Dyslexia is the most common of all neurodevelopmental disorders, affecting around 20 percent of the population. Without the proper support, students with dyslexia can struggle to develop foundational reading skills, which can have lasting effects on their academic performance and contribute to challenges with self-confidence over time.
Screening students who are at risk for dyslexia is a critical step in bringing these learners the appropriate classroom support, and AI-powered screening tools offer a cost-effective and time-saving solution to traditional methods. With improved accuracy and greater availability in recent years, AI tools for dyslexia detection are now poised to support millions of students by enabling earlier and more appropriate interventions.
Artificial Intelligence In Dyslexia Detection
While over 40 states have laws that require dyslexia screening in early grades, traditional screening methods typically rely on manual assessments that can be time-consuming and cost-intensive, as they typically need to be administered by specialists.
Screening tools that are faster and more scalable are incredibly valuable to educators. In a recent national survey by The Learning Agency, speech recognition literacy tools were ranked the highest of ed tech interventions that would be most helpful to preK-12 educators. Artificial intelligence in dyslexia detection could be revolutionary for both students and teachers.
Over the last several years, there have been significant advances in the application of AI for literacy screening, particularly in the development of tools that aid in the early detection of dyslexia. These include tools that use machine learning for early dyslexia diagnosis by analyzing real-time speech for risk factors, assessing reading data, or tracking eye movements.
Generally, individuals with dyslexia can face challenges with the following literacy skills:
- Decoding: The skill of sounding out written words by matching letters to the sounds they make.
- Fluency: The ability to read smoothly and easily, at the right pace and with accuracy.
- Encoding: Translating spoken words into writing or spelling.
AI models today can capture key literacy indicators, such as those above, by drawing on speech recordings and gamified interactions with learners. AI-enabled dyslexia screening tools can be easily accessed on standard devices, take only a few minutes of engagement time, and provide prompt results. Early detection of dyslexia using AI is fast, engaging, and accessible, helping educators screen students without the need for intensive training or one-on-one administration.
This article is based on a workshop from the Rethinking Reading: AI for Literacy Achievement workshop series, a set of webinars on education AI applications organized in collaboration with InnovateUS and the Burnes Center for Social Change at Northeastern University.
Interested in Learning More?
- View the full session for more on this topic.
Check out other webinars featured in the Rethinking Reading series.
AI Tools for Dyslexia Screening
Several AI-based tools for dyslexia screening are available on the market today. While not all are wholly AI-based, many feature some combination of AI and non-AI-based components.
Dsytech and ROAR are two example organizations using AI for identifying dyslexia risk factors. Both were winners of the Tools Competition, a Renaissance Philanthropy program operated by The Learning Agency, which seeks to discover and support emerging talent in the field.
- Dystech is an Australia-based organization with an AI-powered screener that uses speech analysis to predict dyslexia risk in under 10 minutes. The screener prompts students to read a combination of real and pseudowords (made-up) words out loud. After capturing the audio recordings, the tool analyzes them against six metrics tied to reading speed and fluency, which are then used to calculate the likelihood of dyslexia immediately.
- ROAR, the Rapid Online Assessment of Reading, is a gamified assessment suite that provides real-time data to educators while screening foundational reading skills across PreK-12. It uses a mix of AI and non-AI components and has been approved by the state of California as an approved dyslexia screening tool. ROAR includes tools ROAR-Word and ROAR-Sentence, both of which can be administered within a few minutes. ROAR-Word is a child-friendly assessment that guides children through identifying real words and pseudowords. ROAR-Sentence is a read-silently assessment prompting children to determine if a sentence is true or false. Both of these assessments connect to an educator dashboard, with different score reports, including classroom, district, and individual reports.
Screening tools that are faster and more scalable are incredibly valuable to educators. In a recent national survey by The Learning Agency, speech recognition literacy tools were ranked the highest of ed tech interventions that would be most helpful to preK-12 educators.
Other organizations using AI to identify and support readers at risk for dyslexia include:
- EarlyBird Education offers a game-based screener designed for early learners, ages 4-8, flagging early literacy challenges and delivering tailored action plans for both educators and parents.
- Amira Learning listens to students reading aloud, uses AI to detect dyslexia risk, and then matches learners with micro-lessons and teachers with instructional support.
- Lexplore applies eye-tracking and AI for rapid reading assessments. While this platform doesn’t explicitly screen for dyslexia, it does help identify students at risk for reading difficulties.
The Future of AI in Dyslexia Screening
Despite having come a long way with artificial intelligence in dyslexia detection, future research is still needed to help make these tools more generalizable to diverse populations.
As many of these screening tools rely on audio input from children, building for more linguistic and speech variability is key to ensuring screening tools have strong speech recognition accuracy. For example, tools need to be able to accurately assess diverse accents and the voices of children from multilingual backgrounds. Without careful consideration and representation of these students in normative samples, screeners, and the schools or organizations implementing them, run the risk of over- or under-identifiying children with dyslexia.
Current datasets are limited in number and not publicly available. More representative benchmark datasets informing tool refinement are needed to ensure the screening tools reach broad populations and support clear connections to effective and personalized interventions based on screening data.
Current datasets are limited in number and not publicly available. More representative benchmark datasets informing tool refinement are needed to ensure the screening tools reach broad populations and support clear connections to effective and personalized interventions based on screening data.
Reshaping Screening for Better Support
Artificial intelligence in dyslexia detection is revolutionizing dyslexia screening, making it more accessible and scalable. With numerous tools on the market that are rapidly deployable, educators are able to more easily identify students at risk of dyslexia and ensure they get the interventions and support needed to thrive in the classroom and beyond.
