A high school student can do publishable neuroscience research without a wet lab by analyzing public brain-signal and brain-imaging data, building computational models of neurons, or running a careful behavioral study. The students who get published pick one narrow question, use a method they can finish in a few months, and write it up for a peer-reviewed conference or journal.
This guide covers the questions that work, the methods a student can carry out from a laptop, where neuroscience papers by students get published, which competitions fit, and the neuroscience papers YRI Fellows have already written.
Why neuroscience is a strong field for student research
Neuroscience sits where biology, medicine, psychology and computing meet. That makes it one of the best fields for a student who wants a paper that reads as serious science but does not have access to a university lab.
Three things work in a student's favor:
- The data is public. Large collections of EEG recordings, MRI scans and eye-tracking data are free to download for research.
- The questions matter to people. Alzheimer's, depression, paralysis and concussion are problems every reader understands, which helps with judges and admissions readers.
- Computation does the heavy lifting. Much of modern neuroscience analysis is signal processing, statistics and machine learning, all of which run on a laptop.
Pure wet-lab neuroscience, such as recording from live neurons or working with animal brains, is usually out of reach before college. You do not need it to publish.
Neuroscience research areas a student can work in
| Area | What you study | What you need |
|---|---|---|
| EEG and brain signals | Electrical activity recorded from the scalp | Public datasets, Python |
| Neuroimaging | MRI and other brain scans | Public datasets, a free GPU notebook |
| Computational neuroscience | Simulated neurons and networks | Python, a simulation library |
| Brain-computer interfaces | Turning brain signals into commands | EEG data, sometimes low-cost hardware |
| Behavioral and cognitive studies | Attention, memory, reaction time, eye movements | Participants, consent, a careful protocol |
| Neurological disease detection | Early signs of Alzheimer's, Parkinson's, depression | Public clinical or signal datasets |
Example research questions
A good student question is narrow enough to answer in 10 to 12 weeks and specific enough that a reviewer can tell whether you answered it. These are the kind of questions that work:
Brain signals:
- Which EEG features best separate imagined left-hand from imagined right-hand movement, and does that hold up across many subjects or only the cleanest ones?
- How much accuracy does a seizure-detection model lose when the EEG signal is noisy?
Neuroimaging:
- Can a lightweight model classify Alzheimer's stages from MRI as well as a much larger one?
- Which brain regions does a scan classifier rely on, and do they match what clinicians expect?
Computational models:
- Does replay during rest change how much a simulated network remembers?
- How does changing one learning rule in a spiking network change what it forgets first?
Behavior and disease:
- Do eye-movement measures such as fixation length and saccade size separate Alzheimer's patients from healthy controls?
- Does a short mindfulness exercise change reaction time on an attention task?
Notice that each question names a measurement, a comparison and a group. That is what turns an interest in "the brain" into a project a mentor can help you finish.
Methods a student can actually carry out
1. Analyze public EEG and physiological data
PhysioNet hosts free physiological datasets, including EEG recordings, and OpenNeuro hosts openly shared neuroimaging and EEG studies. A student can download a dataset, clean the signal, extract features such as power in different frequency bands, and compare classifiers. This is the most common path to a first neuroscience paper because the data is already collected and approved for research use.
2. Build machine learning models on brain scans
MRI classification projects ask whether a model can detect disease, estimate severity, or explain which regions drive its answer. The strongest student papers here do not chase the highest accuracy. They test something specific, such as how a model behaves when data is degraded, or whether a smaller model can match a larger one.
3. Simulate neurons and networks
Computational neuroscience uses mathematical models of neurons to test ideas about learning and memory. You state a prediction in advance, run hundreds of simulations, and check whether the prediction holds. It needs no participants and no lab, only patience and solid math.
4. Run a behavioral study with human participants
Reaction time, memory recall and attention tasks can be run with classmates or volunteers. This path needs the most paperwork. Any study with human participants needs informed consent and review before data collection, and ISEF has detailed human participants rules that apply to projects entering affiliated fairs. Read them before you start, not after.
5. Combine fields
Some of the most original student papers borrow a tool from another field. One YRI Fellow adapted signal processing methods from astronomy to find Alzheimer's patterns in brain data. Cross-field work is easier to defend as new because fewer people have tried it.
Neuroscience papers by YRI Fellows
These are papers YRI Fellows wrote with their PhD mentors, taken from our published work list:
- Suriya D., a 9th grader, built machine learning models on eye-tracking measures to separate Alzheimer's patients from healthy controls, tested under signal degradation. His paper was accepted at IEEE EMBC 2026 in Toronto, the annual conference of the IEEE Engineering in Medicine and Biology Society. Read his story.
- Aditya A. classified motor imagery straight from EEG for brain-computer interfaces, benchmarking six models on the PhysioNet Motor Movement/Imagery dataset across all 109 subjects. First-author paper accepted at an IEEE conference. Read his story.
- Ashwajit W. built a 1,000-neuron spiking network model of memory replay and consolidation, confirmed 3 of 3 pre-stated predictions across more than 700 simulation runs, published in Springer Lecture Notes in Electrical Engineering, and presented it at BICA 2026 in Merida, Mexico. Read his story.
- Devansh G. studied ECG microstate instability as a possible biomarker for Major Depressive Disorder and became a first-author IEEE author at 14. Read his story.
- Ayaan R. adapted astrobiology signal processing to detect Alzheimer's patterns, reached 88.9% accuracy, was published in IEEE, and placed 3rd at his science fair. Read his story.
- Vaanya S. wrote "AlzAware-MSA-Net: Lightweight Dual-Path Morphology-Aware Multi-Scale Attention for Ordinal Alzheimer MRI Classification," accepted at an IEEE conference.
- Shivi K. had an oral presentation and publication accepted at the 2026 Congress of Neurological Surgeons Annual Meeting in Washington, DC, as the only presenter under 25, and was then invited to give a research talk at Yale School of Medicine.
None of these students needed a university lab. Every one of them needed a question narrow enough to finish and a mentor who knew the field.
Where student neuroscience papers get published
Engineering and medicine conferences. IEEE runs many conferences where neuroscience, signal processing and medical AI papers are accepted, including EMBC. Conference papers are peer reviewed, usually move faster than journals, and appear in IEEE Xplore once published.
Springer and Elsevier proceedings and journals. Cognitive architecture and computational neuroscience conferences often publish proceedings in Springer series, as with Ashwajit's paper.
Clinical society meetings. Neurosurgery and neurology meetings accept abstracts and presentations, as Shivi's CNS acceptance shows. These are harder to reach and usually need a clinical angle.
Student journals and preprints. The Journal of Emerging Investigators publishes hypothesis-driven work by students before university, with a mentor involved. bioRxiv takes biology and neuroscience preprints, which are public but not peer reviewed.
For a full list and how to judge a venue, see our guide to journals that publish high school research.
Neuroscience at science competitions
Regeneron ISEF has no standalone neuroscience category. Neuroscience projects enter through subcategories in other categories:
| ISEF category | Neuroscience subcategory |
|---|---|
| Behavioral and Social Sciences | Behavioral Neuroscience |
| Cellular and Molecular Biology | Neurobiology |
| Computational Biology and Bioinformatics | Computational Neuroscience |
Disease detection projects can also fit Translational Medical Science, which has a Disease Detection and Diagnosis subcategory. Pick the category by your method, not your topic: an EEG classifier and a behavioral study on the same disease belong in different places.
JSHS judges students on a research talk followed by questions, which suits neuroscience projects with a clear story. See how to win JSHS.
The International Brain Bee is a neuroscience competition for teens that runs from local Brain Bees to national rounds to a World Championship. It is a knowledge competition rather than a research competition, but it is a strong signal of interest alongside a paper.
Common mistakes in student neuroscience projects
- Choosing "the brain" as a topic. Pick one signal, one task, one group.
- Reporting only best-case accuracy. Reviewers trust results that hold across all subjects, not just the five cleanest.
- Data leakage. If recordings from the same person appear in training and testing, the accuracy is inflated. Split by subject.
- Ignoring consent. A behavioral study without proper review cannot enter most fairs.
- Claiming a diagnosis tool. Say what the model detects in your dataset. Leave clinical claims to clinical trials.
Frequently asked questions
Can I do neuroscience research in high school without a lab?
Yes. Public EEG, MRI and eye-tracking datasets, plus computational models, cover most of the neuroscience papers students publish. Every YRI Fellow listed above worked without a university lab.
Do I need to know how to code?
It helps, but you do not need to start with it. Python is the main language for signal processing and machine learning, and many Fellows learned what they needed during the project with a mentor's help.
Is neuroscience research the same as psychology research?
They overlap. Psychology research usually measures behavior with surveys and experiments. Neuroscience research usually measures the brain or nervous system directly, through signals, scans or models. See our psychology research guide for the survey and experiment path.
What ISEF category should a neuroscience project enter?
It depends on the method. Behavioral studies go to Behavioral and Social Sciences, signal and modeling projects often go to Computational Biology and Bioinformatics, and disease detection can go to Translational Medical Science.
How long does a neuroscience project take?
With public data and a mentor, a focused project can reach a full paper draft in about 10 weeks. Behavioral studies that need participant approval take longer.
How the YRI Fellowship helps you publish neuroscience research
The YRI Fellowship pairs you 1:1 with a PhD mentor in your field for 10 weeks of weekly sessions, about 5 to 8 hours of work a week. Your mentor helps you narrow your question, choose a dataset or model, and run the analysis. An editing team works through every draft, and a publishing team gets the paper into IEEE, Springer or Elsevier journals and conferences, with free help until it is published.
87% of Fellows publish and 73% place at science competitions. The Fellowship includes ISEF and JSHS preparation and a letter of recommendation from your mentor. No research experience is needed, and students from middle school up apply. Decisions come within 48 hours.
Tuition is $2,997, or 3 payments of $999, with scholarships considered case by case. See pricing, or read more student stories from Fellows who started where you are now.
Sources
- Society for Science: ISEF categories and subcategories
- Society for Science: ISEF human participants rules
- PhysioNet (physiological and EEG datasets)
- OpenNeuro (open neuroimaging and EEG data)
- International Brain Bee
- Journal of Emerging Investigators
- bioRxiv
- YRI Published Work
- YRI case study: Suriya D.
- YRI case study: Aditya A.
- YRI case study: Ashwajit W.
- YRI case study: Devansh G.
- YRI case study: Ayaan R.
