AI Model Helps Clinicians Tell Epilepsy Apart from Functional Seizures Using Routine Clinical Notes
Key Takeaways:
- Researchers developed “EpiScreen”, an AI system built on fine-tuned large language models, which distinguished epilepsy from psychogenic non-epileptic seizures (PNES) using only routine clinical notes, achieving an AUC of up to 0.992 within one institution and 0.980 across institutions.
- When clinicians were given EpiScreen’s predictions and sentence-level explanations, neurology residents’ diagnostic accuracy rose from 76.3%–79.3% to 87.7%–89.0%, and senior attending physicians’ accuracy rose from 81.7% to 89.3%.
- The authors propose EpiScreen as a decision-support tool used after neurology referral and before confirmatory video-electroencephalography (vEEG), though prospective real-world trials are still needed.
An AI approach to a long-standing diagnostic challenge
In a recent study published in the journal npj Digital Medicine, researchers describe the development and validation of “EpiScreen”, a novel artificial intelligence (AI) system designed to distinguish epilepsy from psychogenic non-epileptic seizures (PNES) using clinical notes from routinely documented electronic health records (EHRs).
The retrospective, two-institution study drew on 13,633 patient records. EpiScreen outperformed comparison models, achieving an area under the receiver operating characteristic curve (AUC) of up to 0.992 in within-institution testing and 0.980 in cross-institutional testing. AUC measures a model’s ability to distinguish between the two conditions across different classification thresholds. In a retrospective case-review assessment, EpiScreen also improved clinicians’ classification accuracy.
Why telling the two conditions apart matters
Epilepsy is a chronic neurological disorder characterised by recurring, unprovoked seizures caused by sudden, abnormal electrical discharges in the brain. Approximately 50 million people live with the condition worldwide.
Diagnosis can be complicated by PNES, also termed functional or dissociative seizures. These are paroxysmal events (sudden, temporary seizure-like episodes) that resemble epileptic seizures but are not caused by epileptic electrical activity. Because the two conditions have different underlying mechanisms and management needs, accurate differentiation is critical. Misdiagnosis remains common: prior studies indicate that 20%–30% of people referred to tertiary centres for presumed drug-resistant epilepsy go on to receive a PNES diagnosis.
Prolonged inpatient video-electroencephalography (vEEG), which records synchronised brain waves alongside behavioural data, is the diagnostic gold standard. However, it often costs over US$15,000 per admission in the United States, with reported waiting times of 8 to 12 weeks that vary across healthcare systems. Diagnostic delays can also lead to unnecessary treatment and postpone appropriate care.
Routine clinical notes contain extensive diagnostic signals, but seizure semiology (the symptoms, signs and sequence of a seizure-like event) overlaps considerably across disorders, and earlier narrative-based approaches have shown limited diagnostic performance.
How EpiScreen was developed
EpiScreen was trained and evaluated using 3,451 records from the Beth Israel Deaconess Medical Center’s Medical Information Mart for Intensive Care IV (MIMIC-IV) database and 10,182 records from the University of Minnesota (UMN) Clinical Data Repository. Labels reflected clinicians’ final documented diagnoses, and notes were processed to approximate the information available before diagnosis. People with coexisting epilepsy and PNES were excluded.
The system was built by instruction fine-tuning open-source LLMs, including Qwen-2.5-32B, DeepSeek-R1-32B, Llama-3.3-70B and Med42-70B, using Quantized Low-Rank Adaptation (QLoRA), a method that reduces the memory and training demands of adapting large models. To explain its reasoning, the researchers used sentence-level Integrated Gradients (IG), a technique that highlights the sentences most influential in the model’s decisions across 11 predefined clinical feature categories.
Four medical teams, made up of six neurology residents and two senior attending physicians, then evaluated 300 MIMIC-IV test cases, with each pair reaching decisions by consensus.
Study findings
EpiScreen consistently outperformed conventional supervised baselines, including ClinicalBERT. Its Med42-based variant achieved AUCs of 0.891 on MIMIC-IV and 0.976 on UMN, while the Qwen- and DeepSeek-based variants reached 0.992 on UMN. When tested on records from institutions that had not supplied their training data, the best AUCs reached 0.875 on MIMIC-IV and 0.980 on UMN.
Working alone, EpiScreen achieved 81.0% diagnostic accuracy, compared with 81.7% for unaided senior attending physicians and 76.3%–79.3% for unaided neurology residents. These differences were not statistically significant. With access to the AI’s predictions and sentence attributions, accuracy rose to 87.7%–89.0% across the resident groups and 89.3% for the attending physicians, with the largest improvement around 12.7 percentage points.
AI assistance also reduced false negatives (missed epilepsy cases) for resident Group A from 36 to 14, while the attending pair reduced false positives from 37 to 19.
Not all diagnostic labels were vEEG-confirmed, although sensitivity analyses in confirmed cases supported the findings. Documentation bias remains possible, and other seizure mimics, including syncope and sleep disorders, were not evaluated.
What this means for clinical practice
The authors suggest EpiScreen could be used during or after neurology referral, before confirmatory vEEG, to support earlier assessment and help clinicians prioritise urgent monitoring. They propose it could reduce diagnostic delays, costs and unnecessary exposure to antiseizure medication, though these remain potential benefits rather than measured outcomes.
The study highlights how AI tools that explain their reasoning may be most valuable when working alongside clinical judgement. Clinicians looking to build confidence in evaluating and using AI in practice may find The College of Contemporary Health’s AI Essentials for Primary Care short course a useful starting point.
Future research must prioritise prospective, real-time trials to confirm how EpiScreen performs in real-world settings using minimally curated notes. The tool is intended to support clinical judgement alongside definitive testing, not replace it.
CCH insight
As AI decision-support tools like EpiScreen move closer to everyday practice, healthcare professionals need the knowledge to assess what these systems can and cannot do. Our AI Essentials for Primary Care CPD course helps clinicians understand the role of AI in healthcare and apply it safely and confidently in practice.
Explore AI Essentials for Primary Care →
Source: npj Digital Medicine
Read More
New Study Suggests Semaglutide May Reduce Epilepsy Risk in Adults with Type 2 Diabetes
Key Takeaways:
- Initiating semaglutide was associated with a significantly lower risk of developing epilepsy or seizures compared with other glucose-lowering therapies.
- The observed effect does not appear to be primarily driven by improvements in blood glucose or body weight.
- Findings are preliminary and should be interpreted as an early signal rather than a basis for changing clinical practice.
Background – seizure risk in type 2 diabetes
People living with type 2 diabetes mellitus face a higher risk of developing seizures and epilepsy. This increased risk is thought to be partly driven by inflammatory processes that affect the central nervous system. While glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are primarily used for glucose regulation, there has been growing scientific interest in their potential neurological effects.
However, until now, the relationship between GLP-1 RAs and seizure risk has remained unclear.
Study design and population
The findings were presented at the American Academy of Neurology Annual Meeting, held in Chicago from April 18 to 22, 2026.
Researchers conducted a retrospective study using a target trial emulation approach, drawing on data from the National Institutes of Health All of Us Controlled Tier Dataset. The analysis focused on adults aged 18 years and older with type 2 diabetes mellitus who initiated one of the following treatments between December 2018 and December 2021:
- Semaglutide
- Sodium-glucose cotransporter-2 inhibitors (SGLT2 inhibitors)
- Other glucose-lowering therapies
Participants were followed through to December 2023.
To ensure comparability between groups, inverse probability of treatment weighting was applied to balance baseline characteristics. Time-to-event analyses were then used to assess the risk of developing epilepsy or seizures.
Cohort characteristics
A total of 393,596 individuals met eligibility criteria. Within this population:
- 8,533 individuals were included in the semaglutide versus other glucose-lowering drug comparison (2,397 vs 6,136)
- 7,455 individuals were included in the semaglutide versus SGLT2 inhibitor comparison (1,650 vs 3,725)
Key findings – reduced risk of seizures
After statistical adjustment, initiation of semaglutide was associated with a lower risk of epilepsy or seizures compared with both comparator groups:
- Compared with other glucose-lowering drugs:
- Hazard ratio: 0.46
- 95% confidence interval: 0.25–0.83
- P = .010
- Compared with SGLT2 inhibitors:
- Hazard ratio: 0.44
- 95% confidence interval: 0.22–0.86
- P = .017
These findings suggest a meaningful reduction in relative risk among those initiating semaglutide.
Absolute risk reduction and clinical interpretation
Further modelling provided estimates of absolute risk reduction:
- Compared with other glucose-lowering therapies:
- Absolute risk reduction: -0.014
- Number needed to treat: 69
- P < .001
- Compared with SGLT2 inhibitors:
- Absolute risk reduction: -0.008
- Number needed to treat: 129
- P < .001
These results indicate that, while the relative risk reduction is notable, the absolute reduction in risk remains modest.
Mechanisms – not driven by glycaemic control alone
To explore potential mechanisms, the researchers conducted mediation analyses. These analyses assessed how much of the observed effect could be explained by changes in glycated haemoglobin or body mass index.
The results showed that:
- Glycated haemoglobin accounted for only 1.1% of the effect compared with other glucose-lowering drugs and 3.6% compared with SGLT2 inhibitors
- Body mass index contributed 0.3% or less in both comparisons
This suggests that the reduced seizure risk may not be primarily driven by improvements in glycaemic control or weight loss, pointing towards other possible mechanisms.
Expert perspective
Yoonhyuk Jang, MD, PhD, Postdoctoral Fellow in the Department of Immunology at Harvard Medical School, commented on the findings:
“This study suggests that semaglutide may be associated with a lower risk of adult-onset seizures or epilepsy in patients with type 2 diabetes, with the signal appearing more pronounced in late-onset cases among adults aged 60 years or older.”
He added:
“Given that age-associated brain insults are major contributors to adult-onset seizures and epilepsy, these findings may have implications beyond seizure risk alone and raise the possibility that semaglutide could also be relevant to broader brain health; however, because this was a retrospective target trial emulation with a relatively small number of events, it is not yet practice-changing and should instead be viewed as a signal that supports further research into the role of GLP-1 receptor agonists in epileptogenesis.”
Limitations and future directions
The authors emphasised that the study design was observational, despite using advanced statistical methods to emulate a clinical trial. As such, causality cannot be definitively established.
Additionally, the relatively small number of seizure events limits the strength of the conclusions. These findings should therefore be interpreted cautiously and seen as hypothesis-generating.
Further prospective and randomised studies will be needed to determine whether GLP-1 receptor agonists such as semaglutide have a direct role in reducing seizure risk or influencing broader neurological health.
Disclosures
One study author reported affiliations with biotechnology, pharmaceutical, or device companies. Full disclosure details are available in the original study source.
Source: Neurology Advisor
Read More