to getting to know everyone here and getting started on the next phase of my research!
An image of the arch with the ‘University of Manchester’ sign
August 4, 2025 at 1:35 PM
tomdonoghue.bsky.social
In 1949, two separate papers reported that i) there was an exponential distribution of energy in the EEG (Motokawa) & that autocorrelation analyses suggested 'aperiodic motions' in EEG activity (Imahori & Suhara)!

In the 75 years since, there have been lots of cool papers, many seemingly forgotten.
A reprint of Figure 1 from Motokawa, 1949, showing an exponential distribution of energy in the EEG. A reprint of a figure from Imahori & Suhara, 1949 showing autocorrelation results of a rythmic and non-rhythmic section of the EEG.
November 21, 2024 at 6:59 PM
tomdonoghue.bsky.social
The premise is that there are many different kinds of methods that measure (probably) related features of the data - especially when thinking of 'aperiodic' activity.

This project seeks to investigate how these different methods relate to each other!
A reprint of figure 1 of the manuscript, showing different simulated colours of noise (white noise, pink noise, etc), and different kinds of methods that get applied to them (autocorrelation, entropy, etc).
September 18, 2024 at 8:07 PM
tomdonoghue.bsky.social
Reading this old EEG paper, and we really don't design behavioral tasks like we used to, huh...
Screenshot of text from the paper, which reads: The six tasks were as follows: 1) observation of a swinging double pendulum; 2) to name aloud nouns which started with the letter M (e.g., mouse, man, milk, morning); 3) to determine which of six pieces of sandpaper was the smoothest with the right index finger; 4) the same task with the left index finger; 5) to image and experience an extremely positive time in their past in which they had felt in love (without actual sexual imagery) and 6) to image the same type of extremely positive experience which included a sexual experience.
February 15, 2024 at 6:37 PM
tomdonoghue.bsky.social
A fun find in an old paper - in 1956, Mary Brazier noting that of course EEG contains both periodic and aperiodic activity, and relating it to the auto-correlation based analyses they were developing at the time:
Screenshot of text that reads: "When both classes of potentials are present, both periodic and aperiodic, as is usual in the EEG, the correlogram shows a periodically recurrent correlation, alternately positive and negative, riding on the initial decay curve caused by the aperiodic components”
January 7, 2024 at 10:28 PM
tomdonoghue.bsky.social
I have a few projects this SfN on two main themes:

For human single-unit work, see our posters Sunday afternoon, or my poster on Monday morning!

For spectral parameterization work, I'll be at our sleep poster on Monday afternoon!

+ collaborator projects!

If anything's of interest, come say hi!
November 10, 2023 at 9:46 PM