Epitel’s Roadmap from Seizure Detection to Brain Health Platform

Spend five years building a dataset to detect seizures, then use it to detect Alzheimer’s years before symptoms appear. That’s not feature creep—it’s platform strategy.

In a recent episode of Category Visionaries, Mark Lehmkuhle, CEO and Founder of Epitel, a brain health technology platform that’s raised over $20 million, revealed an expansion roadmap that transforms a single-use medical device into a comprehensive brain health platform. The lesson for B2B founders isn’t about neurology—it’s about how initial infrastructure investments unlock adjacent markets that would be prohibitively expensive to enter independently.

The Infrastructure That Enables Everything

Epitel’s platform strategy didn’t start as platform strategy—it started as necessity. When Mark needed training data for their seizure detection AI, he discovered that hospitals “just delete the data because we don’t have the storage for it.” This forced Epitel to spend years building their own dataset, “we used the wired EEG as kind of the ground truth and then trained our machine learning on our sensor data.”

That dataset, built to solve one problem, became the foundation for a platform. The same AI trained to identify seizure patterns in EEG data can be retrained to identify patterns associated with other central nervous system disorders. The same wireless sensor hardware that captures seizures can capture brain activity markers for multiple conditions. The same regulatory pathways proven for one indication become templates for others.

This reveals a critical principle for medical device companies: the infrastructure you build to solve your first problem determines which adjacent problems you can tackle economically. Epitel didn’t just build a seizure detector—they built the data collection, AI training, and regulatory competency infrastructure that makes brain health monitoring viable as a category.

From Diagnostic to Monitoring

Epitel’s current product serves two markets: hospitals diagnosing seizures and outpatient monitoring to capture rare seizure events. But Mark’s next move shifts from episodic diagnosis to continuous monitoring for patients already diagnosed with epilepsy.

“What I want to do next with this is for people who have been diagnosed with a seizure disorder,” Mark explains, describing a future device that “really fits behind the ear and looks like a hearing aid. It’s rechargeable, reusable, you can use it whenever you want, and it would count the number of seizures that you’re having.”

This shift from diagnosis to monitoring changes the business model fundamentally. Diagnostic tests are one-time events with reorder cycles measured in months or years. Continuous monitoring creates recurring revenue with devices patients wear daily. The economic model looks more like continuous glucose monitors than traditional medical tests.

The monitoring use case also enables objective treatment optimization. “How do we know that this anti seizure medication is working?” Mark asks. Currently, epilepsy patients rely on self-reporting, but “a lot of seizures can occur at night while you’re sleeping. You wouldn’t even know that you’re having it. So self reporting is terrible.”

Objective seizure counting enables physicians to balance efficacy with side effects—”could we balance seizure control with side effects for a drug that’s working”—and even adjust medication timing to match each patient’s seizure patterns, since “there’s a chronicity to most people’s seizure events, whether it’s daily, weekly, monthly, or somewhere in between.”

The Real-Time Alerting Layer

Beyond counting, continuous monitoring enables real-time intervention. Mark describes alerting “a parent or a caregiver or a loved one to an ongoing seizure,” which addresses a specific mortality risk: “People with epilepsy have a higher incidence of what’s called sudden, unexplained death in epilepsy. It’s really kind of a cardiac arrest event where you have a seizure where your face is in the pillow.”

Real-time alerting transforms the product from diagnostic tool to safety device. This creates new stakeholder conversations—not just neurologists but families, caregivers, and potentially insurance companies interested in preventing adverse events. Each new stakeholder potentially opens new revenue streams or reimbursement pathways.

Forecasting as the Next Layer

Mark’s vision extends to predictive capabilities: “What’s my probability of having a seizure in the next hour? I want to go to the grocery store. I have a high probability of having a seizure. Maybe I’ll go tomorrow.”

Forecasting requires understanding each patient’s seizure chronicity through long-term data collection. “With long term monitoring, we can then really determine what do those cycles look like for you?” This isn’t just about convenience—it’s about quality of life and independence for patients whose seizure unpredictability limits their activities.

The technical foundation enabling forecasting—long-term EEG data collection, pattern recognition AI, and personalized modeling—was built to solve the initial diagnostic problem. Each new capability layers onto existing infrastructure rather than requiring separate development efforts.

The Preventive Brain Health Vision

Mark’s most ambitious expansion moves from treating known conditions to preventing future ones. His vision: “You go into your primary care physician’s office for your annual physical, and you’re recording eeg with our system for ten minutes while you do a task. And that becomes kind of like the baseline for your brain health year after year.”

This positions Epitel alongside continuous glucose monitors—moving from tools for diagnosed patients to preventive monitoring for everyone. “Could we detect the symptoms of Alzheimer’s through eeg years before you have any physical symptoms? And in that way, we could have an intervention early on that either slows the progression or even prevents the disease altogether.”

The market size of this vision dwarfs epilepsy alone. Annual physical EEG baselines could become as routine as blood pressure checks, with Epitel’s AI tracking changes that indicate developing conditions. The same dataset built to detect seizures becomes the training foundation for detecting early Alzheimer’s, Parkinson’s, or other neurodegenerative diseases.

Adjacent Acute Applications

Beyond chronic monitoring, Mark sees applications in acute care settings. “Could we identify people who are suspected of having a stroke before it becomes an event that we’re all familiar with?” he asks. “Could we differentiate those who may have a large vessel occlusion versus aneurysm versus a small vessel occlusion right there in the ambulance?”

This addresses a critical triage problem: “If it’s a large vessel occlusion, you know, only a specialized stroke center can handle that. And if we take them to your neighborhood hospital, all they’re going to have to do is transfer you again. And time is brain, and, you know, getting them the right care as quick as possible is super important.”

Each new application leverages the same core technology—wireless EEG sensors and pattern recognition AI—but opens entirely new customer segments and use cases. The infrastructure investment that took five years to build for seizure detection becomes reusable across multiple indications.

The Platform Economics

Mark acknowledges the scope of this vision: “I’ve got this, like, roadmap that is huge of all the different spaces that we want to go with this technology. And, like a golden retriever, where I get, you know, distracted by everything.”

But this isn’t distraction—it’s deliberate platform strategy. The marginal cost of adding new indications is dramatically lower than the original infrastructure build. The first indication required building hardware, collecting datasets, training AI, proving regulatory pathways, and establishing clinical relationships. Each subsequent indication reuses most of that infrastructure.

This economics enables Epitel to expand into adjacent markets that would be impossible for new entrants to tackle economically. A startup trying to build Alzheimer’s early detection from scratch would face the same multi-year data collection and AI training process Epitel already completed. Epitel’s platform moat isn’t just the technology—it’s the infrastructure that makes new applications economically viable.

The Takeaway for Platform Builders

Epitel’s expansion strategy reveals how single-use medical devices become platforms. The principle extends beyond healthcare to any market where initial infrastructure investments are high but enable adjacent applications.

The framework: identify what expensive infrastructure you’re building to solve your first problem, then map which adjacent problems share that infrastructure. The dataset that took five years to build for one use case becomes your platform advantage. The regulatory pathways proven for one indication become templates for others. The customer relationships built for one application create channels for others.

Mark’s advice to founders pursuing grants applies equally to platform strategy: “To stick with it. If you truly think you have something innovative, all it takes is time.” Build infrastructure once, reuse it everywhere. That’s how a seizure detector becomes a brain health platform.