Verbal fluency test require to produce as much words as possible in one or two minutes. This test is highly sensitive to main brain diseases and are therefore widely used in clinical routine for diagnostic purpose. The verbal fluency task requires several cognitive processes including executive and linguistic processes for which it is difficult to extract the origin of the deficit. For this reason, fluency tests are variably interpreted in terms of executive or language. The implementation of an experimental protocol exploring each of these processes separately and studying the links between the verbal fluency task and each of these processes should allow a better understanding of the origin of the verbal fluency deficit after brain injury and improve the identification of key brain structures.

Indeed, the lesion determinants of this task remain to be clarified despite remarkable advances due to the evolution of imaging techniques (voxel-based lesion-symptom mapping (VLSM); voxel-based morphometry (VBM)).

Furthermore, while the direct assessment of linguistic process, semantic memory, and processing speed is well defined, the examination of the executive component (i.e., strategic search process) remains unsettled and will be undertaken in this study.

This work will take advantage of data from previous multicenter work, validated methodologies for both analysis and interpretation of cognitive performance as well as anatomic-clinical correlations at the voxel level and will be performed in cognitive neurodegenerative and cerebrovascular disease.

Inclusion Criteria:

• For Cognitive cortical neurodegenerative diseases PATIENTS
• informed consentPatients between 40 and 85 years old,
• French native language,
• Social Security affiliation,
• Can read, write and count (up to 36) and knows the alphabet
• Assessed in Amiens University Memory Clinic for:
• Mild severity impairment (MMSE> 19) or major severity impairment related to:
• AD according to Albert criteria and McKhann criteria
• DCL according to McKeith criteria
• FTLD according to Rascovsky criteria
• CBD according to Armstrong's criteria
• PSP according to Höglinger criteria
• For Stroke PATIENTS
• informed consent
• Patients between 40 and 85 years old,
• French native language,
• Social Security affiliation,
• Can read, write and count (up to 36) and knows the alphabet
• Having suffered a stroke, hospitalized in neurology, visualized by imaging

Exclusion Criteria:

• Mental retardation or guardianship
• Other current or past brain condition affecting cognition, including:
• Severe head trauma
• Epilepsy prior to stroke still requiring previous treatment
• Parkinson disease, multiple sclerosis
• Brain tumor or brain radiation therapy
• Current or past schizophrenia or psychosis
• Active or past psychiatric impairments requiring a stay> 2 days in a specialized environment
• Contra indication to MRI
• Comorbidity with life expectancy <1 year
• Comorbidity affecting cognition in particular:
• Alcohol (> 3 glasses / day) or history of alcohol withdrawal syndrome
• Opiate or cocaine addiction or opiate withdrawal syndrome
• Renal failure (dialysis or creatinine clearance <30)
• Hepatic failure (spontaneous INR> 1.5 or PT <60%)
• Respiratory failure requiring oxygen therapy
• Heart failure (orthopnea> 2 pillows)
• Persistent vigilance disorder (NIHSS1a score ≤1)
• Cancer with paraneoplastic syndrome
• Treatment with gold salts, D Penicillamine or other treatment with cognitive effect
• Patient under guardianship or curators or private under public law
• Pregnant and / or lactating woman