The S-ICD defibrillator developed by Boston Scientific Inc. is a fully subcutaneous automatic defibrillation system that prevents and treats sudden cardiac arrest. The subcutaneous system offers the advantage of avoiding the risks associated with transvenous access, limiting serious infections and lead failures. The detection system is based on the identification of rapid ventricular arrhythmias using one of three bipolar vectors defined by the subcutaneous lead (distal and median electrode) and the generator. However, before implanting the device in a patient, it is essential to carry out a screening procedure to ensure that the patient's QRS signal amplitude is sufficient to be detected by the S-ICD. This screening procedure is carried out by recording a specific three-lead electrocardiogram (ECG), which represents the three bipolar vectors of the S-ICD. The aim of this research is to show that the recording of a standard 12-lead ECG can be sufficient to predict the eligibility of patients for S-ICD implantation. More specifically, it aims to identify electrocardiographic markers on a standard 12D-ECG that can be used to determine which patients may benefit from implantation of a subcutaneous defibrillator, thereby eliminating the need for a 3-lead screening ECG. The developments that will be carried out will be based on the development of signal analysis algorithms and decision support using artificial intelligence.

Patients who meet the study's eligibility criteria will receive an information note describing the study by email or post one to two weeks before their cardiology consultation. Patients will be offered inclusion in the study by their cardiologist during a follow-up cardiology consultation as part of their routine care. During this visit, and for patients meeting the eligibility criteria, oral consent will be obtained. The patient will then be included in the study.

Two arms will be set up depending on the screening result: a group with patients who screened negative and a group with patients who screened positive.

No follow-up is planned as part of the study. Clinical data will be collected in a parameterised and secure eCRF. Data analysis: the 12-lead ECG scans will be analysed using deep learning methods (convolutional neural network), which will be prototyped and trained to predict the outcome of the S-ICD screening with a high accuracy score.

Inclusion Criteria :

• Patient undergoing cardiology consultation, whatever the pathology
• Patient over 18 years of age
• Patient having been informed and having given their oral non-opposition
• Patient and relatives affiliated to a social security scheme

Exclusion Criteria :

• Patients receiving cardiac stimulation
• Patient taking part in a therapeutic trial which may interfere with the results of the research
• Patients under guardianship.

Adult patients undergoing cardiology consultation and not having a pacemaker