Navigating according to a specific goal is a common activity of everyday life. Spatial navigation requires the implementation of motor and perceptual functions (sight, walking, proprioception), but also various cognitive functions (executive functions, memory, spatial orientation skills). Many people affected by a neurodegenerative disease have topographical difficulties which have a major impact on their autonomy in daily life, by gradually limiting their movements outside their home, then inside their home, and which are the main factor leading to the institutionalization of this population. People with dementia or MCI (Mild Cognitive Impairment) of the Alzheimer type, according to the definition of the NIAA (National Institute on Aging-Alzheimer's Association) have navigation and spatial memory disorders, with difficulty in acquiring mental spatial representations of their environment. These topographical difficulties have been shown to be related to the reduction in volume of the temporal cortex, in particular the hippocampal regions, as well as to atrophy of the retrosplenial cortex.

Augmented Reality (AR), often defined as an intermediary between the totally artificial world of VR and the real world in which we operate, makes it possible in particular to add summary information to the natural environment in which a participant operates. Even if, in particular in outside environment, AR must solve many challenges, such as the integration of the real and virtual worlds in real time, the selection of the modalities of restitution of information, its use is exponential in the medical field, in particular in surgery for the assistance of the practitioner, but also in the field of sensory substitution, in particular to facilitate the movements of people with visual impairment. Other works focused in helping people with dementia of the Alzheimer type, such as those of Quintana and Favela (2012) who proposed preliminary systems of annotations in AR. Hervás et al. were the first in 2014 to test the use of augmented reality to provide navigation assistance to people with dementia. In 2017, Firouzian et al. as well as Sejunaite et al. implemented related systems. Firouzian et al. have developed spectacle frames comprising around ten LED lights in order to provide directional indications to people who moved outdoor. However, the influence of this system on navigation performance has not been tested yet. On the other hand, although simple to develop, this system requires training on the part of the users and only makes it possible to provide directional information, which is not recommended for the implementation of a tool for this population. Finally, Sejunaite et al. used an environmentally tested smart glasses to allow users to display information in the form of a map to help older people navigate independently. However, the literature review indicates that even increased card use does not seem to be suitable for people with dementia or Alzheimer's-type MCI. Finally, these two tools do not allow navigational information to be co-located in the field of vision of people, which represents one of the major advantages of augmented reality. To our knowledge, there is not yet an AR device providing co-located information in the environment dedicated to outdoor navigation of people with dementia or Alzheimer's type MCI.

The ARIADE project was preceded by:

• A phase of systematic literature review and selection of the different types of visual aids to be tested (completed)
• A design phase with implementation of these visual aids in augmented reality and determination of methods for measuring the effectiveness of devices for detecting wanderings (criteria linked to changes in gait: length of steps, frequency, regularity, direction , pauses) with detection using accelerometers (completed)
• A phase to assess the a priori acceptability of an Augmented Reality device by people with Alzheimer's disease or an Alzheimer-type MCI and by their relatives (completed)

The ARIADE project (9 months) will take place in a controlled and reproducible ecological environment (Ker Lann gymnasium) and will aim at assessing the effectiveness of an Augmented Reality assistance, and of the devices adapted for detecting wandering, the safety of the patient when traveling with an Augmented Reality headset, and their acceptance of the device.

Three routes each comprising seven intersections will allow us in 20 patients to compare the three different visual aids offered in augmented reality, i.e. arrows, path illumination and a little companion. We will assess the effectiveness of these aids to complete the course without trajectory errors, the safety of the participant throughout the course, the effectiveness of the system for detecting wanderings, and the acceptance of the device by the participant.

The participation of a patient in this phase will not exceed one hour, including breaks.

Patients will be called beforehand by one of the study investigators from a list of patients who volunteered to participate in clinical research studies held by the CMRR of Rennes University Hospital. Verification of the inclusion and non-inclusion criteria will be carried out by CMRR's investigating physicians from the patient's paper file and then during a usual follow-up consultation. The travel of participants to the Ker Lann gymnasium will be compensated.

Visual aids tested:

Based on previous experimental results, we aimed at testing the usability of three types of directional stimuli, i.e. information provided regarding the location of the user as well as the direction regarding the path to follow to reach a destination.

1. The first type of aid proposed, which consists of the use of a directional arrow, is classic since it is the one whose use is most regularly reported in the literature. Arrows will be placed at ground level at each intersection.
2. The second type of assistance consists of a path illumination. This type of help has only been tested with people with dementia as part of a single decision-making task. However, previous studies based on the illumination of certain areas indicate that this method is promising.
3. Finally, the third type of help consists of the use of an animated companion standing in front of the person. This character does not interact with the person. However, it provides a social presence during the ride. The use of a personal aid should help on reducing the anxiety generated by navigation tasks. In the long term, and according to the results of this first study, the use of a learning companion, ie a type of educational agent that can provide a complex form of social presence and emotional feedback in a controlled environment, could be considered. To the best of our knowledge, a personified navigation aid has never been tested with people with dementia.

Exploratory, prospective, single-center, pilot study.

Inclusion Criteria:

• Existence of a probable Alzheimer's disease (AD) according to the NIAA criteria (McKhann et al., 2011) with a Mini Mental State Examination (MMSE) greater than or equal to 20 (threshold defined in order not to include certain participants whose comprehension problems would not be compatible with the ARIADE project) OR Existence of an Alzheimer-type Mild Cognitive Impairment (MCI) according to the criteria of Albert et al. (2011)
• Adult person
• Affiliation to a social security scheme
• Free, informed and signed written consent The diagnosis of dementia or MCI of the probable Alzheimer type will be made by the medical and paramedical team of the Memory, Resources and Research Center (CMRR) of the Rennes University Hospital (Department of Dr Serge Belliard). The proposal for voluntary participation in the ARIADE project will be made by one of the project's clinical investigators to CMRR patients who have already expressed their interest in participating in a clinical research project (list already updated by the CMRR team) and to members of the Association Bistrot Mémoire de Rennes, partner of the ARIADE project, after validation of the inclusion and non-inclusion criteria by the CMRR team at Rennes University Hospital.

Exclusion Criteria:

• Clinical history of neurological event in the central nervous system
• Untreated anxiety-depressive syndrome with a sub-score A (anxiety) or D (depression) on the HADS questionnaire, i.e. Hospital Anxiety and Depression Scale, greater than or equal to 11
• Moderate or greater deafness (loss of at least 40dB from the "good" ear) or an asymmetry of more than 10dB between the left and right ear on tone audiometry
• Insufficient corrected visual acuity (decree of August 31, 2010 for driving a light vehicle) (incompatibility if binocular visual acuity is less than 5 / 10th OR if the acuity of one eye is less than 1 / 10th and the acuity of the other eye is less than 5 / 10th)
• Severe comprehension disorder, preventing the understanding of the instructions
• Severe behavioral disorder, not allowing participation in the experiment (DSM-V)
• Acute decompensated pathology (example: heart failure)
• Lack of social security
• Safeguarding effective justice
• Severe balance disorders noted when performing a Timed Up and Go Test with a risk of falling noted by the assessor or an inability to perform a 6-minute walk test.
• Non-French speaking person
• Pregnant or breastfeeding woman

The existence of a simple curatorship, a reinforced curatorship or a tutorship will not constitute a criterion for non-inclusion. Indeed, people with a neurocognitive disorder under legal protection are those who probably present the most loss of autonomy and loss of social participation and inclusion in the city, in connection with the disease and / or with isolation. social and family. These are therefore the people most in need of technological aids that we are going to offer. It therefore seems illogical to exclude them insofar as they will represent the sub-part of this population that will need them the most and will probably derive great benefit from them, with minimal risks in this study.