When Memory Fades, Is the Problem in the Neurons or the Connections Between Brain Networks?

What happens when we stimulate one brain region and observe how the entire network responds?

Alzheimer's Disease · TMS–EEG · fMRI · Brain Connectivity · Connectomic Reserve · Neuroplasticity

Imagine meeting someone familiar and recognizing their face but struggling to remember their name. The information seems close, yet the memory remains incomplete. At another moment, someone may recall individual events but struggle to organize them into a sequence.

What is happening? Have certain neurons stopped functioning properly? Or are brain regions no longer communicating in ways that support the same cognitive operations?

In September 2026, Giacomo Bertazzoli, Elisa Canu, Chiara Bagattini, Marta Bortoletto, and a large international research team published Network-targeted neurophysiological biomarkers of dysconnectivity and cognitive decline in Alzheimer's disease in Alzheimer's Research & Therapy.

Their investigation introduces an important approach: stimulating regions belonging to specific brain networks and recording how the brain responds to that perturbation.

The researchers aimed to identify neurophysiological alterations associated with cognitive impairment and the progression of Alzheimer's disease.

1. The brain is not a collection of isolated regions

Remembering an experience, retrieving information, and making decisions require coordinated activity across multiple brain regions.

Alzheimer's disease involves cellular and molecular changes, but it also progressively disrupts functional and structural communication between networks.

Two networks received particular attention in this investigation.

The default mode network (DMN) contributes to autobiographical memory, internally directed thought, and information integration.

The executive control network (ECN) participates in planning, working memory, and cognitive control.

These networks do not operate independently. Their interactions contribute to how we organize experiences.

Bertazzoli and colleagues asked whether brain responses to targeted stimulation could reveal alterations that conventional assessments might not easily detect.

2. How did the researchers design the experiment?

The study included 86 participants, divided into three groups: 28 cognitively healthy controls, 27 individuals with mild cognitive impairment, and 31 individuals with Alzheimer's dementia.

The researchers combined resting-state functional magnetic resonance imaging (resting-state fMRI) with transcranial magnetic stimulation and electroencephalography, known as TMS–EEG.

First, fMRI enabled the identification of individual brain regions belonging to the DMN and ECN.

The team then delivered magnetic pulses to these targets and recorded the resulting electrical brain responses.

This individualized approach matters because the functional location and organization of brain networks may vary between individuals.

The investigation also included cognitive assessments and structural brain measurements, including hippocampal volume and white matter integrity.

In 55 participants, cognitive functions were followed longitudinally for approximately 18 months.

The study therefore combined anatomical, functional, neurophysiological, and clinical information.

3. What actually happens when we apply TMS?

Transcranial magnetic stimulation uses a coil positioned near the head to generate a changing magnetic field.

This field induces an electric field in brain tissue, capable of temporarily modifying the activity of excitable neural elements.

When TMS is combined with EEG, researchers can record responses known as TMS-evoked potentials (TEPs).

Imagine striking one instrument within an interconnected musical ensemble and observing which other instruments respond, when they respond, and how strongly.

The analogy illustrates the experimental principle, although neural propagation is considerably more complex.

Bertazzoli and colleagues focused particularly on early responses recorded within the first 50 milliseconds following stimulation.

These responses can provide information about cortical reactivity and interactions among connected brain regions.

However, EEG cannot directly identify which individual synapses contributed to a particular response.

4. The discovery: differences emerge within milliseconds

One of the most relevant findings emerged following stimulation of the left DMN.

The researchers identified a contralateral frontal response, called N20F, whose amplitude was greater in participants with mild cognitive impairment and Alzheimer's dementia than in healthy controls.

This measure distinguished healthy controls from participants with mild cognitive impairment, achieving an area under the receiver operating characteristic curve (AUC) of 0.82.

This is promising, but it does not mean the test has been validated for individual clinical diagnosis.

Within the executive control network, another response, called N20C, showed patterns that depended on the stimulated hemisphere.

These differences varied according to clinical status and stimulation side.

The central finding is that brain responses depend not only on the stimulated region, but also on the network being investigated, the hemisphere, and the individual's clinical condition.

This highlights the importance of examining the specific functional organization of neural circuits.

5. Can a network response anticipate cognitive decline?

The researchers identified associations between TEP components, structural brain characteristics, and cognitive measurements.

An N20C response following left DMN stimulation was associated with the integrity of the superior longitudinal fasciculus, a major white matter pathway.

Associations were also observed between certain ECN responses and executive performance.

During longitudinal follow-up, lower N20C amplitude following left DMN stimulation was associated with greater global cognitive decline.

Among participants with mild cognitive impairment, responses involving the right ECN also showed predictive value for executive changes.

These findings suggest that network-targeted neurophysiological measurements may help characterize individual differences and monitor cognitive trajectories.

However, larger independent investigations, including biological confirmation of Alzheimer's pathology, are needed before establishing their clinical usefulness.

6. What does this research actually demonstrate about connectivity?

There is a fundamental difference between observing spontaneous correlations and experimentally provoking a response.

Resting-state fMRI allows researchers to investigate brain regions whose signals fluctuate in related patterns.

TMS–EEG adds the possibility of introducing a controlled perturbation and observing the system's response over time.

This provides a perspective on effective connectivity and causally evoked dynamics, although interpretation requires careful control of sensory artifacts, signal propagation, and indirect effects.

Stimulating one region and recording a response elsewhere does not automatically demonstrate a direct anatomical connection between them.

Transmission may involve intermediate circuits and multiple synaptic interactions.

The research team's achievement lies in combining complementary techniques to examine network alterations across different clinical conditions.

It is also important to emphasize that the study was observational in its clinical purpose: it did not demonstrate that TMS restored memories or treated Alzheimer's disease.

7. Connectomic Reserve: are there pathways that could still respond?

An interesting connection emerges with the hypothesis proposed by Roberto Lent, Fernanda Tovar-Moll, and Diego Szczupak in The secret connections of the brain: a connectomic reserve for neuroplasticity?.

The authors propose that pre-existing anatomical connections, some relatively underused or functionally inhibited, may represent potential resources for neural reorganization.

This raises a question for BrainLatam:

Could differences in TMS-evoked responses help investigate the functional availability of circuits belonging to a Connectomic Reserve?

This is an experimental possibility, not a finding demonstrated by Bertazzoli and colleagues.

Converging structural and functional evidence, together with controlled interventions, would be necessary to determine whether previously underused connections become functionally engaged.

The concept of a Relational Connectomic Reserve, explored by BrainLatam, extends this discussion toward investigating how accumulated experiences and social interactions relate to the functional repertoires available to an organism.

It remains a theoretical hypothesis requiring experimental operationalization.

8. From an immediate response to transformation between encounters

From the perspective of Metanoia, we can propose two levels of investigation.

At the first level, we observe A, the brain state before the magnetic pulse; B, the response evoked by TMS; and the subsequent return toward baseline conditions.

At the second level, we follow the organism through learning experiences or cognitive changes, comparing its responses across successive encounters.

The distinction is crucial: an evoked response lasting milliseconds does not demonstrate lasting transformation.

To investigate transitions from A→B→A, we would need to measure persistence, reversibility, and dependence on the trajectory taken.

TMS–EEG provides tools for formulating these questions, but it cannot replace behavioral measures or longitudinal follow-up.

9. New technologies, new scientific questions

The anticipated arrival of BrainAmp II, with its public presentation planned for SfN 2026, alongside Research TMS technologies, increases interest in experimental architectures combining recording, stimulation, and network analysis.

However, integration between specific devices must be technically validated.

Scientific innovation does not consist merely of using a newer amplifier.

It consists of investigating when, where, and under which brain states a perturbation produces different responses, while recognizing the original contributions of the researchers.

Bertazzoli and colleagues demonstrated the value of a network-targeted approach for understanding neurophysiological alterations associated with Alzheimer's disease.

10. When memory weakens, what should we investigate?

The opening question does not require choosing between neurons and connections.

Memory depends on cells, synapses, networks, and physiological conditions that continuously interact.

Bertazzoli and colleagues add an important dimension: we can investigate the functional integrity of brain networks by observing how they respond to a carefully targeted experimental perturbation.

Their contribution opens possibilities for more individualized biomarkers and new questions about plasticity, compensation, and brain organization.

But it does not demonstrate memory recovery or direct recruitment of a Connectomic Reserve.

In Blog 05, we will move from networks involved in cognition to language: does the language we learn change our expectations and the way we process our next encounter with words?

Main scientific reference: Bertazzoli, G., Canu, E., Bagattini, C., et al. (2026). Network-targeted neurophysiological biomarkers of dysconnectivity and cognitive decline in Alzheimer's disease. Alzheimer's Research & Therapy.

Meta description: Discover how a 2026 study uses TMS–EEG and fMRI to investigate brain network dysconnectivity, cognitive decline, and Alzheimer's disease.

BRAINLATAM | W47/2026 | BLOG 04/12 — ENGLISH VERSION