Anterior temporal epilepsy is the most frequent form of focal epilepsy (more than 50% of focal epilepsies). The epilepsy is drug-resistant when epileptic seizures persist despite antiepileptic treatment (25% of cases). In this case, it is possible to offer a surgical solution to the patient: an anterior temporal lobectomy. This surgery consists of removing the entire portion of the temporal lobe responsible for epilepsy (epileptogenic zone), that is to say a major part of the temporal pole, the hippocampus and the ventral anterior temporal lobe (vATL). The goal of the surgery is to lead to a disappearance of the seizures while preserving the functions of the patient. This is why a pre-surgical assessment is systematically carried out in order to locate the epileptic focus and to predict the risks of neurological and cognitive deficits. The vATL is of particular interest because it is a highly functional region, involved in naming, semantic processing and face recognition. A resection of this region by anterior lobectomy can therefore impact these functions and lead to cognitive deficits (for example, up to 65% decline in naming), which can be disabling in the lives of patients, even if they are free from seizures. One of the major challenges of epilepsy surgery is therefore to predict the postoperative neuropsychological outcome. The prediction of the neuropsychological outcome of the post-surgery patient is largely based on the mapping of functional regions preoperatively, carried out by functional MRI (fMRI), or by electrical stimulation carried out during intracerebral exploration by StereoElectroEncephalography (SEEG). However, current techniques have drawbacks. Electrical stimulations are based on an invasive exploration (SEEG), are time-consuming and sometimes difficult to interpret. The fMRI sequences used in clinical routine do not make it possible to visualize the entire vATL region because of artifacts related to the auditory canal. Thus, the signal is strongly diminished in this region, rendering a large area of the vATL invisible. This results in insufficient visualization of activated vATL regions when performing tasks such as naming, semantic processing, and face recognition. Important functional regions can therefore be removed during surgery and negatively impact the patient's neuropsychological outcome. In the CARTA study, original methods are associated in order to increase the signal-to-noise ratio in vATL. On the one hand, the Multi-Band sequence, an innovative fMRI sequence, will be used. On the other hand, a particular method of presentation of visual stimuli will be used, called fast periodic visual stimulation (FPVS: Fast Periodic Visual Stimulation), during which the stimuli are presented periodically (fixed frequency). Individually, these methods improve vATL (signal enhancement) exploration. The investigators assume that the combination of the two methods may have a potentiating effect, compared to the standard SMS (Simultaneous Multi-Slice) sequence. fMRI exploration will not influence the surgical management of the patient included in the study because it is the beginning of the development of this technique, but could be used, in the longer term, to guide the surgeries of epileptic patients. Thus, the goal of this study is to precisely map the vATL, using innovative methods in fMRI. This mapping will make it possible to study the cerebral functions of the vATL involved in naming, semantic processing and face recognition, and ultimately improve the postoperative neuropsychological prognosis of epileptic patients.