Seven tracks for studying and exploring neurotech, one buildable project at a time. Each folder is a self-contained subproject with its own study notes and build path.
No BCI hardware (OpenBCI, Muse, etc.) is available — projects that would normally use live acquisition are scoped to run on recorded/public datasets instead (see the hardware note in each affected project's README).
Tracks 06 and 07 are deliberately narrow: each was scoped around a specific set of applied skills (06: stimulation and prosthetics modeling; 07: clinical and health data engineering) rather than as a general survey of the field — see each README's scope note.
| # | Track | Starter project |
|---|---|---|
| 01 | Applied ML on Neuro Data | Sleep stage classification |
| 02 | Computational Neuroscience | Hodgkin–Huxley model |
| 03 | BCI / Signal Processing | Focus / attention classifier |
| 04 | Connectomics | Graph-theoretic analysis (C. elegans) or Tracing QA |
| 05 | Multimodal Biosensing | Multimodal emotion/state fusion (Galea-style) |
| 06 | Neurostimulation & Prosthetics |
NEURON stimulation modeling |
| 07 | Clinical / Health Data Engineering |
Healthcare data standards interop (EDF → FHIR) |
- 01-A Sleep stage classification — clean labels, classic EEG pipeline (filter → features → classifier).
- 02-A Hodgkin–Huxley — equations-level single-neuron intuition.
- 03-A Focus / attention classifier — consumer EEG + product narrative (FocusAnalyze).
- 04-B Tracing QA tool — connectomics + NeuroGlass-relevant tooling (or 04-A for a faster graph win).
Then continue with B / C / D (and E for MEG) in each track as you go deeper.
- Tracks are numbered
01–04; projects inside each track use letter prefixesA–E(starter =Awhere marked). - Each project folder has a
README.md(goal, dataset, stack, milestones). - Put notebooks in
notebooks/, scripts insrc/, figures infigures/, and local data underdata/(gitignored when you init repos). - Prefer small, demable v1s over large unfinished systems.