What are you looking at? Or rather, how do you look? Discover the world of eye tracking and understand what lies behind glances and blinks.

Imagine you want to measure an employee at work. Of course, we want to disturb our participant as little as possible. Otherwise, our measurements may be biased or even invalid. We always try to minimize interfering measurements, but the brain cannot be completely tricked: It typically knows when it is being measured. In order to measure people, to monitor the brain without “directly” questioning the brain and thereby disturbing it, we have to find alternative ways.

There are less intrusive methods to examine the mental state. Perhaps you have heard of magnetic resonance imaging (MRI) or functional MRI (fMRI; Smith, 2004)? This makes it possible to observe the brain during work on the basis of changing blood oxygen levels (blood-oxygen-dependent, BOLD signal). The disadvantage, however, is that this method is very expensive and complicated. The MRI is a static device that cannot be accommodated in a cockpit. A more flexible approach is the electroencephalogram (EEG; Casson, 2019). With the EEG, we can observe brain waves that are related to concentration or sleep. The disadvantage, however, is that a full EEG cap is required, which takes a long time to set up. In addition, analyzing the data is not as easy as it may sound.

So why not monitor, or rather track, the eyes? There is actually a simple idea behind this: the eyes are where the mind is (Just & Carpenter, 1980). The eye-mind hypothesis states that everything the eyes focus on is processed by the mind. An important categorization must be made here: Eye tracking vs. gaze tracking. In eye tracking, we measure the movement of the eyes or associated features. If we combine this data with information from outside, then gaze tracking is performed. With gaze tracking, we can associate eye movements with world coordinates. When we combine gaze tracking with an image of the world the subject sees, we get a better understanding of what the brain is processing. But that is not all. When tracking the eyes, parameters such as blink rate or eye closure percentage can help determine sleepiness or draw conclusions about other mental states.

Faster than a rabbit

Have you ever thought about which part of the body can perform the fastest movements? Without a doubt, the eye is at the top of the list. Blinking and saccades, i.e. parallel and rapid movements of both eyeballs between two fixations, are considered to be the fastest movements that mammals can perform (Fuchs, 1967). Saccades are seen as an involuntary scanning pattern of the eyes to build up a dimensional image of the scene. Fixations, as previously mentioned, occur in between saccades, when the gaze maintains a location.

Depending on the scope and given situation, different approaches to measuring eye movements are possible. Overall, there are several ways to measure the eyes or associated functions. The systems can be differentiated according to their mobility. In head-mounted eye-tracking systems, a special eyeglass-based camera system is worn by the operator. Typically, these systems consist of an infrared light source that illuminates the cornea, a camera system that captures the corneal reflection to estimate gaze, and an external camera. This brings a high degree of mobility, but as these devices are quite heavy and may be unfamiliar for the operator to wear, they can be perceived as quite intrusive.

Other systems, in contrast, are more static and contactless. So-called remote eye tracking systems consist of a series of several camera systems that are aimed at the eyes of an operator sitting in front of a screen. These systems are usually based on the same cornea reflex method, i.e. an infrared light source is also used. The gaze pattern can then be linked to the content of a screen. It is clear that these systems can be rather static and the virtual area in which the head can be located is limited. It should be noted that recent developments allow for simple eye-tracking applications using affordable webcam-based systems. However, the scientific quality might be limited. A third way to measure eye-based features is through electrode-based systems. In these systems, electrodes are placed around the eyes and the electric potential field can be recorded. One advantage is that these measurements can be carried out in complete darkness and with eyes closed.

Fit for purpose?

Depending on the research question, certain systems may be superior to others. No eye or gaze tracking systems are currently used in normal aviation operations. They are mainly used in research projects, but their use in pilot and air traffic controller training is also being investigated. When tracking and recording gaze during simulated critical situations, the gaze pattern could be analyzed in a debriefing. If the student has perhaps failed to recognize a critical cue, eye tracking can be used to uncover any gaps in attention. In research projects, eye tracking measurements can be used to analyze gaze patterns in order to improve the human-machine interface. A different gaze pattern could also provide information about faster or slower information processing when assessing new designed workplaces. Measuring blink rate, blink velocity or percentage eye closure (PERCLOS) appears to be a good way of detecting sleepiness. Potentially, this could be used in research studies, but also in normal operations. If data security and privacy are ensured, systematic detection of sleepiness could help to improve shift systems and rosters.

The potential of eye and gaze tracking is certainly there, but it requires expertise in conducting studies and analyzing and interpreting data. This makes eye tracking not the easiest and most accessible method. But the insights that can be drawn from it could be worth it.

References 

Casson, A. J. (2019). Wearable EEG and beyond. Biomedical engineering letters, 9(1), 53-71.

Fuchs, A. F. (1967). Saccadic and smooth pursuit eye movements in the monkey. The Journal of physiology, 191(3), 609.

Just, M. A., & Carpenter, P. A. (1980). A theory of reading: from eye fixations to comprehension. Psychological review, 87(4), 329.

Smith, S. M. (2004). Overview of fMRI analysis. The British Journal of Radiology, 77(suppl_2), S167-S175.