فاعلية تقنية تتبع العين في اكتشاف المصابين باضطراب طيف التوحد : مراجعة منهجية
Artificial intelligence (AI) techniques play a prominent and crucial role in the rehabilitation and diagnosis of children with autism spectrum disorder (ASD). This study aimed to conduct a systematic review of studies published in scientific journals from 2019 to 2024 that discussed the effectiveness of eye-tracking technology in detecting children with ASD aged 2 to 12 years. A systematic review of MEDLINE, Embase, and Scopus databases was conducted in April 2024. The results demonstrated the effectiveness of eye-tracking technology in detecting children with ASD, with most studies (n=24) concluding its efficacy in early detection. The sample sizes ranged from 14 to 141 children, with 95.8% of the samples comprising children with ASD from foreign environments, while only one study was conducted in an Arabic setting. 100% of the studies (n=24) proved the effectiveness of eye-tracking technology in detecting children with ASD, albeit with varying degrees ranging from 85% to 96%. Additionally, 4.16% of the studies (n=1) were applied to individuals under 6 years old, 54.16% (n=13) to samples aged 6-12 years, and 41.66% (n=10) to individuals under 12 years old. The methods of stimulating the child's eye varied, including video clips, web pages, images, social situations, or animated cartoons. Furthermore, 87.5% of the studies employed experimental or quasi-experimental methodologies, while 12.5% (n=3) utilized a mixed-methods approach. The findings of this review demonstrated the possibility of accurately identifying individuals with ASD, particularly in the case of pre-school children, using eye-tracking technology. However, these results cannot be relied upon with certainty due to the discrepancies among studies in the application of eye-tracking technology and the small sample sizes used in most studies, leading to unreliable results. Therefore, well-designed and executed studies with transparent and comprehensive reports are necessary to determine the optimal model for using eye-tracking technology in detecting children with ASD. (Published abstract)