Stereoelectroencephalography — SEEG — changed epilepsy surgery by allowing physicians to record seizures from multiple structures inside the brain in three dimensions.
Its importance was not simply that electrodes were placed deeper than scalp EEG could reach. The method developed by Jean Talairach, Jean Bancaud and their colleagues at Sainte-Anne Hospital in Paris was built around a different way of thinking about epilepsy.
The starting point was a clinical hypothesis. Which structures might be involved in the patient’s seizures? In what order were they recruited? How did the electrical discharge relate to the symptoms seen during the attack?
The electrodes were then placed to test that hypothesis.
This page follows the historical development of SEEG as a method. The site’s separate modern SEEG page explains how intracranial recordings are used in present-day epilepsy care.
Before SEEG: localisation was already changing epilepsy surgery
By the first half of the twentieth century, epilepsy surgery had already moved far beyond simply operating on visible scars or tumours.
Hughlings Jackson’s clinical localisation, experimental cortical stimulation, electrocorticography and the work of centres such as the Montreal Neurological Institute had shown that seizure symptoms could provide clues to where abnormal activity began and spread.
But important limitations remained.
Scalp EEG had limited ability to detect activity arising from deep or medial structures.
Electrocorticography recorded directly from the exposed cortical surface but usually during an operation and mainly from regions accessible to the surgeon.
Subdural electrodes sampled the brain surface well, but epilepsy did not always respect the surface.
Some seizures seemed to involve deep structures or networks extending across several lobes.
The problem was becoming three-dimensional.
Stereotaxy provided the technical foundation
SEEG could not have developed without stereotactic neurosurgery.
Stereotaxy uses a coordinate system to guide instruments accurately to targets inside the brain.
Experimental stereotactic apparatus had existed since the work of Victor Horsley and Robert Clarke in the early twentieth century, but human stereotactic neurosurgery expanded rapidly in the late 1940s.
Jean Talairach became one of the central figures in that development in France.
Jean Talairach
Jean Talairach began his medical career at Sainte-Anne Hospital in Paris and initially trained in psychiatry before moving increasingly toward neurosurgery and stereotactic methods.
He developed his own stereotactic apparatus, reference systems and anatomical methods rather than simply adopting existing frames.
Historical work on Talairach describes one of his early human stereotactic procedures in 1948.
His larger contribution, however, was the construction of a method for relating an individual patient’s brain anatomy to reproducible three-dimensional coordinates.
1952: the anterior–posterior commissural reference
In 1952, Talairach proposed using the anterior commissure and posterior commissure as internal reference landmarks for stereotactic localisation.
The line connecting these structures — the AC–PC line — became one of the most influential reference systems in human brain mapping.
This was important because skull measurements alone did not provide sufficiently reliable information about where internal cerebral structures were located.
By anchoring the coordinate system to structures inside the brain, Talairach could account better for differences between individual heads and brains.
1957: a stereotactic atlas
Talairach and colleagues published an important stereotactic atlas of deep grey structures in 1957.
The atlas drew on extensive anatomical work and used the AC–PC reference system together with Talairach’s frame and grid apparatus.
Its significance went beyond epilepsy.
It provided a framework for stereotactic neurosurgery, deep-brain targeting and later brain mapping.
For epilepsy investigation, it supplied something equally important: a geometrical language for saying precisely where an intracerebral electrode contact was located.
Jean Bancaud brings the epilepsy question
Jean Bancaud was a neurologist and epileptologist whose thinking had been influenced by the developing field of cortical localisation, including the work of Wilder Penfield.
He became convinced that existing electrophysiological techniques could not adequately explain many complex focal seizures.
A scalp trace might show that a seizure was lateralised without revealing its deep origin.
A surface electrode might detect activity only after a discharge had already propagated from another structure.
Bancaud’s clinical questions and Talairach’s stereotactic technology were therefore unusually complementary.
Sainte-Anne becomes the laboratory
At Sainte-Anne Hospital, Talairach and Bancaud worked with a multidisciplinary group that included Georges Szikla, Alain Bonis, Pierre Tournoux and others.
The centre already had a developing epilepsy surgery programme.
What emerged there in the 1950s was not merely a new type of electrode implantation.
It was a complete investigative methodology combining clinical seizure analysis, stereotactic anatomy, intracranial electrical recording and electrical stimulation.
Why depth electrodes were not themselves new
Talairach and Bancaud did not invent the general idea of placing electrodes inside the brain.
Earlier investigators had used depth recordings.
For example, chronic intracerebral recordings had been attempted in the 1940s, and stereotactic instruments for human neurosurgery were being developed by several groups internationally.
The distinctive contribution of the Sainte-Anne group was to turn multiple intracerebral recordings into a systematic, anatomically defined, hypothesis-driven method for epilepsy surgery.
The defining idea: anatomo-electro-clinical correlation
The Sainte-Anne method tried to correlate three things.
Anatomy: exactly which cerebral structures were being recorded.
Electrical activity: where the seizure discharge appeared and how it propagated.
Clinical semiology: what the patient experienced and what observers saw as those electrical changes unfolded.
This became known as anatomo-electro-clinical correlation.
The method asked not merely, ‘Where is there an abnormal spike?’ but, ‘How does activity in this structure relate in time to this particular seizure symptom?’
The implantation was planned from a hypothesis
One of the most important principles of SEEG was that electrodes were not supposed to be scattered randomly through the brain in the hope of finding a seizure.
Before implantation, the team formed a hypothesis from the patient’s seizure history, semiology, scalp EEG and available anatomical information.
The electrode plan was then designed to test competing possibilities.
If temporal epilepsy was suspected, the investigation might sample several temporal structures and connected regions.
If frontal, insular or multilobar involvement was possible, the implantation had to be designed differently.
This principle remains one of the historical hallmarks of SEEG.
How electrodes were placed before CT and MRI
Modern SEEG planning benefits from high-resolution MRI, CT angiography, computerised trajectory planning and robotic guidance.
The original investigators had none of these.
They relied on stereotactic radiography, ventriculography, angiography, anatomical atlases and Talairach’s coordinate system.
Ventriculography helped identify internal landmarks such as the ventricular system and commissural structures.
Angiography was critical for identifying blood vessels so that electrode trajectories could be planned to avoid them.
The procedure required painstaking geometrical reconstruction.
The Talairach frame and double-grid system
Talairach developed a stereotactic frame and a double-grid system that allowed trajectories to be planned and reproduced in three-dimensional space.
The grids helped define entry points and directions relative to the patient’s individual stereotactic anatomy.
Electrodes could then pass through multiple structures along one trajectory, with several contacts recording at different depths.
This allowed one electrode to sample a sequence of cortical and deeper regions rather than only a single point.
Why three dimensions mattered
A scalp EEG is largely interpreted as a projection of electrical activity recorded from the head surface.
SEEG approached the problem differently.
An array of intracerebral contacts could sample medial temporal structures, insula, cingulate cortex, deep sulci and other regions that surface recordings might poorly represent.
The epileptic process could therefore be reconstructed through depth as well as across the cortical surface.
This three-dimensional view became one of SEEG’s defining characteristics.
1962: ‘functional stereotaxic exploration of epilepsy’
In 1962, Talairach, Bancaud, Bonis, Szikla and Tournoux published Functional Stereotaxic Exploration of Epilepsy.
The period around this publication is usually associated with the formal emergence of the term stereo-electroencephalography or stereoencephalography for their method.
By then the group had moved well beyond technical proof that intracerebral recordings were possible.
They were using the recordings to study seizure organisation and guide surgical decisions.
SEEG was designed to record seizures, not just spikes
Interictal epileptiform abnormalities were useful, but Bancaud and Talairach were especially interested in the seizure itself.
Where did ictal activity begin?
Which contacts became involved next?
At what point did the patient’s aura, loss of responsiveness or motor behaviour occur?
The temporal order mattered.
A structure showing spikes between seizures was not automatically assumed to be the true origin of the habitual attack.
This distinction helped separate an irritative region from the organisation of the seizure itself.
Electrical stimulation became part of the method
The implanted electrodes could also stimulate brain tissue.
Stimulation served several purposes.
It could identify regions involved in movement, sensation, language or other functions.
It could produce sensations or experiences resembling part of a patient’s usual seizure.
And under carefully controlled conditions it could provoke electrical or clinical responses that helped the investigators understand functional relationships between sampled structures.
Recording and stimulation therefore complemented one another.
The patient’s seizure became a sequence through space and time
This was one of the most important conceptual advances.
Earlier localisation often sought a single cortical point.
SEEG made it easier to see a seizure as a process.
An electrical discharge might begin in one structure, recruit another within fractions of a second, spread through a connected pathway and only then produce the symptom visible to an observer.
The seizure was therefore not only somewhere.
It unfolded.
Bancaud and the epileptogenic zone
The Sainte-Anne school helped develop the concept of the epileptogenic zone.
In this tradition, the relevant surgical target was not defined solely by the first electrode contact to change.
It was the cortical territory necessary for the generation of the patient’s seizures and whose removal or disconnection would be required for seizure freedom.
This was a more sophisticated concept than a single ‘epileptic spot’.
It acknowledged that the visible onset, symptom-producing regions and broader epileptogenic organisation might not be identical.
A precursor to the network view of epilepsy
Talairach and Bancaud did not use modern network neuroscience in the contemporary computational sense.
But their recordings encouraged a dynamic view of focal epilepsy.
Seizures involved relationships between structures.
Propagation pathways mattered.
Symptoms could arise when the discharge recruited a region some distance from the earliest recorded activity.
This made the Sainte-Anne tradition an important historical precursor to later concepts of epileptic networks.
Frontal and temporal epilepsies were re-examined
During the 1960s and 1970s, the Sainte-Anne group used SEEG to investigate complex temporal and frontal lobe epilepsies.
These studies helped demonstrate how similar outward seizure behaviours could arise through different anatomical pathways.
They also revealed that some regions could participate in seizures without producing obvious early symptoms.
Such ‘silent’ involvement was particularly important before CT and MRI could show subtle structural abnormalities.
The insula became accessible
Deep and opercular structures posed a particular challenge to surface-based methods.
The insula lies buried within the Sylvian fissure beneath the frontal, parietal and temporal opercula.
Intracerebral stereotactic electrodes allowed investigators to sample regions such as the insula directly.
This contributed to the gradual recognition that some seizures attributed to neighbouring lobes could involve deeper cortex.
SEEG therefore expanded not only the accuracy of localisation but the anatomical territory that could realistically be investigated.
SEEG and the Montreal tradition were different, not opposites
The Montreal Neurological Institute and Sainte-Anne developed different but overlapping approaches to presurgical localisation.
Montreal became strongly associated with cortical stimulation, electrocorticography and mapping exposed cortex during surgery.
The Paris school emphasised stereotactic implantation and prolonged three-dimensional intracerebral analysis.
It would be misleading to portray this as a simple rivalry in which one method replaced the other.
Both traditions contributed fundamentally to epilepsy surgery, and later centres combined elements of both.
Why SEEG did not immediately spread everywhere
For decades, SEEG remained concentrated mainly in France and a relatively small number of European and Canadian centres.
There were practical reasons.
The original methodology was technically demanding.
It required specialised stereotactic expertise, detailed anatomical planning, invasive implantation and a team experienced in interpreting intracerebral recordings.
Other centres developed strong traditions using subdural strips and grids instead.
Institutional training patterns therefore shaped which invasive method became dominant in different countries.
Claudio Munari and the Milan school
An important later chapter occurred in Italy.
Claudio Munari trained within the French SEEG tradition and helped establish the methodology in Milan.
The Italian school became one of the major centres preserving and developing SEEG while the technique remained relatively uncommon elsewhere.
This continuity proved important when international interest in SEEG increased again decades later.
CT changes stereotactic planning
The introduction of computed tomography in the 1970s began changing intracranial electrode planning.
For the first time, cross-sectional imaging could show brain structures and lesions directly during life without relying entirely on ventriculographic and angiographic reconstruction.
CT did not immediately eliminate older stereotactic methods, but it made anatomical localisation more direct.
The relationship between electrical data and visible pathology could now be examined in new ways.
MRI changes it again
MRI later transformed presurgical evaluation even more profoundly.
Many lesions that had been invisible to previous imaging became detectable.
This altered the role of invasive investigation.
Some patients who once would have required extensive exploratory intracranial recording could now undergo a more targeted evaluation.
At the same time, MRI made SEEG planning more anatomically precise and allowed suspected deep and subtle abnormalities to be sampled directly.
The method survives technological change
Modern SEEG bears little visual resemblance to the original Sainte-Anne procedure.
Computerised planning has replaced hand-drawn geometrical reconstruction.
CT and MRI have replaced ventriculography for most anatomical planning.
Digital angiographic imaging helps identify blood vessels.
Robotic systems can guide electrode trajectories rapidly and reproducibly.
Yet the core intellectual method remains recognisably Talairach and Bancaud’s.
Start with a hypothesis.
Sample the structures needed to test it.
Record habitual seizures.
Correlate anatomy, electrical activity and clinical symptoms.
Use the result to decide what can safely be treated.
SEEG is not simply ‘depth EEG’
This distinction has been emphasised repeatedly by practitioners in the Talairach–Bancaud tradition.
A depth electrode is a device.
SEEG is a methodology.
Simply inserting an intracerebral electrode does not reproduce the original SEEG approach.
The defining elements include hypothesis-driven implantation, stereotactic anatomical definition, electroclinical correlation and interpretation of the seizure as a spatial and temporal network process.
Why the revival came so late
From the 2000s onward, SEEG expanded rapidly beyond its traditional European centres.
Several technological developments helped make that possible.
High-resolution MRI improved target selection.
Digital planning software simplified trajectories.
Robotic stereotactic systems reduced implantation time.
Modern recording systems could handle large numbers of intracranial channels.
Experience also accumulated showing that stereotactic depth-electrode investigations could be performed with acceptable complication rates in specialist centres.
The method that had been developed in Paris more than half a century earlier became increasingly international.
The terminology: SEEG, sEEG and stereo-EEG
Historical and modern publications use several forms of the abbreviation.
SEEG, sEEG and stereo-EEG can all be encountered.
The original French tradition often used stéréo-électroencéphalographie.
The variation in spelling does not indicate fundamentally different methods, although centres may differ considerably in implantation strategy and interpretation.
What Talairach and Bancaud did not invent
They did not invent EEG.
They did not perform the first intracranial electrical recording.
They did not originate cortical electrical stimulation.
They did not invent stereotaxy itself.
Their contribution came from integrating stereotactic anatomy, multiple intracerebral electrodes, seizure semiology, electrical recording and stimulation into a coherent presurgical method for focal epilepsy.
That synthesis is what made SEEG historically distinctive.
The central historical lesson
The history of SEEG is not primarily the story of putting electrodes deeper into the brain.
It is the story of changing the question.
Instead of asking only, Where is the abnormal EEG?, Talairach and Bancaud asked, How is this individual patient’s seizure organised through anatomical space and time?
The difference was profound.
A seizure could have an origin, a pathway of propagation, symptom-producing regions and functionally critical structures that were related but not identical.
SEEG provided a method for studying those relationships directly.
Decades before modern network neuroscience became routine language, the Sainte-Anne school was already treating focal epilepsy as a dynamic three-dimensional process rather than a single electrical spot.
Sources and further reading
Stereoelectroencephalography at Sainte-Anne Hospital, Paris, France. Neurophysiologie Clinique. 2025.
The start and development of epilepsy surgery in Europe: a historical review. Neurosurgical Review. 2015.
Jean Talairach (1911–2007): A life in stereotaxy. Human Brain Mapping. 2019.
Jean Talairach: a cerebral cartographer. Journal of Neurosurgery. 2020.
Jean Talairach (1911–2007): an untold story of the pioneer of stereotactic and functional neurosurgery. 2022.
Talairach J, Bancaud J, Bonis A, Szikla G, Tournoux P. Functional stereotaxic exploration of epilepsy. Confinia Neurologica. 1962.
Bancaud J, Talairach J. Methodology of stereo-EEG exploration and surgical intervention in epilepsy. 1973.
The history of invasive EEG evaluation in epilepsy patients. Seizure. 2016.
Stereoelectroencephalography in epilepsy, cognitive neurophysiology and psychiatric disease: safety, efficacy and place in therapy. 2019.
French guidelines on stereoelectroencephalography (SEEG). Neurophysiologie Clinique. 2018.
Historical evidence checked: September 2026.