As healthy users move their head, such as when walking, their eyes compensate for this motion and enable them to fixate a target. This is how players can keep focused on the goal while sprinting across a soccer field for example. When a person loses function on both sides of vestibular system (i.e., the balance part of the inner ear), they lose their sense of balance as well as the ability to fixate targets during motion, perceiving their vision blurry instead. This is known as bilateral vestibular loss (BVL) or hypofunction (BVH) (S. Robinson, Bilateral Vestibular Hypofunction). One of the main causes of this blurred vision is a malfunction of the compensation mechanism when a person moves their head. In patients with BVS, their eyes moves slower than their head, which leads to erratic eye movements and the inability to keep focused on targets.
In this project, you will develop a compensation algorithm for BVS using Virtual Reality (VR), which will be similar to digital image stabilization. By tracking the user’s eyes and the user’s motions while moving in VR, an algorithm should be developed that stabilizes the image displayed to the user. To inform the development of this algorithm, you will gather and analyze gaze and motion data from healthy users and compare them to patients with BVS. The final algorithm will be evaluated with patients on a treadmill and in a free-walking experiment. This project is in collaboration with the Universitätsspital Zürich (Dr. Andreas Eckhard) and Prof. Christian Holz from ETH Zurich.
The goal of the project is to 1) analyze stabilization mechanisms based on data gathered from healthy users and patients with BVS, 2) implement efficient algorithms that enable real-time compensation, and to 3) test the algorithm with a small set of patients (this evaluation will be conducted together with experimenters at Universitätsspital).
- Literature review on VR and BVS, and analysis of eye movement and gain patterns of earlier experiments
- Implementation of a VR test environment in Unity
- Create VR environment that enables testing of algorithm with multiple visual tests (e.g., Snellen chart)
- Simulate BVS for healthy users (mis-compensation and erratic eye movement)
- Implement calibration procedure to match algorithm to patient parameters
- Evaluation that studies the efficacy of the algorithm
(optional) Depending on the quality of the solution and results, this project may warrant submitting the project to a journal venue
AIT Lab, ETH Zurich
Bilateral Vestibular Loss
IDEA League Student Grant (IDL)
CLS Student Project (MPG ETH CLS)
ETH Organization's Labels (ETHZ)
Information, Computing and Communication Sciences