Out of 30,000 new cases per year in France, 30% of epileptic patients are drug-resistant. Neurosurgery, which consists in resecting the epileptogenic zone, is the only chance of cure. In the case of temporal epilepsy of the language-dominant hemisphere (TLE), this procedure presents a high risk of increasing cognitive difficulties and may even be contraindicated for this reason alone. The difficulties found are impairments in lexical access (anomia) and verbal memory and affect more than 60% of patients. Preoperative cognitive rehabilitation could influence brain plasticity mechanisms but there are currently no recommendations on this topic. In this context, a speech rehabilitation procedure specific to the needs of ELTPR patients was developed. Investigators rely on cognitive hypotheses explaining the disorders but also on models of rehabilitation-induced neural plasticity likely to improve cognitive reserve before surgery. Investigators hypothesize that preoperative cognitive language rehabilitation in ELTPR patients may decrease surgical risk and improve postoperative language prognosis.

The main objective is to demonstrate the efficacy of preoperative speech therapy on language performance and to evaluate possible protective effects on postoperative language prognosis.

Single case study following the Single Case Experimental Design (SCED) methodology involving the prior definition of the following elements: a repeated measure of the target behavior (naming abilities), the sequential introduction of an intervention (speech therapy), whose effect will be evaluated according to SCED specific analysis and statistics (visual analysis, Tau -U, randomized tests).

Investigators expect patients' naming performance to be stable before the introduction of speech therapy. It is expected that patients will progress in the trained words from the beginning of speech therapy. Finally, in the postoperative period, investigators predict that for the trained words, patients will show performances superior or equal to the pre-rehabilitation period. This result would support a protective effect of preoperative speech therapy.

Inclusion Criteria:

• Patient of legal age,
• Patient with signed informed consent,
• Patient whose epileptogenic area involves the temporal structures of the hemisphere specialized for language,
• Patient with known hemispheric specialization for language (fMRI),
• Patient who is a surgical candidate and accepts resective surgery,
• Patient for whom a surgery date compatible with the study design has been set;
• Patient of French nationality and mother tongue (1st language learned by the patient),
• Patient affiliated to or benefiting from a social security system,
• Patient declaring to be familiar with the use of a computer and having access to an Internet connection from home.

Exclusion Criteria:

• Pregnant or breastfeeding patients,
• Patients deprived of liberty,
• Patients under guardianship or curatorship
• Patients with speech disorders that may impair intelligibility and compromise the use of the interface,
• Patients with uncorrected hearing impairment,
• Patients with low overall intellectual functioning: total intelligence quotient (IQ) <70 (data from the preoperative neuropsychological assessment performed in the year preceding the surgery)