Jin Hyung Lee | Visualizing Brain Circuits with EEG and AI

Dr. Jin Hyung Lee is an associate professor at Stanford and founder of LVIS, a company translating years of brain circuit research into clinical software for neurological care. In this episode, we discuss how her team thinks about the brain as a circuit system, why neurological disorders can be difficult to diagnose with today’s tools, and how LVIS’s NeuroMatch platform aims to use EEG data to visualize brain network function in a more accessible way.

We also talk about why EEG remains clinically useful despite its limitations, how software can help clean and interpret noisy brain signals, and why Dr. Lee believes brain health should become something we monitor and maintain rather than only address after major disability occurs. The conversation touches on epilepsy, memory problems, traumatic brain injury, anxiety, depression, sleep, clinical access bottlenecks, and the broader future of brain health clinics.

Top 5 takeaways

1. Brain disorders may be better understood as circuit problems.
Dr. Lee frames neurological disease around how signals flow through the brain, rather than only around symptoms or broad diagnostic categories. Her goal is to visualize brain network function so clinicians can better understand where communication may be disrupted.

2. EEG is imperfect, but accessible.
EEG is noisy and spatially limited, but it is far easier to deploy than MRI or invasive recordings. LVIS chose EEG because a 19-channel clinical EEG system could potentially make brain network analysis more accessible to patients and clinicians.

3. Software may help extract more useful information from traditional brain recordings.
Dr. Lee explains that the NeuroMatch approach combines signal processing, machine learning for data cleaning, and years of circuit neuroscience insight to extract meaningful patterns from EEG. The claim is not simply “AI finds everything,” but that software can help make noisy clinical data more interpretable.

4. Access to neurological testing is a major bottleneck.
Although EEG may seem simple from an engineering perspective, clinical EEG requires trained technicians, physicians, interpretation time, and infrastructure. Dr. Lee argues that automation and cloud-based software could reduce delays and make neurological evaluation more available.

5. Dr. Lee’s long-term vision is preventive brain health.
Rather than waiting until someone is severely impaired, she imagines a future where people monitor brain function the way they monitor blood pressure, glucose, weight, or car maintenance. That is the central philosophical shift of the episode: brain health as something measurable, trackable, and potentially maintainable.

0:00 Introduction to Dr. Jin Hyung Lee, Stanford, LVIS, and NeuroMatch
0:50 Translating 15 years of Stanford brain circuit research into clinical tools
1:30 Why brain disorders require understanding how signals flow through the brain
2:35 NeuroMatch, EEG, and visualizing individual brain network status
3:30 Epilepsy, memory problems, TBI, anxiety, depression, sleep, and other applications
5:00 What a brain network scan might reveal about symptoms
7:30 Measuring brain performance like glucose, weight, or other health metrics
8:10 Why use EEG despite its noisy and low-resolution nature?
9:20 Standard 19-channel clinical EEG and extracting more information from limited data
12:25 Why this was not possible before: hardware, signal processing, and circuit knowledge
15:00 Novera Brain Health Institute and a new model for brain health clinics
18:40 Why clinical EEG is less accessible than it seems
21:30 Software as a medical device and keeping physicians in the loop
23:30 Dr. Lee’s origin story: electrical engineering, her grandmother’s stroke, and brain circuits
27:00 The five-year vision: maintaining brain health before disability occurs
29:00 Early clinical use and the hope for changing the future of brain disorders