The paper examines how eye-tracking (ET) method make possible to early diagnosis of autism by explaining some specific studies -especially Klin & Warren’s study- about the topic. In contrast to some perceptions that autism takes place at some point later on in human’s life, researcher’s assumption is that the autism disorder has to be traced to the infancy. For infants, they can always follow gazes of people since even before they communicate through language; they try to decipher their environment by looking at it. Infants also learn attitudes as well as body languages to develop meaning, but if they are affected by autism spectrum disorder; the distinction between attraction and attention is a challenge which impairs the process of sense making (Klin & Jones, 2006).
The inner world of infants and individuals should be understood to have meaningful interpretations about the spectrum. Therefore, one important research subject is to explore activities infants engage in, the things that matter to them as well as their perceptions of the world. Once we are able to screen and analyze the spectrum clinically, we have the power to develop effective intervention and prevention methods which include cues about attention, attraction, and attitude in the case of autism. Fortunately, there are already successful technologies used by researchers to study perspectives, perception, and experiences of autistic individuals. One important example is the eye-tracking technology, which allows one to see what newborns, children, and people are engaging within every moment. This can be an effective tool to grasp attention attraction and attitude of ASD people while providing meaning at the same time (Falck-Ytter, Bölte, & Gredeback, 2013).
This paper carries out a critical assessment of the eye tracking method for autism spectrum diagnosis, especially in infants. First, the strengths and weaknesses of tracking the eyes will be discussed. Then, the eye tracking method and autism spectrum disorder relation will be reviewed based on their validity and reliability. Finally, outlook will be provided for future directions.
Autism Spectrum Disorder and diagnosis
Autism spectrum disorder is termed as spectrum and complex because there are some individuals with the disorder that have notable intellectual challenges while some possess special abilities. Those people with the disorder are unique regarding strengths and differences (Klin, Shultz, & Jones, 2015). Autism spectrum disorder-commonly known as Autism- is a complex condition which is characterised by a group of neurodevelopment disorders that causes challenges in social skills, non-verbal communications speech, and interaction as well as having repetitive behaviours. The latter is caused by majorly by a different combination and interaction of both genetic and environmental conditions (Durkin, Maenner, Christensen, Daniels, & Fitzgerald, 2017). However, in regard to environmental effect, researchers claim that the disorder is strongly associated with genetic components of developmental disorders which usually begin before the baby is even born (Thushara Vijayakumar, 2016). Moreover, people with the disorder show a deficit in nonverbal communication that is necessary for social interaction. They also demonstrate a deficit in developing and understanding relationships, the presence of repetitive behaviour patterns as well as a deficit in social and emotional reciprocity (Semrud-Clikeman, Jenifer, Wilkinson, & Minne, 2010). However Klin et al., (2006) hypothesized that people with autism also have social interaction with an environment; the difference is that they attend to physical stimulus rather than human movement (see Appendix A) (Constantino, Kennon-McGill, Marrus, & Klin, 2017).
Before it was coined as ASD, the disorder had various distinctive subtypes which included the Childhood Disintegrative Disorder, Asperger’s Syndrome, Autistic Disorder as well as Persuasive Developmental Disorder not otherwise specified. At present autism diagnosis usually requires that behavioural characteristics are visible as a symptom before the age of about two or three years. In most cases, autism spectrum’s most obvious signs start to appear when the child is at the age of about two years. However, in some cases, it is possible to diagnose ASD as early as six months of age. Since autism spectrum disorder has various causes, there is a need to focus on the first two years of development (Klin, Shultz, & Jones, 2015).
Neuman et al., (1996) states that experts may have different assessments with respect to eye tracking as either an inadequate measure of brain function or as an improved measure of cognition, yet it complements brain and behavioral measures at the same time. Due to this complementation, we are able to do research on human behaviors by objectively measuring and quantifying eye movements, eye tracking is exactly such method that allows us to study human movement and thus, the method is useful to examine the autism spectrum behavior patterns. One important reason for this is that the eye tracking method includes ocular measures which provide further knowledge about accuracy and response times. Due to its high temporal resolution, it makes possible to measure moment by moment how people respond to certain tasks (Neumann & Sanders, 1996). Moreover, Eye tracking is a technology that is used in measuring a person’s point of gaze (the point where an individual is looking at) or the movement of the eye. In essence, it refers to the recording of eye movement and its relative position in an environment (Aslin, 2012). The technique is faster than any other input media, and therefore also useful for conducting usability studies as well as understanding how users interact with their environment. In relation to its speed, sampling rates of eye tracking range indicate that the faster eye trackers achieve sub-millisecond temporal resolution, just as happen in EEG method (Neumann & Sanders, 1996). Although the eye-tracking measure is an indirect measure of brain function, eye-tracking has many positive sides when we compared to EEG and fMRI. For example, participants feel comfortable during the study because they are provided suitable seating, and testing takes place in the natural environment rather than the noisy environment in contrast to the fMRI scanner. Later, the eye tracker equipment is adaptable and portable, thus, can be easily placed in a hospital or school settings. Accordingly, eye tracker provides a large amount of population for researchers. Lastly, contemporary eye trackers have fast calibration procedures thus it allows us to begin an experiment quickly. This is helpful for researchers who seek to reduce and minimize testing time (Ecksteina, Carrilloa, Singleya, & Bunge, 2017).Finally, the technique is enabling to track reflections from the corneal and assess the visual attention of an individual (Luna, Katerina, & Charles, 2008). This is one of the main reasons why Klin and colleagues used eye-tracking as a tool in their study on the interaction of children with autism, with their world and environment. Their study found out that, children with autism had attention toward sound synchronized to motion rather than social one (human movement).
Moreover, contemporary eye trackers dominantly utilise the technology of visible light, which are often called infrared-free (IR-free) eye trackers. The IR illumination is generally used in hands-free interfaces as Kunkar and Kostek point out: ” In order to estimate the fixation (see Appendix B) point the eye is illuminated by infrared diode light which is invisible to the user and does not disturb his/her interaction with the computer. IR sources, appropriately installed on the camera, to generate unique reflections on the user’s eye” (Kunka & Kostek, 2009). To be more precise, the camera with relatively high-resolution has capabilities of tracking gaze positions and directions. The technology works in a such a way that the IR light is focused on the pupil of an eye resulting to reflections in the cornea which are eventually tracked by a high-resolution camera. Thus, eye tracking method can reveal the visual elements that likely attract quick attention as well as those elements that maybe be ignored and overlooked (Bryn Farnsworth, 2016).
Eye Movement and Autism
Autism spectrum disorders (ASD), are the neurological conditions that are characterised by impairments arising from deficits in social interaction, nonverbal communication as well as a deficit in imagination. ASD affects the functioning of the brain which has also been linked with a deficit in cognitive processes such as perceptions as well as atypical attention (APA, 2016). In relation to that, eye movement technology has been applied in exploring the perceptual and the cognitive processes which are associated with autism, and tracking the eyes provides an online measure of cognitive processing as they occur at various tasks (Rayner, 2009). Benson et al., (2011) argues that the patterns of fixation provide information on the existence of saccadic systems (see Appendix B) of orientation. This provides inferences on the significance of holding, as well as capturing attention in autism. The technique of eye movement explores the major aspects of autism from past to present. The first aspect is, the basic characteristics of oculomotor were examined using the low-level eye movement control. Another aspect involves exploring eye movement in the perceptual processing of more complex stimuli. The other aspects involve the exploring of eye movement technique in providing insights of the social impairments which are the most common characteristic of autism. These examinations of aspects are important due to atypical eye movements may indicate of main processing of problems per se, irrespective of social function, also there is some evidence to offer that this may be the point of autism (Benson & Fletcher-Watson, 2011).
In terms of basic oculomotor control, the ET involves the measurement of saccade amplitude, as well as the velocity. For instance; as an initial experiment Rosenhall and colleagues investigated oculomotor control in autism, this study includes 11 children with infantile autism or autistic-like conditions, the majority of the participants with ASD exhibited velocities reduced at maximum as well as saccadic positions of landing which fell short in locating the target (Rosenhall, Johansson, & Gillberg, 1988).
On the other hand, the measurement of the frequency in eye movement revealed that individuals with ASD disorder showed more eye movements. Also, the language development is regarded as a modulating factor of autism, in oculomotor controls (Benson & Fletcher-Watson, 2011). With regard to the perceptual processing of complex stimuli in autism, the eye movement method has identified a relationship between the movement patterns with the imagination or flexibility elements of autistic impairment triad when we compared to previous methods such as, Visual search task and EFT, they had limited work on eye movement control in complex stimuli tasks (Manjaly, Bruning, Neufang, Stephan, & Brieber, 2007). In the case of social impairments, ET shows that people with ASD process face typically, for instance, participants ‘exhibits reduced fixation on major features of the face especially the eyes (Pelphrey, Sasson, Reznick, & Paul, 2002).
Hence it is concluded that the ET method remains efficient in diagnosing autism disorder by analyzing and measuring different sides of the autism.
Autism Diagnosis by Eye Tracking
The method of ET was successful in detecting autism symptoms in infants at the age of six months. Klin and Jones have developed a measure which was compounded by frames and attention funnels (Klin and Jones, 2006). They first tracked a “typical” two years age child’s eyes (sample size is 35). They then tracked a two-year-old child with an autism disorder by letting them both watch a video.
It was noted that the “typical” child focused on the little scenes of fights in the video while the child with autism focused on the opening and closing of the door in the same scenes. Klin and Jones then developed a growth eye fixation chart or the focus percent of the eye for ages two, three, four, five, six and nine months until the age of twenty-four months. It was observed that typical children focused on peoples’ eyes initial times by the time it remains stable. For babies with autism spectrum, it started high up and mild orientation with people but no traction, then a free fall. They then used the data to look six months’ infants and concluded that the eye fixation for the two groups (the typical child and the child with autism disorder) could easily be separated. Also, that measure with other similar measurements can be useful in identifying autism conditions early in life. When compared to early times, it is thus not necessary to wait until a child is two years old or more for autism disorders to emerge for diagnosis. Klin adds that we need to provide this technology to clinics or hospitals; it needs to be eligible for everyone because science has meaning if only it has utility for society. However, the technique is relatively expensive and difficult to learn as well. Not every person can work with eye trackers due to certain attributes associated with it, such as long eyelashes. It is also relatively difficult to interpret the data in this technique because eye movements are often non-intentional. Thus, careful interpretation is necessary to avoid unwanted user action responses (Navalpakkam & Churchill, 2014).Whereas; the main point of tracking the eyes to reveal subconscious movement of an eye, to providing objective measures rather than subjective ones thus non-intentional movement can be considered as beneficial. In addition, ET allows us to quantify visual attention when we compare to other techniques because as it objectively monitors what, when and where people look at.
Moreover, by ET, Klin and colleagues supported their hypothesis by tracking the above, below or average attention in the scenes which exist in children with autism and they made it possible to obtain objects that are overlooked and ignored.
In terms of diagnosing autism in infants, the study of ET in infants through corneal reflections has been successful. It involved the study of gaze performances in infants as well as estimation of gaze location with a high accuracy of infrared reflection. It’s suggested that the early symptom of autism can be found in brain activity level rather than behavioral level. Thus, in addition to Klin and Warren’s study, the study by Chawarska et al.’s demonstrates that, behavioral signs of ASD are present in early six months of age, which is suitable to the earliest age that brain-based markers have been identified (Chawaska, Macari, Dicinola, & Shic, 2016). The most positive aspect of ET over other methods is that it has the ability to capture the dynamics of gaze behavior, even though the experiment is conducted in complicated environments. In a social interaction, correct timing of gaze is likely to be critical, and a slight delay may mean missing important information, which will reduce the observer’s chances to engage in meaningful interactions with other people (Falck-Ytter, Bölte, & Gredeback, 2013).
We can conclude that the ET technology is useful in diagnosing autism spectrum for individuals, but especially for infants. Although there are strengths about ET and autism predictive measures, some aspect of autism remains to be explored. For instance, children with autism spectrum have poor imitation and linguistic ability, but little research has been so far on imitation capabilities and linguistic ability of children. Further goal needs to evaluate eye tracking in terms of rehabilitation methods and diagnostic strategies to enrich both literature and clinical settings (Kok & Jarodzka, 2016).