The eyes have it: linking eye tracking and cognitive measures across neurodegenerative diseases

A new ONDRI publication¹, the summation of several years of work, establishes the links between oculomotor (eye movement) parameters, a set of robust cognitive measures (such as executive function and memory), and associated differences between neurodegenerative diseases.

This publication was led by Heidi Riek, a Doctoral student in the laboratory of Doug Munoz PhD, at Queen’s University in Kingston, Ontario.

The study in steps

This research accessed ONDRI’s Foundational study (2013-18), which collected extensive data on 520 people living with one of five diseases. The control was a pre-established group of volunteer study participants, described in a previous study²; this group was broadly matched with ONDRI participants for age, with oculomotor measures corrected for sex. ONDRI participants and controls both completed the same eye tracking tests.

Step 1

The researchers established a set of twelve key parameters³, selected to robustly describe eye movement behaviour (e.g., reaction time, amplitude). Some of these parameters are widely accepted in the literature, others were driven by data patterns within the study and the Munoz lab’s expertise in eye tracking.

Step 2

Next, a comparison was made between results of the abovementioned disease subgroups and controls at the individual parameter level.

Step 3

Four overarching factors, or themed groups, were created — using a statistical technique called factor analysis — into which the 12 parameters were naturally grouped based on patterns in the data itself.

Step 4

These four factors were then correlated with the same participants’ scores in five neuropsychology-based cognitive domains (attention/working memory, executive function, language, memory, visuospatial function).

Results

Overall, the data found within each factor were correlated with the domain-specific neuropsychological score, per the hypotheses of the researchers. Put more simply, the grouped eye tracking parameters that were hypothesized to be related to executive function, for example, were shown to have a statistically significant correlation to this cognitive area.

Secondly, when researchers split the variable groups (factors) into individual variables — i.e., they went back to data at the individual level and looked across/between disease groups – participants who were more cognitively impaired had performed worse on eye tracking tasks, across all diseases.

Riek et al believe that each of the four processes identified through the factor analysis exercise happens in sequence in the brain over the course of each eye tracking trial (or individual activity). This represents a significant cognitive endeavour for each 3.5 second trial.

More on the eye tracking trials

The circuitry related to eye movement largely overlaps with disease-impaired circuitry. Therefore, performance on eye tracking tasks can reveal the location and severity of disease processes. At the same time, it can accurately gauge the integrity of both types of circuits, motor and cognitive.

There are three types of eye movement tasks that were part of ONDRI’s Foundational Study protocol:

• Free viewing behaviour – simply watching a screen displaying video clips
• Prosaccade task behaviour – looking at a particular object within your field of vision
• Antisaccade task behaviour – suppressing the urge to look at a particular object on the screen and actively looking away from it

Why is this research important?

There are a number of studies that have preceded this study. However, this study differed in three primary ways in that it provided:

1. insight into the associations between commonly and rarely reported antisaccade-related measurements — potentially illuminating both their shared and distinct circuitry.
2. enhanced understanding of links between pro- and antisaccade behaviour, and specific aspects of cognitive processing, via robust neuropsychology measures encompassing multiple test scores, and
3. a view of a large cohort of multiple neurodegenerative and cerebrovascular diseases and varying levels of cognitive impairment, enabling direct examination and comparison of both.

“We are truly excited to publish this important study and continue our work,” said Riek. “We believe one of the great possibilities of eye movement assessment is that it can be made even more naturalistic (similar to real world, requiring little to no instruction) which is an avenue we are exploring currently,” she continued.