2022-01-21: Summary of Eye Movement and Pupil Measures: A Review

The increasing ubiquity of gaze-aware technologies (commonly referred to as eye trackers)  have enabled assessing patterns in our complex visual interaction with the surrounding world. We can characterize these complex interactions through gaze and pupil signals and extract them through eye trackers. Recent technological advancements in computer vision, image processing, and deep learning have led to the development of eye-trackers with a wide range of utility from consumer to research applications. As a result, gaze-aware technologies have the capability of mass adoption. 

Analysis of oculomotor events (left), analysis methods (center), and measures (right)

With the increasing utility of eye-tracking measures in human-computer interaction and research, a practitioner needs an aggregate body of knowledge. Our study, Eye Movement and Pupil Measures: A Review, tries to address the requirement by reviewing a selection of relevant prior research studies. For this purpose, we first identify and discuss the key oculomotor events and measurable properties during our visual interactions. Then we review the eye movement and pupillometric measures along with their applications. We conclude the study by discussing the recent developments, limitations, potential challenges, and our recommendations with the measures. 

We conducted the study in collaboration with Sundararaman Rengarajan (Physical Therapy, Movement & Rehabilitation Sciences), Dr. Leanne Chukoskie (Physical Therapy, Movement & Rehabilitation Sciences and Art and Design) from Northeastern University, and Dr. Joseph Snider from University of California San Diego. The study was supported in part by the U.S. National Science Foundation grant CAREER IIS-2045523.

The study also identifies research gaps in eye-tracking, such as applications of a particular measure in a domain. In our study, we discuss possible research gaps and the reasons. A researcher can use the information from our study to identify and address the research gap.

Depending on the requirement, our review is helpful from multiple perspectives. Firstly, a researcher exploring eye tracking in studies can use our study as a reference guide for utilizing eye tracking in studies. Since we categorize and attribute measures with internal ocular mechanisms, the reader can quickly identify domains that use a particular measure. Moreover, our study aggregates knowledge from a broader range of domains. By referring to our study, one can gather implications from other domains, such as the psychological implications of human-computer interaction experiments.

Please find our publication (open access) at https://doi.org/10.3389/fcomp.2021.733531 and cite our paper, 

Mahanama Bhanuka, Jayawardana Yasith, Rengarajan Sundararaman, Jayawardena Gavindya, Chukoskie Leanne, Snider Joseph, Jayarathna Sampath, "Eye Movement and Pupil Measures: A Review", Frontiers in Computer Science, vol. 3, 2022.

--Bhanuka (@mahanama94)