[Eeglablist] Is this what we see in the EEG?

[Cedric Cannard]

Quick comment just to say that BOLD signals are also uneven and their sensitivity is debated. See for example this fascinating recent paper: https://urldefense.com/v3/__https://www.nature.com/articles/s41593-025-02132-9__;!!Mih3wA!Bei0EQ_4djbkxzzW7eKji44_i6---H5G2_CpjJMkk9FFp77kfSFQYuAn8i8kBnRZ0bmMM-b0e801x3ZlX7j8qML83A$ 

Blood flow is again causing problems... :) 

Cedric


Sent with Proton Mail secure email.

On Wednesday, May 20th, 2026 at 9:10 PM, Makoto Miyakoshi via eeglablist <eeglablist at sccn.ucsd.edu> wrote:

> Hi Yevgeny,
> 
> > when the Tsar ordered Dmitri Mendeleev to discover the secret of French
> smokeless powder, he examined the factory dump.
> 
> I like that.
> Let's use this example. Now, imagine that Dmitri uses a stethoscope to hear
> the sound of the factory instead of examining the factory dump. This
> hypothetical example, which is a poor choice and more challenging, may be
> closer to the situation of scalp EEG measurements. Instead of getting
> quantitative and qualitative information of chemical compounds used, he
> would learn temporal dynamics of the manufacturing process and types of it
> by performing time-frequency analysis, etc.. With a microphone array
> attached to the factory wall and roof, he may be able to estimate locations
> of different types of processing machines and their sizes. But he would
> never know the precise ratio of ingredients, processing temperature, etc..
> Most importantly, some processes that make little noise would be
> transparent to his measurement such as electrolysis.
> 
> My point is that God did not create EEG as an official window to study the
> brain. Instead, we hack the brain's hypothetically epiphenomenal byproduct
> to estimate its function. Thus, there is no guarantee on what brain
> functions can measure: spatial distribution of EEG's sensitivity is uneven,
> unlike BOLD signals. Hence we cannot use Gaussian random field theory in
> inferential statistics.
> 
> Another analogy that may be interesting and useful is that EEG research is
> more like open-source intelligence (OSINT) than signal intelligence
> (SIGINT). OSINT is basically data mining on open information. In contrast,
> molecular biology is an example of biological SIGINT because it deals with
> the DNA's (de)coding principle.
> 
> Makoto
> 
> 
> On Tue, May 19, 2026 at 4:11 AM Евгений Машеров <emasherov at yandex.ru> wrote:
> 
> >
> >
> > > Hi Yevgeny and Jack,
> > >
> > >> True brain activity manifests itself in the higher frequency range and
> > >
> > > over a very broad frequency band (which corresponds to the conclusions of
> > > information theory regarding bandwidth and information transmission
> > > intensity).
> > >
> > > Mathematics captured the universe, but did information theory so too?
> > There
> > > seems to be different types (definitions) of information than that in
> > > Shannon's sense. People talk about brain's interregional 'connectivity'
> > by
> > > showing phase coherence etc., but has anyone proven what is communicated
> > > between the regions? What is the principle of brain's calculation?
> > >
> > > I even think these questions are perhaps wrong in Wittgenstein's sense.
> > >
> > > Another random skepticism: Electric fields in the extracellular space is
> > > mostly epiphenomenal in the sense that they do not carry any functions
> > (an
> > > exception is ephaptic coupling, which occurs in much smaller
> > > spatial scale). Am I correct? I always think that measuring EEG to
> > estimate
> > > brain's state is like using a stethoscope to diagnose a car engine. A car
> > > engine makes noise, but its purpose is to drive a car. If this is the
> > case,
> > > then how can we be so sure that critical information must be, and always,
> > > contained in the signal? In fact, we already know that scalp
> > > EEG's sensitivity to various brain functions and states are patchy.
> > >
> > Information is what can change our actions. Shannon's definition, through
> > probability, is simply a way to make measurable what is intuitively
> > estimable. In this case, it's a way to poorly measure something that can't
> > be measured otherwise. Clearly, any analogies with technology are
> > imprecise, but nevertheless, similar problems sometimes generate similar
> > solutions.
> > It seems to me that electric fields are secondary to ionic concentration
> > (I'd say "epiphenomenon," but I'm afraid to confuse epileptologists). And
> > in this sense, changes in the field at one point can be related to the
> > field at another, albeit indirectly. As for the analogy with listening to
> > the knock of an engine to detect a malfunction, that's a good one, but I
> > can offer another: when the Tsar ordered Dmitri Mendeleev to discover the
> > secret of French smokeless powder, he examined the factory dump.
> >
> > Eugen Masherov
> >
> _______________________________________________
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu or visit https://sccn.ucsd.edu/mailman/listinfo/eeglablist .