After a stroke, 80% of patients have an upper limb deficit, limiting activity. Some develop a non-use: they can, but do not, use their paretic limb. Non-use is a general phenomenon applied to all situations where the patient applies unnecessary compensation. Several rehabilitation techniques are effective to counter non-use, but there is insufficient knowledge to choose the most suitable technique. Optimal control theory could help guide these choices. It assumes that the chosen coordination satisfies the constraints of the task (force, amplitude, tolerance) while reducing the cost of the movement. This study will assess non-use by anticipating the sensitivity to the constraints of force and precision deduced from the logic of optimal control. The study authors expect to observe a weakness effect: in a reaching task (i.e. when the person has to touch an object placed in front of them), lightening the paretic arm makes it possible to reduce non-use, and a precision effect: in a reaching task, non-use increases with the required spatial precision.

Inclusion Criteria:

Inclusion criteria specific to the stroke group:

• Patient having had a 1st stroke
• Patient whose stroke occurred more than a month ago.
• Patient having had a supratentorial stroke
• Patient able to touch the opposite knee with the paretic arm

Inclusion criteria specific to the control group:

• Subject who has never had a stroke
• Subject without motor or orthopedic impairment of the upper limbs and in particular of the shoulder

General inclusion criteria:

• Subject who has given free and informed consent.
• Subject who signed the consent form.
• Subject affiliated or beneficiary of a health insurance plan.

Exclusion criteria:

• Pregnant, parturient or breastfeeding subject.
• The subject is participating in a category 1 interventional study, or is in a period of exclusion determined by a previous study
• The subject refuses to sign the consent
• It is impossible to give the subject informed information
• The patient is under safeguard of justice or state guardianship
• Patient with cognitive disorders incompatible with a good understanding of the use of the device
• Patient with other neurological or osteoarticular history that may limit the performance of the task - criterion left to the deiscretion of the investigator

Anonymised raw datasets and code used for the analysis will be available on reasonable request.