Attention deficit hyperactivity disorder affects more than 22 million Americans, yet clinicians still lack a widely accepted biological tool to distinguish between its three major subtypes. Firefly Neuroscience, Inc. has announced research findings suggesting that brain wave biomarkers derived from its Evoke System EEG platform could help differentiate between inattentive, hyperactive-impulsive, and combined ADHD presentations.

The discovery, based on analysis of resting electroencephalography and event-related potential signals captured using the FDA-cleared Evoke System, highlights how artificial intelligence-assisted neurophysiology may move ADHD diagnosis closer to objective neurological measurement rather than symptom-only classification.

How AI-driven EEG biomarkers could reshape subtype classification in attention deficit hyperactivity disorder

For decades, ADHD diagnosis has relied primarily on behavioral assessments rather than measurable biological indicators. Clinicians typically apply criteria from the Diagnostic and Statistical Manual of Mental Disorders, which classify the disorder based on patterns of inattentiveness, impulsivity, and hyperactivity reported across multiple settings.

This approach has enabled broad clinical recognition of ADHD but has also created persistent ambiguity around subtype identification. Behavioral symptoms often overlap between categories, and patient reports can vary depending on context, observer interpretation, and developmental stage.

Industry observers note that the lack of an objective biological signal has long limited the precision of ADHD classification. Without biomarker support, clinicians must rely on symptom clusters rather than neurological signatures.

Firefly Neuroscience’s work suggests that distinct electrophysiological patterns may exist across ADHD subtypes. By analyzing resting EEG signals and cognitive event-related potentials, the company believes subtype-specific neural activity patterns can be identified.

If validated in further research, this could introduce a measurable neurological layer into ADHD classification. That shift would represent a notable step toward biomarker-informed psychiatry, a field that has historically struggled to identify reliable physiological indicators for complex neurodevelopmental disorders.

Why objective brain signal markers could change treatment strategy and clinical decision-making

The practical implications of subtype differentiation extend beyond diagnosis. ADHD treatments vary widely in mechanism and response profiles, ranging from stimulant medications to behavioral therapy and neurofeedback interventions. Clinicians tracking the field note that some ADHD presentations respond well to stimulant medications such as methylphenidate or amphetamine-based therapies, while others demonstrate weaker responses or experience tolerability challenges.

Without objective markers, determining which treatment strategy may work best often requires trial-and-error prescribing. This process can delay effective therapy and complicate long-term care planning.

Firefly Neuroscience suggests that subtype-specific EEG biomarkers could provide clinicians with a more precise understanding of underlying neurological patterns. In theory, this could help guide therapy selection by identifying which patients may benefit from stimulant treatment, behavioral interventions, or alternative approaches such as neurofeedback.

The company also highlights the potential role of EEG-based monitoring in tracking treatment response. If neurological signals associated with ADHD symptoms shift in response to therapy, clinicians could potentially observe treatment effects directly in brain activity data rather than relying exclusively on behavioral reports. While such capabilities remain exploratory, they align with a broader movement toward precision medicine in neuropsychiatry.

What this discovery reveals about the emerging role of large-scale brain data in neuropsychiatric research

Another notable element of Firefly Neuroscience’s research is the scale of the dataset supporting its artificial intelligence models. The company reports that its broader platform development effort draws on a repository of more than 191,000 brain scans.

Large neurophysiological datasets have historically been difficult to assemble due to technical variability in EEG acquisition and the fragmented nature of clinical data collection. Advances in data aggregation and computational analysis are beginning to change that dynamic.

Industry analysts note that machine learning approaches increasingly depend on large datasets to identify subtle patterns in brain activity that might otherwise remain undetected. By combining EEG signals with AI pattern recognition tools, researchers aim to identify neural signatures associated with specific cognitive states or disorders.

Firefly Neuroscience’s broader ambition involves building what it describes as an EEG and event-related potential brain foundation model. In concept, such a model would function similarly to large language models in artificial intelligence but would instead map neurological patterns across different cognitive conditions. If successful, such a platform could support diagnostic research across multiple neurological and psychiatric disorders beyond ADHD.

How EEG biomarker research compares with other diagnostic approaches in neurodevelopmental disorders

The search for objective biomarkers in ADHD is not new. Researchers have explored various biological indicators over the past several decades, including neuroimaging signatures, genetic markers, and neurochemical pathways.

Electroencephalography has long been considered a promising candidate because it directly measures neural electrical activity and is relatively inexpensive compared with advanced imaging technologies. However, translating EEG signals into clinically actionable biomarkers has proven difficult. Brain wave patterns vary widely between individuals, and differences associated with ADHD are often subtle rather than definitive.

Some earlier studies suggested that certain EEG patterns, such as altered theta-beta ratios, may correlate with ADHD symptoms. Yet these signals have not achieved universal acceptance as diagnostic standards due to variability across patient populations and research methods.

Firefly Neuroscience’s approach differs by incorporating artificial intelligence analysis of complex signal patterns rather than relying on single EEG ratios or markers. Industry observers believe this multi-signal pattern analysis may improve reliability compared with earlier approaches that focused on isolated EEG features.

Nevertheless, experts caution that biomarker discovery alone does not guarantee clinical adoption. Demonstrating consistent diagnostic accuracy across diverse patient populations remains essential before regulators or clinicians consider widespread implementation.

What clinicians, regulators, and industry observers are likely to watch as this research progresses

Despite the promising nature of EEG biomarker research, several questions remain unresolved. One central issue is whether the observed brain wave patterns can consistently distinguish ADHD subtypes across large and heterogeneous patient groups.

Clinical validation will likely require prospective trials comparing EEG biomarker predictions with established diagnostic methods. Regulators and clinicians will want evidence that biomarker-based classification improves patient outcomes rather than simply replicating existing diagnostic categories.

Another factor involves integration into clinical workflows. Even if EEG biomarkers prove reliable, clinicians must determine how such data should be incorporated alongside behavioral assessments and psychiatric evaluation.

Reimbursement considerations also remain uncertain. Diagnostic tools in psychiatry often face reimbursement challenges unless they demonstrate clear value in guiding treatment decisions or reducing healthcare costs.

Industry observers also note that the broader vision of a brain foundation model will depend heavily on data quality and standardization. EEG data collected across different devices and clinical environments can vary significantly, which may affect the consistency of AI training datasets.

Finally, competition in the neurodiagnostics space continues to intensify. Several technology companies and academic groups are exploring AI-driven neurological biomarkers for conditions ranging from depression and schizophrenia to Alzheimer’s disease.

Firefly Neuroscience’s ADHD biomarker discovery therefore sits within a larger race to apply artificial intelligence to brain signal analysis. While the company’s findings remain at an early research stage, they highlight the growing belief among neuroscientists and technology developers that objective brain-based biomarkers may eventually transform how psychiatric disorders are diagnosed and managed.

  ADHD, AI in neuroscience, attention deficit hyperactivity disorder, EEG biomarkers, event related potentials, Evoke System, Firefly Neuroscience Inc., neurodiagnostics