Review Article

Computational Techniques for Eye Movements Analysis towards Supporting Early Diagnosis of Alzheimer’s Disease: A Review

Table 1

Research on eye movements and Alzheimer disease on pubmed since 2013.

Cite Methods Findings Participants/Apparatus

2016 [25] Subjects responded to targets presented on a hemispherical screen with diverse eccentricity. PwAD recognized less targets in the center. No difference was found with CG on the peripheral targets.AD: 18 CG: 20 Apparatus: Hemispherical screen Octupus 900 with camera used for eye tracking.

2017 [26] The King-Devick test (with saccadic and other movements) was applied to subjects. The King-Devick test may a tool to detect cognitive impairment associated with AD.AD: 32 CG: 135 MCI: 39 Apparatus: N/A

2016 [27] Subjects looked a series of slides containing four images of different emotional themes. PwAD with apathy had diminished attentional bias toward social-themed stimuli.AD: 36 (Apathy: 17 Not apathy: 19) Apparatus: Binocular eye tracking system developed by EL-MAR Inc.

2016 [11] Eye movements from subjects were examined during reading regular and high predictable sentences. PwAD gaze was longer than CG gaze. CG decreased gaze duration with high predictable sentences suggesting reading enhancement using stored information.AD: 35 CG: 35 Apparatus: EyeLink 1000. Chinrest to control eye movements.

2015 [28] Subjects performed a variety of tasks: walking, through stairs, through a room with and without obstacles. The Posterior Cortical Atrophy (PCA) patient had longer mean fixation durations than PwAD and CG. Mean fixation duration between PwAD and CG was similar.AD: 1 CG:1 PCA: 1 Apparatus: SMI mobile eye tracker

2015 [29] Eye movements from subjects were examined while read sentences. PwAD had more fixations on regular and high predictable sentences. PwAD spend more time reading the sentence. CG had less frequent second pass fixation over sentences.AD: 35 CG-elderly: 35 Apparatus: EyeLink 1000. Chinrest to control eye movements.

2015 [19]Longitudinal study with Gap and overlap paradigms.PwAD had slower reaction times than CG. Prosaccades did not deteriorate after the 12-month longitudinal study in AD.AD: 11 CG elderly: 25 Apparatus: ExpressEye

2015 [30] Subjects made saccadic movement to photographs to target instructed scenes (natural vs urban, indoor vs outdoor) Were found differences between controls and PwAD on accuracy but not saccadic latency.AD: 24 CG age-matched: 28 CG young: 26 Apparatus: Eye tracker (Red-M, Senso-Motoric Instruments)

2015 [23] Eye movements from subjects were examined while read proverbs. PwAD have less word predictability than CG.AD: 20 CG: 40 Apparatus: EyeLink 1000. Chinrest to control eye movements.

2014 [31] Eye movements from subjects were examined while read low and high predictable sentences. CG have shorter gaze duration on high predictable sentences. PwAD have similar gaze duration on both low and high predictable sentences. PwAD gaze duration is longer than CG.AD: 20 CG age-matched: 40 Apparatus: EyeLink 1000. Chinrest to control eye movements.

2014 [32] Eye movements from subjects were examined while read sentences PwAD have altered visual exploration and absence on contextual predictability.AD: 18 HC age-matched: 40 Apparatus: EyeLink 2K. Chinrest to control eye movements.

2013 [33] Eye movements from subjects were examined while read sentences PwAD evidences marked alterations in eye movement behavior during reading.AD: 20 CG age-matched: 25 Apparatus: EyeLink 1000. Chinrest to control eye movements.

2014 [12] Subjects were asked to spot an animal target contained in Colored photographs along with other distracting items.PwAD were significate less accurate than elderly controls. Elder were less accurate than young controls.AD: 17 mild AD. CG elderly: 23 CG young: 24. Apparatus: Eye tracker (Senso-Motoric Instruments)

2014 [34] Subjects were required to look to a small fixation cross for 20 seconds on the center of a screen. CG and PwAD showed significantly differences of microsaccade direction.AD: 18 MCI: 15 CG age-matched: 21 Apparatus: Eye See Cam

2013 [35] Visual targets were presented to subjects in a dim room. Prosaccade and antisaccade trials. The antisaccade taks performance serves as a measure of executive function on PwAD.AD: 28 MCI: 36 CG elderly: 118 Apparatus: Dual Purkinje Image Tracker. Heads stabilized on a chinrest.

2013 [36] Pro-saccade and anti-saccade tasks. Gap and overlap paradigms. PwAD have an excessive proportion of uncorrected errors in the antisaccade test.AD: 18 Parkinson disease: 25 CG-young: 17 CG elderly: 18. Apparatus: Head mounted device ExpressEye eyetracker.

2013 [37] Horizontal and vertical saccades. Gap and overlap paradigms on a black computer screen. A link between MMSE and saccade latency.AD: 25 Amnestic MCI: 18 CG elderly: 30 Apparatus: Head mounted Eyeseecam

CG: Control Group; MCI: Mild Cognitive Impairment; MMSE: Mini Mental State Examination.