Inclusion rate too slow / infeasible.

The main objective of this study is to determine among the patients included a subpopulation that does not need a defibrillator.

This will be done by comparing the number of discordant segments (mismatch) between patients who have a severe arrhythmia (ventricular tachycardia or ventricular fibrillation) with appropriate electrical therapy within 3 years of monitoring and others. "Mismatch" in the definition of this objective is the usual definition: score 3.4 in I123-MIBG (Iodine 123 metaiodobenzylguanidine) and PET at 0,1,2.

Inclusion Criteria:

• Patient with ischemic heart disease with heart failure
• Patients who received myocardial perfusion scintigraphy
• Indication for the implantation of an ICD installed according to current recommendations : LVEF <30% and NYHA (New York Heart Association) stage II, III.
• The subject must have given his informed consent and signed consent
• The subject must be a member or beneficiary of a social security system

Exclusion Criteria:

• Pregnant or lactating women
• History of "open heart" cardiovascular surgery
• Inability to sign informed consent (patient under guardianship)
• Patients with diabetes mellitus
• Patient with proven dementia
• Person refusing to sign the consent
• Someone who already has an ICD at baseline
• Life expectancy less than one year (NYHA IV, other diseases such as neoplastic disease in an advanced stage)

Coronary patients that should benefit from ICD (Implanted Cardiac Defibrillator) implantation for primary prevention according to current recommendations, that is to say with LVEF (left ventricular ejection fraction) <30% in NYHA (New York Heart Association) stage II or III and within one month of an myocardial infarction or within 3 months of a complete revascularization.