DOCTORAL RESEARCHER IN NEUROSCIENCE · ERLANGEN
From hearing aids to human brains.
The short version of a long detour — and why a researcher who spends his days on brain scans is writing about how you think at work.
I’m Mageshwar Selvakumar, a doctoral researcher in neuroscience in Erlangen, Germany. My path here was not a straight line. I trained as an electronics engineer in Chennai, moved to Germany for a master’s in audio signal processing, and spent years at Fraunhofer working on speech enhancement — teaching hearing aids to pull a single voice out of a noisy room. Today I use multi-parametric MRI to study how chronic pain reshapes the human brain.
Different instruments, the same craft: finding weak signals in noisy data, stating hypotheses precisely enough to be proven wrong, and drawing honest conclusions from small samples.
What I actually do all day
Most of my work is not dramatic. I design studies, recruit and scan participants, and then spend long stretches cleaning and modelling data — deciding what is genuine signal and what is an artefact of movement, physiology, or the scanner itself. A single defensible claim about the brain rests on weeks of careful, unglamorous preprocessing and statistics.
The lesson repeats in every project: the quality of an answer is set long before the analysis — in how carefully the question was framed and the measurement was designed. Get those right and the statistics are almost bookkeeping. Get them wrong and no analysis will rescue you.
Why I avoid the fashionable methods
Human neuroimaging studies are small — thirty to forty people is normal, because scanning is expensive and patients are hard to recruit. That constraint shapes how I think. I deliberately avoid data-hungry methods like deep learning on studies of this size, however impressive the results would look, because thirty people cannot honestly support them.
Matching your confidence to your evidence is not a limitation to apologise for; it is the discipline that makes a conclusion trustworthy. It is also the single habit I most want to pass on to anyone who makes decisions on small, messy, incomplete data — which is to say, everyone at work.
Why this site exists
AI is making answers cheap. What it has not made cheap is knowing which question to ask, which evidence to trust, and which consequence nobody has priced in yet. Those are thinking skills — methods, not talents — and they are exactly the tools I use every day in the lab.
Ask the Brain Scientist is where I translate them into a form you can use in an ordinary workweek: one article at a time, with the evidence graded honestly, and one monthly email that does the same. If that’s useful to you, the newsletter is the best place to start.
SIGNAL — THE MONTHLY EMAIL
Start with the five models.
The free guide — 5 Mental Models to Future-Proof Your Career — plus one calm, evidence-backed idea from neuroscience each month. No productivity spam, no AI panic.
Free PDF, then one email a month. Confirm by email (double opt-in). Unsubscribe anytime.