Fabian Löbel

Advisors
Wolfgang Mehringer (M. Sc.), Veronika König (M. Sc.), Prof. Dr. Björn Eskofier, Milos Wieczorek, Prof. Dr.-med. Georg Michelson

Duration
10 / 2022 – 04 / 2023

Abstract
Visual field analysis, also called perimetry, is a procedure to screen for visual field abnormalities caused by diseases like Glaucoma [1]. In-office clinical perimetry is commonly performed using systems like the Octopus, Dicon, Henson Perimeters, or the Humphrey Visual Field Analyzer (HFA) [2]. These require patients to schedule a meeting at the medical practitioner’s office. In case of the state-of-the-art devices, a test subject has to physically be able to sit in front of the machine for longer periods of time, with their head in a fixed position [3]. Additionally, to ensure accurate lighting conditions in combination with the open concave canvases the visual field tests have to be performed in darkened rooms. All of these drawbacks usually require the test to be performed in a separate room.

This is improved upon by newer solutions, making use of the improvements of Virtual Reality (VR) headset technology, which can be used outside a clinical environment. Studies using different VR headsets have shown, that similar performances compared to a clinically used perimeter can be achieved, with respect to a number of parameters in healthy and glaucoma patients [4, 5, 6, 7]. Mentioned procedures all use the Standard Automated Perimetry (SAP) algorithms, like Swedish Interactive Threshold Algorithm (SITA). To confirm the event of seeing a stimulus, a physical button is used in most cases. However, this approach has multiple shortcomings, as it allows the subject to trigger a seen event, even though no point should have been visible. Additionally, it inhibits the testing of patients with impaired motor functions.

To circumvent these issues, eye-tracking as an additional interaction method for the test person is already looked at by multiple sources. An eye-tracking based interaction approach, in combination with a common computer monitor, showed results similar to the gold standard HFA in young, normally sighted adults [2, 8]. The recent Saccadic Vector Optokinetic Perimetry (SVOP) [9] method already indicates agreement with the gold standard HFA method, which commonly uses the button pressing confirmation [8].

To make use of the mentioned advantages of VR combined with the advantages of eye-tracking, a comparison of button pressing and an eye tracking based trigger interaction was already attempted, which concluded that the latter approach is preferred by the patients and yields similar results [10]. This conclusion was based on an early VR headset, with limited eye tracking capabilities, and also lacked test type randomization during the executed study.

Rapid development in the world of consumer VR headsets, has led to eye-tracking capable devices getting more affordable, such as the HTC Vive Pro Eye. Recent studies showed, that this model only has a ±25◦ area, in which the eye-tracking precision is accurate enough for visual field testing [11]. Nevertheless, as it is one of the more common headsets with built-in eye-tracking, it is chosen for following study proposition, which addresses the different ways of triggering a “seen” event. The study will compare the different interaction types, to reevaluate which is more suitable for VR based perimetry. Confirmation procedures taken into account are manual confirmation, gazebased confirmation as well as a hybrid approach. The test subject satisfaction will be evaluated in combination with the test result, to point out benefits and disadvantages of the respective approaches as well as their possible applications. The starting point will be the Unity (runtime and development environment for games) based perimetry framework, previously developed in a project by the author of this thesis [12].

References:
[1] Tatham Andrew J., McClean Pam, Murray Ian C., et al. Development of an Age-corrected Normative Database for Saccadic Vector Optokinetic Perimetry (SVOP) Journal of Glaucoma. 2020.
[2] Jones Pete R.. An Open-source Static Threshold Perimetry Test Using Remote Eye-tracking (Eyecatcher): Description, Validation, and Preliminary Normative Data 2020;9:18–18.
[3] Racette Lyne, Fischer Monika, Bebie Hans, Holló Gábor, Johnson Chris, Matsumoto Chota. Visual Field Digest . 2019.
[4] Stapelfeldt Jan, Kucur S˛erife Seda, Huber Nina, Höhn René, Sznitman Raphael. Virtual Reality-Based and Conventional Visual Field Examination Comparison in Healthy and Glaucoma Patients 2021;10:10–10.
[5] Mees Lukas, Upadhyaya Swati, Kumar Pavan, et al. Validation of a Head Mounted Virtual Reality Visual Field Screening Device 2019;29:1.
[6] Shetty Vijay, Sankhe Prachi, Haldipurkar Suhas S, et al. Diagnostic Performance of the PalmScan VF2000 Virtual Reality Visual Field Analyzer for Identification and Classification of Glaucoma Journal of Ophthalmic and Vision Research. 2022.
[7] Preliminary Report on a Novel Virtual Reality Perimeter Compared With Standard Automated Perimetry ;30.
[8] Tatham Andrew J., Murray Ian C., McTrusty Alice D., et al. A case control study examining the feasibility of using eye tracking perimetry to differentiate patients with glaucoma from healthy controls Scientific Reports. 2021;11:839.
[9] Murray Ian C., Perperidis Antonios, Cameron Lorraine A., et al. Comparison of Saccadic Vector Optokinetic Perimetry and Standard Automated Perimetry in Glaucoma. Part I: Threshold Values and Repeatability Translational Vision Science & Technology. 2017;6:3-3.
[10] Wroblewski Dariusz, Francis Brian A., Sadun Alfredo, Vakili Ghazal, Chopra Vikas. Testing of Visual Field with Virtual Reality Goggles in Manual and Visual Grasp Modes 2014;2014:206082. Publisher: Hindawi Publishing Corporation.
[11] Sipatchin Alexandra, Wahl Siegfried, Rifai Katharina. Eye-Tracking for Clinical Ophthalmology with Virtual Reality (VR): A Case Study of the HTC Vive Pro Eye’s Usability 2021;9.
[12] Löbel Fabian. Modular framework for perimetry using a virtual reality headset Department Artificial Intelligence for Biomedical Engineering, Machine Learning and Data Analytics, Friedrich-Alexander-Universität Erlangen-Nürnberg 2022.