Research prototype

ARBingo

AI-adaptive spatial neglect training in augmented reality

A headset-based rehabilitation research system designed to make spatial neglect practice active, repeatable, and easier to review across sessions.

01 Active visual exploration

02 Natural hand interaction

03 Adaptive support over time

Visual exploration Cross-midline movement Spatial cueing AI-adaptive support Session continuity

Why it exists

Spatial neglect rehabilitation should feel active, familiar, and repeatable.

ARBingo uses an augmented-reality bingo task to encourage visual exploration and cross-midline movement while preserving the feel of working in a real physical environment.

Clinical context

Designed around spatial neglect after stroke.

The task format supports repeated practice of exploration, reaching, and placement behaviors relevant to supervised rehabilitation research.

Interaction

Physical-room training, not flat-screen tapping.

Patients interact with virtual task objects while remaining seated at a real table, keeping the experience grounded in everyday movement.

Adaptation

Cueing support can change with practice.

The system is structured for patient-specific support across repeated sessions without exposing technical model details on the public site.

Workflow

01

Explore

Search for targets in the surrounding workspace.

02

Move

Reach, grasp, and place objects through natural hand interaction.

03

Adapt

Receive cueing support that can change across repeated sessions.

Status

Currently presented as a research prototype for supervised rehabilitation workflows, not as a standalone medical product.