Effects of intracranial electrical brain stimulation on neuronal dynamics recorded by tetrodes microelectrodes in epileptic patients – DYNEUMICS

How?

A novel stimulation protocol

Rather than relying on rigid, one-size-fits-all procedures, the project introduces a new stimulation strategy based on brain physiology. Electrical stimulation frequencies are no longer fixed at standard values but are selected according to the brain region being explored and its natural rhythms (theta, alpha, beta, etc.). This approach allows researchers to test how different frequencies interact with local brain dynamics, with the goal of improving the reliability and clinical relevance of intracerebral electrical stimulations.

Unique hybrid intracerebral electrodes

The project relies on a new generation of hybrid electrodes developed in collaboration with the French manufacturer DixiMedical and implanted for the first time in patients in Toulouse. These electrodes combine two complementary technologies within the same device:

macro-contacts, which record and stimulate large-scale brain networks, as used in conventional SEEG;

micro-contacts, organized in groups of four (tetrodes), capable of recording the activity of individual neurons.

Using such hybrid electrodes during electrical stimulations is a world first. This technological innovation makes it possible to observe, simultaneously and in real time, how the same stimulation affects both global brain networks and microscopic neuronal activity.

A multiscale analysis of brain responses

Thanks to this unique setup, the research team can analyze the effects of intracerebral stimulation across multiple levels of brain organization:

at the level of single neurons, by tracking changes in firing activity;

at the level of local circuits, by studying interactions between neighboring neuronal populations;

and at the level of large-scale brain networks, through intracranial EEG signals.

This multiscale approach provides an unprecedented view of how electrical stimulations propagate through the brain, sometimes producing subtle neuronal effects that remain invisible on standard clinical recordings.

Towards clinically optimized stimulation parameters

By combining tailored stimulation frequencies with hybrid electrodes and multiscale analysis, the project aims to identify stimulation parameters that are both more effective and more informative for clinical decision-making. Ultimately, this methodology is designed to improve the precision of pre-surgical epilepsy evaluation, shorten stimulation sessions, and deepen our understanding of how electrical stimulation interacts with the human brain.

The different studies conducted within the DYNEUMICS project show that intracerebral electrical stimulations used in patients with epilepsy can be made more effective, faster, and more precise by adjusting their parameters, rather than applying standardized settings to all patients.

A first major finding concerns stimulation frequency. In the temporal lobe, stimulation at 7 Hz—a frequency close to the brain’s natural rhythms—elicits relevant responses more reliably than the classically used frequency of 1 Hz. At equivalent intensity and duration, 7 Hz stimulations induce a higher number of typical epileptic seizures, characteristic electrical abnormalities, and clinically meaningful signs that help clinicians better delineate the boundaries of the epileptogenic network.

In contrast, stimulations at 50 Hz—another frequency that has been widely used for decades—although effective in certain situations, more frequently generate misleading responses in non-pathological brain regions, thereby increasing the risk of false positives. These results demonstrate that stimulation frequency plays a critical role and should be adapted to the brain region being explored.

A second set of results focuses on the duration of stimulation. The analysis of several thousand stimulations performed in 117 patients shows that when a clinical effect occurs (movement, sensation, or behavioral change), it most often appears very rapidly, on average around 2 seconds after stimulation onset. This contrasts with current clinical practice, where stimulations typically last between 5 and 20 seconds.

Nearly 99% of all observable reactions occur within the first 10 seconds, and extending stimulation beyond this duration provides very little additional information. Although response latencies vary depending on the stimulated brain region and the type of clinical effect, these findings indicate that long stimulations are most often unnecessary. Shortening stimulation duration could therefore improve the efficiency of data acquisition while also enhancing patient comfort during SEEG procedures.

Finally, thanks to the use of intracerebral microelectrodes, preliminary results suggest that electrical stimulations alter the activity of individual neurons, both locally and at a distance, even when no clear effect is detectable on standard clinical EEG recordings. This points to the existence of stimulation-induced effects that remain “invisible” to conventional clinical tools, yet may be crucial for understanding the organization and dynamics of epileptic networks.

Taken together, these results indicate that shorter stimulations, delivered at better-chosen frequencies and tailored to each brain region, could improve the quality of pre-surgical epilepsy evaluation, reduce examination time, and refine our understanding of how epileptic brain networks function.

The results obtained so far, together with the extensive datasets that are still under analysis, are expected to have major clinical, methodological, and conceptual impact.

For patients, these advances could directly translate into:

Shorter and more efficient pre-surgical evaluations, by reducing the duration and redundancy of stimulation procedures;

More accurate diagnoses, through a finer delineation of epileptogenic networks;

More precisely targeted surgeries, minimizing unnecessary tissue resection;

And ultimately, higher chances of seizure freedom, with fewer cognitive or neurological side effects.

By optimizing stimulation parameters and adapting them to the physiological properties of each brain region, DYNEUMICS aims to improve both the quality and comfort of invasive epilepsy evaluations.

For science and medicine, the project opens several major perspectives:

It provides a deep, multiscale understanding of how electrical stimulation interacts with the human brain, from individual neurons to large-scale networks;

It generates a unique, high-density database combining single-neuron recordings, local microcircuit dynamics, and network-level activity, all acquired simultaneously while brain activity is actively modulated by electrical stimulation;

This resource will support future exploratory and hypothesis-driven analyses that go far beyond the original scope of the project.

A key outstanding feature of DYNEUMICS is its potential contribution to standardization. By identifying stimulation parameters that are both physiologically grounded and clinically effective, the project could inform the development of evidence-based guidelines for intracerebral stimulation during SEEG, helping to harmonize practices across centers and reduce variability in patient care.

The approach developed in DYNEUMICS is also highly transferable. Its principles can be extended to other invasive neuromodulation contexts, including deep brain stimulation, chronic implanted devices, and adaptive stimulation systems. As such, the project lays conceptual and methodological foundations that may benefit other neurological and psychiatric disorders, such as Parkinson’s disease, chronic pain, and psychiatric conditions.

Beyond its clinical applications, DYNEUMICS represents a conceptual shift. Rather than relying on empirical, one-size-fits-all stimulation protocols, it promotes a model of physiology-driven neuromodulation, in which stimulation parameters are tuned to the intrinsic dynamics of brain circuits. This paradigm has the potential to reshape how electrical stimulation is designed, interpreted, and applied in both research and clinical settings.

DYNEUMICS is therefore a pioneering initiative. By combining technological innovation, personalized stimulation strategies, and fine-grained analysis of brain function, it could transform the management of drug-resistant epilepsy and pave the way toward next-generation, brain-informed neuromodulation therapies.

Intracranial electrical brain stimulations (EBS) are used during stereoelectroencephalography (SEEG) in pharmacoresistant epilepsy to delineate the epileptogenic zone by triggering seizures and other electrophysiological changes, and to do functional mapping before a surgery that may make patients seizure free. However, current EBS protocols are empirical and limited (only 2 frequencies are used, 1Hz and 50Hz, regardless of the brain area), with a low yield of clinically relevant effects. Moreover, the direct impact of EBS on neuronal activity, on local field potentials (LFP) or on widespread networks have seldom been explored. Very little is known about the modification of electrical parameters in humans. This is not limited to epilepsy, but to all the fields where EBS are used (e.g., Parkinson’s disease, some psychiatric disorders). Indeed, how EBS modulate neural activity in general is still a conundrum. Because multiscale intracranial recordings across several days and simultaneous stimulations cannot be done easily in any other pathological context, EBS during SEEG in epilepsy can be considered as a model to explore different types of EBS effects and parameters.

The DYNEUMICS project aims overall at helping clinicians determine more optimal electrical parameters for reproducible effects & better yields from EBS sessions: more clinical effects, more seizures triggered and consequently a decrease of SEEG duration, more relevant medical information for the surgical decision, higher efficiency of diagnostic phase and more comfort for the patient.

To achieve this,
- (1) We have designed a new enriched intracerebral EBS protocol (Dyneumics EBS: electrical parameters specifically tailored to each brain location, corresponding to physiological frequencies such as theta, alpha and beta, prominently recorded in specific brain areas), easy to use in clinical practice, We hypothesize that this will increase the percentage of effective EBS compared to standard EBS during the pre-surgical SEEG workup of patients with pharmacoresistant epilepsy. We will therefore contrast the new frequencies (Dyneumics EBS) to the standard frequencies (standard EBS: where only 2 types of EBS at 1 & 50 Hz are applied regardless of the brain location). We will collect and compare intracranial EEG and clinical effects of Standard and Dyneumics EBS.
- (2) We will also perform multiscale analyses to understand the entire cascade of modulations induced by these EBS, i.e. from single neuron activity to large scale networks. These analyses will be performed using a novel hybrid intracranial electrode combining several tetrodes (microelectrodes grouped by 4). Tetrodes are more efficient to isolate single neuron activity and LFP at the micrometer scale. In addition, macrocontacts measure LFP at the millimeter scale at > 120 locations per patient simultaneously, i.e. large-scale neural networks. We have been the first center throughout the world to use and develop expertise with this electrode since 2015. Indeed, we also hypothesize that the most relevant EBS (= leading to more intracranial EEG and clinical effects) will be those that are related to significant changes in single unit dynamics.

Thanks to DYNEUMICS, we should be able to draw a panel of optimal EBS parameters (couple frequency + brain location) for clinicians. We should aslo produce a rational of the panel of EBS selected thanks to a better understanding of the local and large-scale neuronal effects of EBS. Our project should improve our fundamental understanding of EBS effects & influence of EBS parameters, for future & also other applications in any fields other than epilepsy (such as therapeutic EBS perfomed in different brain pathologies).