2024
Diagnostics of dyslexia using eye-tracking and artificial intelligence
ŠVAŘÍČEK, RomanZákladní údaje
Originální název
Diagnostics of dyslexia using eye-tracking and artificial intelligence
Autoři
Vydání
Erasmus+ přednáška na University of Jyväskylä, 2024
Další údaje
Jazyk
angličtina
Typ výsledku
Vyžádané přednášky
Obor
50300 5.3 Education
Stát vydavatele
Česká republika
Utajení
není předmětem státního či obchodního tajemství
Organizační jednotka
Filozofická fakulta
Klíčová slova anglicky
dyslexia; eye-tracking; artificial intelligence
Změněno: 11. 5. 2024 22:23, Mgr. Roman Švaříček, Ph.D.
Anotace
V originále
This study explores an innovative approach to diagnosing dyslexia, integrating eye-tracking technology and artificial intelligence (AI) to enhance diagnostic accuracy and efficiency. Traditional methods of diagnosing dyslexia often rely on subjective assessments and standardized test batteries, which may not fully capture the nuances of individual eye movement patterns. By utilizing eye-tracking data, this research identifies distinct patterns in reading behavior that differentiate dyslexic readers from non-dyslexic readers, such as increased fixation durations and more frequent regressions. AI algorithms, particularly convolutional neural networks, are employed to analyze these patterns, achieving classification accuracies around 90%. This approach not only offers a more objective basis for diagnosis but also provides insights into the specific challenges faced by dyslexic readers, facilitating more personalized intervention strategies. The findings suggest potential for broader application in psychological assessments and educational settings, aiming for early and accurate identification of dyslexia to better support affected individuals.
Návaznosti
| TL05000177, projekt VaV |
|