[Eeglablist] Open online discussion: How Do Cable Theory and AMPA/GABA Balance Compare in Their Contributions to 1/f?

Wed Apr 29 11:45:00 PDT 2026

Hi Ching-Ming,

In our upcoming paper, I hope to argue that while cable theory and
spatial averaging provide a stable biophysical baseline,

Both spatial averaging and cable theory are likely to be dynamic because

   - A contribution ratio between local-highFreq and broad-lowFreq varies
   based on cortical dynamics
   - Synaptic input locations varies depending on which network is dominant
   (exlem thalamocortical = apical dendrites (supragranular) input = more
   low-pass filter, lem thalamocortical = infragranular input = less low-pass
   filter.

In fact, I predict that nothing is truly stationary in the brain in the
timescale of a typical cognitive experiment.

It suggests that
as the exponent increases towards a certain limit, the brain might
lose the specific "complexity" required for high-level information
processing, leading to the topological breakdown we recorded.

Ah--that makes sense. Flatter PSD (i.e., invertebrate) = less AR = less
complexity/memory (?) (closer to white noise = max entropy)

I look forward to your paper! Fingers crossed for you.

Makoto

On Mon, Apr 27, 2026 at 1:37 AM Ching-Ming Lee <chingming.lee at gmail.com>
wrote:

> Dear Makoto,
>
> Thank you for the insightful response and for suggesting the term
> "System Health Index." It perfectly encapsulates our observation that
> the 1/f exponent can serve as a dynamic barometer for the functional
> boundaries of neural networks. We initially embarked on this analysis
> because of a recurring clinical paradox: we noticed that some patients
> showed signs of cognitive decline or 'regression' after excessive rTMS
> sessions. However, without objective metrics, these observations
> remained anecdotal. Our EEG findings, particularly the 'Network
> Collapse' beyond a certain exponent threshold, finally provide a
> biophysical basis for these clinical warnings. It turns out that what
> the clinicians 'felt' was happening is exactly what the neural network
> topology shows.
>
> Regarding your question: No, we have not published this finding yet.
> The manuscript is currently a work in progress. We are currently
> refining our preprocessing pipeline—specifically comparing GEDAI  with
> ASR+ICA—to ensure that the non-linear "Network Collapse" we observed
> at higher exponents (e.g., ~1.8) is robust and not an artifact of
> signal processing.
>
> I am particularly intrigued by your list of contributors to 1/f-ness.
> In our upcoming paper, I hope to argue that while cable theory and
> spatial averaging provide a stable biophysical baseline, the dynamic
> shifts we see under intensive rTMS are likely driven by acute
> modulations in synaptic kinetics and thalamo-cortical state
> transitions.
>
> The invertebrate comparison you shared is brilliant. It suggests that
> as the exponent increases towards a certain limit, the brain might
> lose the specific "complexity" required for high-level information
> processing, leading to the topological breakdown we recorded.
>
> I would be happy to share a preprint with the list once our validation
> is complete.
>
> Best regards,
>
> Ching-Ming
>
> Makoto Miyakoshi via eeglablist <eeglablist at sccn.ucsd.edu> 於
> 2026年4月27日週一 上午7:41寫道：
> >
> > Hi Ching-Ming,
> >
> > when the exponent reaches a certain threshold (~1.8 in our case), we
> > observe a non-linear 'Network Collapse' in global efficiency (via
> > WPLI analysis).
> >
> > This implies that the 1/f exponent might be a 'System Health Index' that
> > tracks how close the neural cable is to its functional boundaries.
> >
> >
> > That sounds very interesting. Did you publish the finding?
> >
> > My understanding is that multiple factors contribute to 1/f^a-ness of EEG
> > power. I can at least list
> >
> >    - Spatial averaging (more spatial averaging for high frequency
> activity)
> >    - Cable theory (aka dendritic filter)
> >    - AMPA_A/GABA receptor (Antonio was absolutely right when he wrote
> "what
> >    matters is their decay kinetics" This is why AMPA_A/GABA stuff can be
> >    'trivial', as Yevgeny once wrote, if you can predict FFT results from
> >    time-domain signals.
> >    - Infraslow EEG component due to quasi-stationary current loop between
> >    dendrites and soma (I will submit this interpretation soon)
> >    - Thalamo-cortical dysrhythmia proposed by Llinas and colleagues in
> case
> >    of pathological conditions
> >
> > Now it may rather be difficult to imagine how 1/f does NOT occur.
> Regarding
> > this question, I found an interesting study reporting that invertebrate
> > shows less 1/f-ness in their nervous system. I quoted their results below
> > for your interest. This comparison seems to provide a useful insight.
> >
> https://urldefense.com/v3/__https://github.com/sccn/OneOverF/discussions/16__;!!Mih3wA!Fd4oBZHv8fvFuWhCS5WK-B-Uz55k_1_LIuy1zBgV_pHndN5Ty_oQ46QBv4QgVr8cSdfhND2LtDFmTDsAPfh2-PpNAyw$
> >
> > Makoto
> >
> >
> > On Wed, Apr 15, 2026 at 8:16 PM Ching-Ming Lee <chingming.lee at gmail.com>
> > wrote:
> >
> > > Thank you all for these profound insights. I find the 'variable RC
> > > circuit' perspective particularly compelling.
> > > From a biophysical standpoint, if we view the neuron as a dynamic
> > > cable, the opening and closing of ion channels (gating) essentially
> > > represent the real-time modulation of resistance (R) and capacitance
> > > (C). In this framework, the E/I balance is not a competing theory, but
> > > rather the functional driver that dictates these parameter shifts.
> > > Our longitudinal n-of-1 data suggests that while the anatomical
> > > structure sets the 'static' baseline of the cable, interventions like
> > > rTMS trigger a dynamic state-shift in these RC properties. The most
> > > intriguing finding is that this shift isn't infinite; when the
> > > exponent reaches a certain threshold (~1.8 in our case), we observe a
> > > non-linear 'Network Collapse' in global efficiency (via WPLI
> > > analysis).
> > > This implies that the 1/f exponent might be a 'System Health Index'
> > > that tracks how close the neural cable is to its functional
> > > boundaries. It’s a beautiful intersection where cable theory provides
> > > the physical medium, while E/I dynamics provide the regulatory signal.
> > > I would be curious to hear your thoughts on whether this 'Network
> > > Collapse' could be interpreted as the point where the RC parameters
> > > reach a state of functional saturation or over-inhibition.
> > >
> > >
> > > Евгений Машеров via eeglablist <eeglablist at sccn.ucsd.edu> 於
> > > 2026年4月15日週三 下午10:42寫道：
> > > >
> > > > >Hi Ching-Ming, Yevgeny, and Cedric,
> > > >
> > > >  >Thank you for sharing your experience and results!
> > > >
> > > >  >*Structural Stability vs. Dynamic Change:* Since the subject's
> neuronal
> > > >  morphology (cable theory properties) cannot change by 60% within 3
> > > hours,
> > > >  this massive shift provides strong evidence that *E/I balance
> > > >  (GABA-mediated inhibition)* is the primary driver of 1/f *dynamics*,
> > > >  even if cable theory sets the *baseline*.
> > > >
> > > >  >Cable theory describes that when a neuron receives synaptic inputs
> at
> > > >  synapses that are located further from the soma, it generates more
> > > low-pass
> > > >  filtered post-synaptic potential. See a nice illustration from Rall
> et
> > > al.
> > > >  (1967)
> > > >
> > >
> https://urldefense.com/v3/__https://github.com/sccn/OneOverF/discussions/8__;!!Mih3wA!DHtN8b_WLfq9LU0kcatC6qilNfGVpUWCoDhdGUr5gz4qcT4adsUlFefUMcDmaZVmsSJ6ai6Kq6mOCrDe05dArQOcMds$
> > > >
> > > >  >Stephanie Jones' human neocortical neurosolver (HNN) has an
> explicit
> > > 2x2x2
> > > >  models: cells at layer 2/3 and 5, distal and proximal inputs, and
> > > >  excitatory (pyramidal neurons) and inhibitory (basket cells)
> neurons.
> > > Among
> > > >  these parameters, 'distal input' is likely to be associated with
> 'more
> > > >  low-pass filtered post-synaptic membrane potentials'.
> > > >
> > > >  >Ching-Ming, thus the question is 'Can TMS (or any other
> intervention)
> > > >  change the balance between distal and proximal inputs?' This is much
> > > >  trickier than thinking whether or not neural morphology can change
> by
> > > 60%
> > > >  in 3 hours.
> > > >
> > > > > We can think of synaptic inputs to apical dendrite in superficial
> > > layers as
> > > >  a typical case of 'distal input'. If TMS (or any other intervention
> > > such as
> > > >  hyperventilation) can increase relative amount of 'distal input',
> that
> > > >  would shift 1/f to 'steeper' = 'as if GABA_A-R is dominant at the
> > > >  measurement site'.
> > > >
> > > > > What can increase inputs to apical dendrite? I published a paper
> last
> > > year
> > > >  that extralemniscal thalamic (EXLEM) projection goes to the
> superficial
> > > >  layers (about 25%; See 'type 2' in this illustration
> > > >
> > >
> https://urldefense.com/v3/__https://github.com/sccn/OneOverF/discussions/17__;!!Mih3wA!DHtN8b_WLfq9LU0kcatC6qilNfGVpUWCoDhdGUr5gz4qcT4adsUlFefUMcDmaZVmsSJ6ai6Kq6mOCrDe05dAARwz5uE$
> > > ). EXLEM is the same as
> > > >  'non-specific thalamus' mentioned by Grey Walter, Robert Galambos
> etc..
> > > >  Recently, Giandomenico Iannetti has been leading the revival of
> EXLEM.
> > > >
> > > >
> > > > It's possible that the presence of an effect from transcranial
> magnetic
> > > stimulation doesn't refute the cable theory's explanation of the
> dependence
> > > (although it is extremely interesting information). This is because
> the RC
> > > circuits formed by dendrites, which transform the pulse shape, differ
> in
> > > their properties from those used in radio engineering. Changing
> resistance
> > > or capacitance by applying voltage to a conventional radio engineering
> > > resistor or capacitor is impossible (at least until the capacitor
> breaks
> > > down or the resistor burns out) – however, varistors and
> varicapacitors are
> > > also used in radio engineering. Ion solutions, under the influence of
> the
> > > current flowing as a result of magnetic induction, change their
> > > concentration, which can change resistance. Channels in membranes can
> open
> > > or close, and the electrical double layer on the membranes can change
> > > (possibly related to the "non-specific benefit" of TMS). All of this
> can
> > > affect conductivity and capacitance, thereby generating a pulse shape
> > > different from that formed before transcranial magnetic stimulation.
> > > >
> > > >
> > > > > Yevgeny, your reviewer will be angry and tell you you should never
> > > call a
> > > >  study with n=167 preliminary. Thank you for sharing your results!
> It is
> > > >  good to know that 1/f is not sensitive to hyperventilation.
> > > >
> > > >
> > > > The number is quite large, but to view this as anything more than a
> > > preliminary experiment, all these observations must be considered and
> some
> > > discarded. Therefore, these are nothing more than experiments during
> the
> > > debugging of a program designed to analyze a more carefully selected
> > > sample. However, since some new information has been obtained, I offer
> it
> > > to your attention, but only as "food for thought," not as conclusive
> > > evidence.
> > > > I plan to obtain results suitable for full publication, not just as a
> > > "personal communication," but to do this, I need to involve clinicians
> who
> > > can assess the suitability of each record for such analysis. I have
> neither
> > > the authority nor the expertise to analyze case histories. This
> requires
> > > time and effort to organize. First and foremost, I need to convince
> them of
> > > the usefulness of this work for them—"it's not so difficult to teach a
> bear
> > > to ride a bicycle than to convince a bear that it needs one." So, what
> I've
> > > submitted for your consideration, I'm not ready to send to a fully
> > > peer-reviewed journal until I've thoroughly checked it. If I say
> something
> > > stupid at work, it's my personal shame; if I publish something stupid
> in a
> > > fully-fledged journal, I'll disgrace my colleagues.
> > > > And another picture, the distribution of the "alpha" by scalp (median
> > > for 167 patients)
> > > >
> > > > Eugen Masherov
> > > >
> > > >
> > >
> https://urldefense.com/v3/__https://ibb.co/cSVsWQBB__;!!Mih3wA!HSuCVd5XYQJMfnY5YTX3pu_k-u1q0EHR_OsYCO9ORIt1DjDy9jpQh-uwUymnqmkz2tLYcVW7RtRBynA8CsD1H5I11io$
> > > > _______________________________________________
> > > > To unsubscribe, send an empty email to
> > > eeglablist-unsubscribe at sccn.ucsd.edu or visit
> > > https://sccn.ucsd.edu/mailman/listinfo/eeglablist    .
> > >
> > >
> > >
> > > --
> > > 李景明
> > >
> > _______________________________________________
> > To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu or visit
> https://sccn.ucsd.edu/mailman/listinfo/eeglablist  .
>
>
>
> --
> 李景明
>