The Cerebral Amyloid angiopathy (CAA) is the leading cause of cortical hemorrhage after 65 years. The presence of cerebral infarction is also reported anatomically in the AAC. MRI studies of these infarcts are rare. They are described as punctate, cortical silent. Frequency and pathophysiology is poorly understood. The investigators put the question of a link with hemorrhagic lesions of the AAC.

Main objective / secondary

The objectives are:

• To assess frequency of cortical infarcts in the cohort of patients recruited consecutively likely AAFC GHPSJ since 2007
• To assess the link between the presence of myocardial infarction and the clinical characteristics of patients and
• To study the relationship between topography and the brain hemorrhage one hand, meningeal hemorrhage other.

Inclusion / exclusion Any patient who was diagnosed as carrying a probable AAC according to the Boston criteria and has had a brain MRI with the following sequences: classic or enhanced diffusion (or DTI B2000), T1, T2 FLAIR, T2EG (T2 * or SWAN)

Methodology This is a non-interventional study single center, including AAC patients hospitalized in the Hospital Group Paris Saint-Joseph from May 2007 to May 2014.

Clinical patient characteristics were collected from their medical records. Patients are aware of the potential use of their data for medical research by information contained in the handbook of the institution.

Brain MRI will be proofread by a neurologist and a neuroradiologist to clarify:

• the number and location of myocardial puncture
• the number and location of macro-bleeding
• the number of microbleeds (micro-bleeding)
• the presence and location of subarachnoid hemorrhage and / or hemosiderosis
• the location of the puncture infarction compared to macro-hemorrhage and subarachnoid hemorrhage / hemosiderosis: ipsilateral <5cm, ipsilateral> 5cm, another location The clinical data and MRI will be entered on a file, with data anonymisation. Statistical analysis will be done by a neurologist service.

Number of topics:

• It is estimated that enrollment of patients with AAC by the UNV GHPSJ is approximately 10 patients per year. The number of screened patients during the study period should be about 70.
• Taking into account those whose MRI will be judged of insufficient quality (excluded), the number of patients included in the analysis should be around 50.

Inclusion Criteria:

• Any patient who was diagnosed as carrying a probable AAC according to the Boston criteria and has had a brain MRI with the following sequences: classic or enhanced diffusion

Exclusion Criteria:

• No conditions

Any patient who was diagnosed as carrying a probable AAC according to the Boston criteria and has had a brain MRI with the following sequences: classic or enhanced diffusion