Feasibility and Preliminary Clinical Validation of the Explainable Digital Biomarker for Dyslexia (EDBD): A Prospective Pilot Study Protocol

For millions of school-age children, the path to learning to read is blocked by a hidden hurdle known as dyslexia. This common learning difference affects roughly one in ten students, yet the system designed to help them often fails to notice the problem until it is too late. By the time a child is formally identified, usually in third grade or later, the critical window for the most effective brain-based interventions has often closed. The science of early detection has advanced significantly over the last forty years, identifying specific signs in how children process sounds, name objects quickly, and remember sequences of words. However, turning these scientific signs into a practical tool that a regular classroom teacher can use has remained a distant goal. The challenge has not been a lack of data, but rather the difficulty of building a system that is both accurate and understandable to the adults who must act on it.

This paper outlines a carefully planned first step toward solving that problem. It does not present a finished product or a final verdict on a new diagnostic tool. Instead, it describes a protocol for a pilot study, a preliminary test designed to see if a new digital screening method can actually work in a real elementary school. The researchers have spent years synthesizing decades of scientific evidence to create a specific framework called the Explainable Digital Biomarker for Dyslexia. This framework is built on four specific markers of reading difficulty that have been proven to predict future struggles. The innovation lies in how the tool presents its findings: rather than giving a simple "yes" or "no" answer, it is designed to show the reasoning behind its conclusion in a way that teachers and doctors can follow. The goal of this study is not to prove the tool works perfectly right away, but to determine if it can be administered to children, if the data it collects is reliable, and if the explanations it provides actually help adults make better decisions.

The study is set to take place in elementary schools, targeting children in first through third grade, roughly ages six to nine. This age range was chosen because it captures the period when the four specific markers are most predictive of future reading difficulties, yet early enough that intervention can still change a child's educational trajectory. The researchers plan to enroll approximately 200 children in the pilot. Every child who agrees will sit down with a trained staff member and complete a short series of tasks on a tablet. These tasks measure how well the child can break words into sounds, name familiar objects quickly, choose the correct spelling, and remember the order of spoken words. The entire session is designed to take no more than thirty minutes.

While all 200 children will complete this initial screening, only a smaller group of about 60 will undergo a deeper, more traditional evaluation. This smaller group will include every child who shows signs of difficulty on the tablet test, plus a random selection of children who performed well. These 60 children will receive a comprehensive, one-on-one psychoeducational assessment conducted by a licensed school psychologist. This traditional evaluation is the current gold standard for diagnosing dyslexia, involving hours of testing and interviews. Crucially, the psychologist performing this deep dive will not know the results of the initial tablet screening. This blinding ensures that the comparison between the new digital tool and the traditional method is fair and unbiased.

The researchers are not just testing whether the tablet can spot reading difficulties; they are also testing whether the tool's design helps or hinders the people using it. A separate part of the study will ask teachers and clinicians to review simulated cases. They will look at the tool's output in two different ways: once with the full explanation showing the evidence, and once with just a raw probability score. The goal is to see if seeing the reasoning behind the score makes the adults trust the tool more, or if it leads them to blindly follow the computer's suggestion. This is a critical question, as tools that are accurate but confusing can be just as dangerous as tools that are easy to use but wrong.

Because this is a pilot study, the researchers are not trying to declare the tool a success or failure based on a single number. Instead, they have set up a "traffic light" system to decide what happens next. If the study runs smoothly, with high participation rates, short administration times, and no technical glitches, the light turns green, and the team will move forward to a much larger, fully powered trial. If there are minor hiccups, such as the test taking a bit longer than expected, the light turns amber, signaling that the team needs to tweak the process before trying again. If the study fails to recruit enough families or the technology breaks down too often, the light turns red, and the team will have to go back to the drawing board. This cautious approach ensures that no time is wasted on a tool that cannot function in the real world.

The paper emphasizes that no results have been collected yet. This document is a blueprint, a promise of how the researchers intend to gather evidence responsibly. They are acutely aware of the risks involved in testing new technologies on children. To protect the participants, the study requires written permission from parents and a separate, age-appropriate agreement from the children themselves. The researchers have also built in safeguards to ensure that the screening results are never used alone to deny a child services or label them permanently. The data collected will be kept anonymous and secure, and the study is designed to include families from diverse backgrounds to ensure the tool works for everyone, not just a privileged few.

Ultimately, this protocol represents a shift in how artificial intelligence is brought into the classroom. Rather than rushing to deploy a complex algorithm, the researchers are taking a deliberate, step-by-step approach. They are building a bridge between decades of scientific discovery and the daily reality of the school system. By focusing first on feasibility and human understanding, they aim to create a tool that does not just identify a problem, but explains it clearly enough to empower teachers and parents to act. If the pilot study meets its criteria, it will pave the way for a larger trial that could finally close the gap between early warning signs and the help children need to succeed.