231 terms · one sentence each, more where it helps
Glossary
Every dotted term on the site points here. Each entry is one plain sentence first. Where the one sentence hides something important, a second paragraph follows. Entries also say where the term is first introduced and which lessons use it.
Phase 0: Orientation
- ADS1299#
- Texas Instruments' 8-channel, 24-bit biopotential front-end chip used in OpenBCI and many research amplifiers.Related: Delta-sigma ADC, OpenBCIUsed in: The six spines, and why you need one, Why being in Utah changes the plan, Set up a bench, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Project C: Make your own electrodes, Instrumentation amplifiers, properly, Analog to digital, Noise, and where the microvolt goes, Electrical safety, with the numbers, KiCad to JLCPCB, Firmware and SPI for a biopotential chip, Project B: Build a lab a rig, Project C: Other windows on the brain, The industry landscape, The math track
- Biodesign#
- Stanford's process for medical technology innovation that starts from an observed clinical need rather than a technology looking for a use.Related: Needs finding, Need statementUsed in: The six spines, and why you need one, Needs finding, The reading track, Career forks
- Brain-computer interface (BCI)#
- A system that turns measured brain activity into control of something outside the body, from a cursor to a wheelchair to a speech synthesizer.Related: Decoder, Closed loopUsed in: The six territories of neuroengineering, The six spines, and why you need one, Why being in Utah changes the plan, How to read a paper, Project E: The zero-dollar version, Alpha and the back of the head, Project F: The software pipeline, Digital filters, and what they do to time, Artifact removal, three ways, The characterization report, Project A: The SSVEP speller, Project F: Reproduce a paper, Event-related potentials and oscillations, Classification done honestly, Bayesian decoding and the Kalman filter, Deep learning for EEG, Statistics for small experiments, Synapses and plasticity, Neural coding, Where oscillations come from, Motor cortex and the systems BCIs target, Sensory systems as prosthesis targets, A reading order for Kandel, Project A: A speller for someone with ALS, Regulatory pathways, Privacy, neuroethics, and the abandoned implant, Project E: Intracortical data, Electrode materials and the foreign body response, Packaging, power, and telemetry, The Utah Array, a history, The industry landscape, The research ladder, The clinic, The math track, The reading track, Community, Career forks
- Characterization#
- Measuring what your instrument actually does, such as its noise floor, bandwidth, and rejection, instead of assuming the datasheet.Related: Noise floor, Phantom, Common-mode rejection ratio (CMRR)Used in: The six spines, and why you need one, How to read a datasheet, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Project D: The phantom head, The characterization report, Design controls and the design history file, Project A: Electrode fabrication, Project B: Build a lab a rig, Packaging, power, and telemetry, The research ladder, Portfolio and writing
- CITI training#
- The online human-subjects research course most universities require before you can touch participant data.Related: Institutional review board (IRB)Used in: The safety charter, Project D: Your own ERPs
- Cochlear implant#
- The most successful neural prosthesis: an electrode array in the inner ear that stimulates the auditory nerve in up to 22 bands.Related: Vocoder (noise vocoder)Used in: The six territories of neuroengineering, Project C: Prosthesis simulators, Synapses and plasticity, Sensory systems as prosthesis targets, A reading order for Kandel, Regulatory pathways, Privacy, neuroethics, and the abandoned implant, Neuromodulation, mechanisms and evidence, Stimulation safety and electrochemistry, The industry landscape, The clinic
- Common-mode rejection ratio (CMRR)#
- How much better an amplifier responds to a difference between its inputs than to a signal shared by both, in decibels; 100 dB means 100,000 times.Related: Common-mode signal, Electrode impedanceUsed in: The six spines, and why you need one, How to read a datasheet, Why so small?, Project B: Reverse-engineer a bought board, Project D: The phantom head, Instrumentation amplifiers, properly, The characterization report
More
A high CMRR is necessary but not sufficient. If the two electrodes have different impedances, the common-mode hum is turned into a real difference at the amplifier's inputs, and no amount of amplifier quality can remove it. That is why electrode preparation matters as much as the chip. - Cross-validation#
- Testing a model on data it never saw during training, repeatedly, so the reported accuracy means something.Related: Data leakage, OverfittingUsed in: The six spines, and why you need one, How to read a paper, Project B: The P300 speller, Project C: Motor imagery, Project F: Reproduce a paper, Classification done honestly, Project E: Large-scale neural data, Project C: Silent speech, The reading track
- Datasheet#
- The manufacturer's document for a chip: pinout, limits, typical performance, and the reference circuit you should start from.Used in: The safety charter, How to read a datasheet, Why so small?, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Instrumentation amplifiers, properly, Analog to digital, Noise, and where the microvolt goes, Electrical safety, with the numbers, KiCad to JLCPCB, Firmware and SPI for a biopotential chip, The characterization report, Project C: Other windows on the brain
- Decoder#
- The algorithm that maps neural signals to intended actions, such as a classifier or a Kalman filter.Related: Classifier, Kalman filter, Brain-computer interface (BCI)Used in: The six territories of neuroengineering, The six spines, and why you need one, How to read a paper, Digital filters, and what they do to time, Project A: The SSVEP speller, Project F: Reproduce a paper, Event-related potentials and oscillations, Bayesian decoding and the Kalman filter, Project D: Neuromatch Academy, Project E: Large-scale neural data, Synapses and plasticity, Neural coding, Motor cortex and the systems BCIs target, Project A: A speller for someone with ALS, Project C: Silent speech, Design controls and the design history file, Regulatory pathways, Privacy, neuroethics, and the abandoned implant, Project E: Intracortical data, Electrode materials and the foreign body response, The Utah Array, a history, The industry landscape, The math track, Career forks
- Deep brain stimulation (DBS)#
- Electrodes implanted deep in the brain and pulsed continuously, the standard treatment for advanced Parkinson's disease and essential tremor.Related: Charge densityUsed in: The six territories of neuroengineering, What was that sound?, Where oscillations come from, A reading order for Kandel, Neuromodulation, mechanisms and evidence, Stimulation safety and electrochemistry, Packaging, power, and telemetry, The industry landscape, The clinic
- Design controls#
- FDA's required process for developing a medical device: user needs, requirements, design, verification, validation, and the records of all of it.Related: Design history file (DHF), Requirements, Verification, ValidationUsed in: The six spines, and why you need one, Project A: A speller for someone with ALS, Design controls and the design history file, The reading track, Career forks
- Design history file (DHF)#
- The complete record of how a device was designed, showing every requirement traced to a test; writing a mock one teaches more than any lecture.Related: Design controlsUsed in: The six spines, and why you need one, Design controls and the design history file, Career forks
- Driven-right-leg circuit#
- A third electrode fed with an inverted copy of the common-mode signal, so the body's hum is actively cancelled instead of merely rejected.Related: Common-mode signal, Ground electrodeUsed in: The notebook, and the build log, How to read a datasheet, Project B: One channel of you, Project C: Muscle to motion, Project D: Eyes as a joystick, Why so small?, Why the hum?, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Instrumentation amplifiers, properly, The math track
More
The name comes from ECG, where the third electrode went on the right leg. In EEG it usually goes on the earlobe or forehead. It is a feedback loop: measure the average of the two inputs, invert, amplify, drive the body with it. Done right it buys 30 to 40 dB of extra hum rejection. - Electrocorticography (ECoG)#
- Recording from electrodes placed on the surface of the brain under the skull, usually during epilepsy surgery evaluation; cleaner and sharper than EEG.Related: Electroencephalography (EEG), Intracortical recording, ECoG gridUsed in: The six territories of neuroengineering, Electrode materials and the foreign body response
- Electrode#
- The conductor that turns ion currents in tissue into electron currents in a wire; its chemistry and contact quality set the noise you live with.Related: Silver/silver-chloride (Ag/AgCl), Electrode impedance, Half-cell potentialUsed in: The six territories of neuroengineering, The six spines, and why you need one, Why being in Utah changes the plan, Set up a bench, The safety charter, The notebook, and the build log, How to read a datasheet, How to read a paper, Project A: The SpikerBox, Project B: One channel of you, Project C: Muscle to motion, Project D: Eyes as a joystick, Project E: The zero-dollar version, What was that sound?, Why so small?, Why the hum?, Alpha and the back of the head, The artifact bestiary, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Project C: Make your own electrodes, Project D: The phantom head, Project E: The eight-channel EMG armband, Project F: The software pipeline, Instrumentation amplifiers, properly, The electrode-electrolyte interface, Analog to digital, Noise, and where the microvolt goes, Electrical safety, with the numbers, Artifact removal, three ways, Spatial filters and montages, The characterization report, Project A: The SSVEP speller, Project C: Motor imagery, Project E: The sleep band, Project H: Play, Event-related potentials and oscillations, Classification done honestly, Designing an experiment that can fail, Project B: The forward and inverse problems, Project C: Prosthesis simulators, Synapses and plasticity, Neural coding, Where oscillations come from, Motor cortex and the systems BCIs target, Sensory systems as prosthesis targets, A reading order for Kandel, Project A: A speller for someone with ALS, Project B: Control for a real user, Project C: Silent speech, Project E: Haptic feedback for a prosthetic hand, Design controls and the design history file, Risk management and the FMEA, Human factors and use error, Regulatory pathways, Privacy, neuroethics, and the abandoned implant, Project A: Electrode fabrication, Project B: Build a lab a rig, Project C: Other windows on the brain, Project D: Stimulation, studied properly, Project E: Intracortical data, Neuromodulation, mechanisms and evidence, Stimulation safety and electrochemistry, Electrode materials and the foreign body response, Packaging, power, and telemetry, The Utah Array, a history, The industry landscape, The clinic, The reading track, Career forks
- Electrode impedance#
- How hard it is to push current through the skin-electrode contact, typically 5 to 50 kilohms with gel and megohms dry, higher at low frequencies.Related: Dry electrode, Conductive gel or paste, Electrochemical impedance spectroscopy (EIS)Used in: How to read a datasheet, Why so small?, Project C: Make your own electrodes, Instrumentation amplifiers, properly, Noise, and where the microvolt goes, The characterization report, Classification done honestly, Project D: Stimulation, studied properly, Stimulation safety and electrochemistry
- Electroencephalography (EEG)#
- Recording the brain's electrical activity from electrodes on the scalp, a few tens of microvolts of blurred, summed cortical activity.Related: Electrocorticography (ECoG), Magnetoencephalography (MEG), Volume conduction, Microvolt (µV)Used in: The six territories of neuroengineering, Set up a bench, The notebook, and the build log, How to read a datasheet, Project B: One channel of you, Project C: Muscle to motion, Project D: Eyes as a joystick, Project E: The zero-dollar version, Why so small?, Why the hum?, What the spectrum did, Alpha and the back of the head, The artifact bestiary, Soldering, breadboards, and the multimeter, Project A: The ADS1299 board, Project C: Make your own electrodes, Project D: The phantom head, Project E: The eight-channel EMG armband, Project F: The software pipeline, Instrumentation amplifiers, properly, The electrode-electrolyte interface, Analog to digital, Noise, and where the microvolt goes, Electrical safety, with the numbers, Artifact removal, three ways, Spatial filters and montages, KiCad to JLCPCB, Lab Streaming Layer and PsychoPy, Project A: The SSVEP speller, Project D: Your own ERPs, Project E: The sleep band, Project G: Neurofeedback, and how to test whether it does anything, Project H: Play, Event-related potentials and oscillations, Classification done honestly, Bayesian decoding and the Kalman filter, Deep learning for EEG, Designing an experiment that can fail, Project B: The forward and inverse problems, Where oscillations come from, Sleep neuroscience, A reading order for Kandel, Project A: A speller for someone with ALS, Project B: Control for a real user, Needs finding, Regulatory pathways, Privacy, neuroethics, and the abandoned implant, Project B: Build a lab a rig, Project C: Other windows on the brain, Electrode materials and the foreign body response, The industry landscape, The clinic, Portfolio and writing, The math track, Community
- Electromyography (EMG)#
- Recording the electrical activity of muscles, hundreds of microvolts to millivolts, much larger than EEG.Related: Myoelectric prosthesis, Electroencephalography (EEG)Used in: The six territories of neuroengineering, How to read a datasheet, Project C: Muscle to motion, Project D: Eyes as a joystick, Tool: The Web Serial plotter, Why so small?, Why the hum?, The artifact bestiary, Project E: The eight-channel EMG armband, The electrode-electrolyte interface, Analog to digital, Noise, and where the microvolt goes, Classification done honestly, Sleep neuroscience, Project B: Control for a real user, Project C: Silent speech, Project E: Haptic feedback for a prosthetic hand, Project F: An adaptive game controller, Risk management and the FMEA, The industry landscape, The clinic
- Epilepsy#
- A tendency to recurrent seizures; its monitoring units are where seizure-detection algorithms get validated and where ECoG data comes from.Related: Epilepsy monitoring unit (EMU), Seizure detectionUsed in: The six territories of neuroengineering, Why being in Utah changes the plan, Project A: The SSVEP speller, Neural coding, Where oscillations come from, A reading order for Kandel, Needs finding, Privacy, neuroethics, and the abandoned implant, Neuromodulation, mechanisms and evidence, The industry landscape, The clinic
- Extracellular recording#
- Recording with an electrode near neurons rather than inside one; you see small, fast blips from nearby cells rather than the full membrane swing.Related: Spike, Local field potential (LFP), Intracortical recordingUsed in: The six territories of neuroengineering, Project A: The SpikerBox, What was that sound?
- FDA device classes#
- Class I is low risk (a tongue depressor), Class II moderate (most EEG amplifiers), Class III highest (implanted BCIs), each with heavier review.Related: 510(k), Premarket approval (PMA), De NovoUsed in: The six spines, and why you need one, Regulatory pathways
- Filter#
- An operation that passes some frequencies and attenuates others; every filter also delays or distorts the timing of what it passes.Related: High-pass filter, Low-pass filter, Notch filter, Phase distortionUsed in: How to read a paper, Project A: The SpikerBox, Project B: One channel of you, Project C: Muscle to motion, Project E: The zero-dollar version, Why so small?, What the spectrum did, The artifact bestiary, Soldering, breadboards, and the multimeter, Python for signals, Project B: Reverse-engineer a bought board, Project D: The phantom head, Project F: The software pipeline, Analog to digital, Digital filters, and what they do to time, Artifact removal, three ways, Spatial filters and montages, Project C: Motor imagery, Project D: Your own ERPs, Project F: Reproduce a paper, Project H: Play, Classification done honestly, Bayesian decoding and the Kalman filter, Deep learning for EEG, Statistics for small experiments, Project B: The forward and inverse problems, Project C: Prosthesis simulators, Project E: Large-scale neural data, Project E: Haptic feedback for a prosthetic hand, The math track
- Firmware#
- The program running on the microcontroller, which configures the chip, reads samples on time, and ships them to the computer.Related: SPIUsed in: The six spines, and why you need one, Tool: The Web Serial plotter, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Digital filters, and what they do to time, Firmware and SPI for a biopotential chip, Project B: Build a lab a rig, Project C: Other windows on the brain, The industry landscape
- FMEA#
- Failure mode and effects analysis: a table of every way the device could fail, how bad and how likely each is, and what you did about it.Related: ISO 14971Used in: The six spines, and why you need one, Electrical safety, with the numbers, Project A: A speller for someone with ALS, Project B: Control for a real user, Competitions: Bench to Bedside, DEBUT, and RESNA, Design controls and the design history file, Risk management and the FMEA, Career forks
- Focused ultrasound#
- Concentrating sound energy deep in the brain to modulate or destroy tissue without opening the skull.Used in: The six territories of neuroengineering, Neuromodulation, mechanisms and evidence, The industry landscape
- Foreign body response#
- The scarring the body builds around an implant, which insulates electrodes and degrades recordings over months to years.Related: Intracortical recordingUsed in: The six territories of neuroengineering, Synapses and plasticity, Project E: Intracortical data, Electrode materials and the foreign body response
- Functional near-infrared spectroscopy (fNIRS)#
- Shining near-infrared light through the scalp and measuring how much comes back to estimate blood oxygenation in the cortex underneath; a pulse oximeter pointed at the brain.Related: Electroencephalography (EEG)Used in: The six territories of neuroengineering, The safety charter, Project C: Other windows on the brain, The industry landscape
- Gain#
- How many times an amplifier multiplies its input; 1000× turns 50 microvolts into 50 millivolts, enough for an ADC to see.Related: Instrumentation amplifierUsed in: The notebook, and the build log, How to read a datasheet, Project A: The SpikerBox, Project B: One channel of you, Project C: Muscle to motion, Project D: Eyes as a joystick, Why so small?, Why the hum?, Alpha and the back of the head, Soldering, breadboards, and the multimeter, Python for signals, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Project C: Make your own electrodes, Project D: The phantom head, Project E: The eight-channel EMG armband, Instrumentation amplifiers, properly, Analog to digital, Noise, and where the microvolt goes, Digital filters, and what they do to time, Firmware and SPI for a biopotential chip, The characterization report, Project E: The sleep band, Bayesian decoding and the Kalman filter, Deep learning for EEG, Synapses and plasticity, The math track
- Haptic feedback#
- Information delivered through touch, such as vibration on the skin that tells a prosthesis user how hard they are gripping.Used in: The six territories of neuroengineering, Sensory systems as prosthesis targets, Project E: Haptic feedback for a prosthetic hand, Competitions: Bench to Bedside, DEBUT, and RESNA
- Hermetic packaging#
- Sealing implanted electronics against body fluid for decades, one of the hardest unsolved engineering problems in the field.Used in: The six spines, and why you need one, Packaging, power, and telemetry
- Human factors#
- Designing so the device fits how people actually behave, especially under stress and fatigue, so that errors are hard to make.Related: IEC 62366, Use errorUsed in: The six spines, and why you need one, Project A: A speller for someone with ALS, Human factors and use error, The industry landscape, Career forks
- IEC 60601#
- The safety standard for medical electrical equipment, including how much current may leak into a patient (microamps).Related: Leakage current, Electrical isolationUsed in: The safety charter, Project B: One channel of you, Why the hum?, Electrical safety, with the numbers, KiCad to JLCPCB
- Input bias current#
- The tiny current an amplifier input draws; through a high electrode impedance it produces an offset and noise.Related: Input impedanceUsed in: How to read a datasheet
- Input impedance#
- How hard the amplifier's inputs are to push current into; it must be far higher than the electrode impedance or the signal is lost in the divider.Related: Electrode impedance, Input bias currentUsed in: How to read a datasheet, Project B: One channel of you, Why so small?, Instrumentation amplifiers, properly
- Institutional review board (IRB)#
- The committee that must approve any research on people before it starts, to protect the participants.Related: Informed consent, CITI trainingUsed in: The safety charter, Project A: The SSVEP speller, Project B: The P300 speller, Project C: Motor imagery, Project D: Your own ERPs, Project E: The sleep band, Project G: Neurofeedback, and how to test whether it does anything, Designing an experiment that can fail, Project A: A speller for someone with ALS, Regulatory pathways, The Utah Array, a history
- Instrumentation amplifier#
- A precision differential amplifier with very high input impedance and high CMRR, built from three op-amps and sold as a single chip such as the INA128 or AD8232.Related: Differential measurement, Common-mode rejection ratio (CMRR), Input impedanceUsed in: Set up a bench, How to read a datasheet, Project A: The SpikerBox, Project B: One channel of you, Project C: Muscle to motion, Why so small?, Soldering, breadboards, and the multimeter, Project A: The ADS1299 board, Instrumentation amplifiers, properly, Noise, and where the microvolt goes
- Intracortical recording#
- Recording from tiny electrodes inserted into the cortex itself, close enough to hear individual neurons.Related: Utah Array, Extracellular recordingUsed in: The six territories of neuroengineering, Project E: Large-scale neural data, Privacy, neuroethics, and the abandoned implant, Project E: Intracortical data
- ISO 14971#
- The international standard for medical device risk management that every company follows.Related: FMEAUsed in: The safety charter, Risk management and the FMEA
- JLCPCB#
- A Chinese board house that fabricates and assembles small-run PCBs for a few dollars, the standard for student projects.Related: Printed circuit board (PCB)Used in: Set up a bench, Project A: The ADS1299 board, KiCad to JLCPCB
- Leakage current#
- Current that flows into the patient through unintended paths; limited to tens of microamps for anything touching the body.Related: IEC 60601, Electrical isolationUsed in: The safety charter, Electrical safety, with the numbers
- Magnetoencephalography (MEG)#
- Recording the tiny magnetic fields produced by cortical currents with superconducting sensors in a shielded room; less blurred by the skull than EEG, and far more expensive.Related: Electroencephalography (EEG)Used in: The six territories of neuroengineering, Event-related potentials and oscillations, Project B: The forward and inverse problems, The math track
- Microvolt (µV)#
- One millionth of a volt. Scalp EEG is tens of microvolts; a AA battery is 1.5 million of them.Related: GainUsed in: The six territories of neuroengineering, Why so small?
- MNE-Python#
- The standard open-source Python library for EEG and MEG analysis: filtering, epoching, ICA, topographies, source localization.Related: Independent component analysis (ICA)Used in: Set up a bench, Project E: The zero-dollar version, The artifact bestiary, Python for signals, Project F: The software pipeline, Digital filters, and what they do to time, Artifact removal, three ways, Spatial filters and montages, Lab Streaming Layer and PsychoPy, Project C: Motor imagery, Project E: The sleep band, Statistics for small experiments, Project B: The forward and inverse problems
- Montage#
- The particular arrangement of which electrodes are subtracted from which, such as referential, bipolar, or average reference.Related: Reference electrode, Common average reference, Surface LaplacianUsed in: Set up a bench, Project E: The zero-dollar version, Why so small?, Python for signals, Project F: The software pipeline, Spatial filters and montages, Project B: The forward and inverse problems, The clinic, The math track
- Myoelectric prosthesis#
- An artificial limb controlled by EMG from the remaining muscles.Related: Electromyography (EMG)Used in: The six territories of neuroengineering, Project C: Muscle to motion, Project E: Haptic feedback for a prosthetic hand
- Needs finding#
- Observing clinical work to catch the moments that are slow, awkward, or unsafe, before proposing any solution.Related: BiodesignUsed in: The six spines, and why you need one, Needs finding
- Noise floor#
- What your instrument records when the input is shorted or connected to a perfectly quiet source; everything smaller than this is invisible to you.Related: Thermal (Johnson) noise, CharacterizationUsed in: Why being in Utah changes the plan, The notebook, and the build log, Project A: The SpikerBox, Project B: One channel of you, What the spectrum did, Soldering, breadboards, and the multimeter, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Project C: Make your own electrodes, Project D: The phantom head, Instrumentation amplifiers, properly, Analog to digital, Noise, and where the microvolt goes, The characterization report, Project B: Build a lab a rig, Portfolio and writing
- Notch filter#
- A very narrow filter that removes one frequency, usually 60 Hz mains hum; it also rings and distorts nearby content, so fix the hum at the source first.Related: Filter, Common-mode signalUsed in: How to read a datasheet, Project A: The SpikerBox, Project B: One channel of you, Project C: Muscle to motion, Why the hum?, Soldering, breadboards, and the multimeter, Digital filters, and what they do to time, Artifact removal, three ways, Project E: Haptic feedback for a prosthetic hand
- OpenBCI#
- The open-source EEG hardware company whose Cyton (ADS1299) and Ganglion boards are the default hobby and research-lite amplifiers.Related: ADS1299, BrainFlowUsed in: Set up a bench, Why so small?, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Project C: Make your own electrodes, Analog to digital, KiCad to JLCPCB, Firmware and SPI for a biopotential chip, Lab Streaming Layer and PsychoPy, The industry landscape, Community
- Phantom#
- A physical stand-in for tissue, such as saline gelatin with embedded wires, used to test an instrument against known signals.Related: CharacterizationUsed in: Set up a bench, Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Project D: The phantom head, Noise, and where the microvolt goes, Firmware and SPI for a biopotential chip, The characterization report, Project B: The forward and inverse problems
- PhysioNet#
- A free archive of physiological recordings including EEG, ECG, and sleep data, run out of MIT.Related: Sleep-EDFUsed in: Set up a bench, Project B: One channel of you, Project E: The zero-dollar version, Alpha and the back of the head, Python for signals, Project F: The software pipeline
- Printed circuit board (PCB)#
- The board that holds and connects your components; designing one is how a breadboard prototype becomes an instrument.Related: KiCad, JLCPCBUsed in: Set up a bench, KiCad to JLCPCB
- Recall#
- A correction or removal of a marketed device; FDA's public database of them is a catalogue of how designs fail people.Used in: The six spines, and why you need one, Project E: The sleep band, Sleep neuroscience, Design controls and the design history file, Risk management and the FMEA, Regulatory pathways, Electrode materials and the foreign body response
- Sampling rate#
- How many measurements per second the ADC takes; EEG is usually 250 to 1000 per second.Related: Nyquist frequency, AliasingUsed in: How to read a paper, What the spectrum did, Analog to digital
- Spike#
- Shorthand for an action potential as seen by an electrode, a brief blip about a millisecond wide.Related: Action potential, Spike sorting, Extracellular recordingUsed in: The six territories of neuroengineering, Project A: The SpikerBox, Project C: Muscle to motion, Project D: Eyes as a joystick, Project E: The zero-dollar version, Tool: The Web Serial plotter, What was that sound?, Why so small?, Why the hum?, What the spectrum did, The artifact bestiary, Instrumentation amplifiers, properly, Analog to digital, Noise, and where the microvolt goes, Digital filters, and what they do to time, Artifact removal, three ways, Project F: Reproduce a paper, Bayesian decoding and the Kalman filter, Project A: Neurons from the equations up, Project D: Neuromatch Academy, Project E: Large-scale neural data, Neuron biophysics, Synapses and plasticity, Neural coding, Where oscillations come from, A reading order for Kandel, Project B: Build a lab a rig, Project E: Intracortical data, Electrode materials and the foreign body response, Packaging, power, and telemetry
- SpikerBox#
- Backyard Brains' small battery amplifier that lets you hear action potentials from a cockroach leg on a speaker; easy to build yourself.Related: Action potential, Instrumentation amplifierUsed in: Set up a bench, The safety charter, Project A: The SpikerBox, Project B: One channel of you, A reading order for Kandel
- Spinal cord stimulation (SCS)#
- Electrodes over the spinal cord, long used for pain and now being used to restore stepping after spinal cord injury.Used in: The six territories of neuroengineering, Neuromodulation, mechanisms and evidence, Packaging, power, and telemetry
- Stentrode#
- Synchron's electrode array mounted on a stent and delivered through a blood vessel to sit against the motor cortex without open surgery.Used in: The six territories of neuroengineering, Electrode materials and the foreign body response, The industry landscape
- Transcranial direct current stimulation (tDCS)#
- Passing a milliamp or two through the head between sponge electrodes; trivially easy to build, weakly supported by evidence, and not for self-experiment.Related: Transcranial magnetic stimulation (TMS)Used in: The safety charter, Neuromodulation, mechanisms and evidence
- Transcranial magnetic stimulation (TMS)#
- A coil on the head that induces currents in the cortex with a brief magnetic pulse; approved for depression, and only for lab use with a trained operator.Related: Transcranial direct current stimulation (tDCS)Used in: The six territories of neuroengineering, The safety charter, Neuromodulation, mechanisms and evidence, The industry landscape
- Utah Array#
- The 10-by-10 grid of 1.5 mm silicon needles invented at the University of Utah, the electrode in nearly every human intracortical BCI study to date.Related: Intracortical recording, Utah Slanted Electrode Array (USEA)Used in: The six territories of neuroengineering, Why being in Utah changes the plan, Project E: Intracortical data, Electrode materials and the foreign body response, Packaging, power, and telemetry, The Utah Array, a history, The industry landscape
- Vagus nerve stimulation (VNS)#
- A cuff electrode on the vagus nerve in the neck, pulsed for epilepsy, depression, and now stroke rehabilitation.Used in: The six territories of neuroengineering, Neuromodulation, mechanisms and evidence, The industry landscape
- Verification#
- Testing that the device meets its requirements; did you build what you specified.Related: ValidationUsed in: Why being in Utah changes the plan, The characterization report, Design controls and the design history file
Phase 1: First contact
- Action potential#
- The brief, all-or-nothing voltage spike a neuron fires, about a millisecond long and about 100 millivolts tall across its membrane.Related: Membrane potential, Ion channel, Spike, Extracellular recordingUsed in: Project A: The SpikerBox, What was that sound?, Project A: Neurons from the equations up, Neuron biophysics, A reading order for Kandel
More
Sodium channels open and the inside of the cell swings positive; potassium channels open and it swings back. From outside the cell, with an electrode nearby, you see a much smaller, faster blip, which is what a SpikerBox amplifies into a click. - Active electrode#
- An electrode with a small amplifier or buffer built into it, so the high-impedance signal travels only millimetres before being made robust.Related: Dry electrode, Input impedanceUsed in: Why so small?, Project C: Make your own electrodes, The electrode-electrolyte interface
- Aliasing#
- A frequency above Nyquist masquerading as a lower one after sampling; nothing downstream can undo it.Related: Nyquist frequency, Anti-alias filterUsed in: Why the hum?, Analog to digital, Digital filters, and what they do to time
- Alpha rhythm#
- A roughly 8 to 12 Hz oscillation over the back of the head that appears when the eyes close and fades when they open.Related: Beta rhythm, Theta rhythm, Thalamocortical loop, Sensorimotor rhythm (mu)Used in: Project B: One channel of you, Why so small?, Alpha and the back of the head, Project A: The ADS1299 board, Analog to digital, Artifact removal, three ways
More
Alpha is the first thing anyone sees in their own EEG. It arises from thalamocortical loops in visual cortex idling; opening the eyes gives that cortex something to do and the rhythm breaks up. The same eyes-closed test is how you check that an electrode setup is actually recording brain. - Amyotrophic lateral sclerosis (ALS)#
- A progressive disease of motor neurons that removes movement and eventually speech while leaving thought intact; the population many BCIs are built for.Related: Locked-in syndrome, Augmentative and alternative communication (AAC)Used in: Project D: Eyes as a joystick, Motor cortex and the systems BCIs target, Project A: A speller for someone with ALS, Needs finding, Design controls and the design history file, Human factors and use error, The clinic
- Analog-to-digital converter (ADC)#
- The chip that measures a voltage at regular instants and turns each into a number.Related: Sampling rate, Quantization, Delta-sigma ADCUsed in: Project B: One channel of you, Tool: The Web Serial plotter, Why so small?, Why the hum?, Analog to digital
- Anti-alias filter#
- A low-pass filter placed before the ADC to remove everything above Nyquist, so nothing can alias.Related: Aliasing, Low-pass filterUsed in: Project B: One channel of you, Why the hum?, Analog to digital
- Artifact#
- Anything in the recording that did not come from the brain: blinks, muscle, heartbeat, movement, sweat, mains hum, electrode pops.Related: Blink artifact, Independent component analysis (ICA), Artifact subspace reconstruction (ASR)Used in: Project B: One channel of you, Project C: Muscle to motion, Project D: Eyes as a joystick, Tool: The Web Serial plotter, The artifact bestiary, Project F: The software pipeline, The electrode-electrolyte interface, Digital filters, and what they do to time, Artifact removal, three ways, Spatial filters and montages, Project C: Motor imagery, Designing an experiment that can fail, Project D: Neuromatch Academy, Project C: Other windows on the brain, The clinic
- Band-pass filter#
- A high-pass and a low-pass together, keeping only a band such as 0.5 to 40 Hz.Related: FilterUsed in: Project A: The SpikerBox, Deep learning for EEG
- Calibration#
- The session at the start of BCI use in which the decoder is trained on this person, today, because brains and electrodes drift.Related: Transfer learningUsed in: Project C: Muscle to motion, Project E: The eight-channel EMG armband, Artifact removal, three ways, Project B: The P300 speller, Project C: Motor imagery, Classification done honestly, Bayesian decoding and the Kalman filter, Privacy, neuroethics, and the abandoned implant, Project E: Intracortical data
- Classifier#
- An algorithm that assigns each trial a label, such as left hand or right hand, after being trained on labelled examples.Related: Linear discriminant analysis (LDA), Support vector machine (SVM), Cross-validationUsed in: Project C: Muscle to motion, Python for signals, Project E: The eight-channel EMG armband, Artifact removal, three ways, Spatial filters and montages, Project B: The P300 speller, Project C: Motor imagery, Project E: The sleep band, Project F: Reproduce a paper, Event-related potentials and oscillations, Classification done honestly, Bayesian decoding and the Kalman filter, Deep learning for EEG, Statistics for small experiments, A reading order for Kandel, Project B: Control for a real user, Project C: Silent speech, Risk management and the FMEA, The math track, The reading track
- Common-mode signal#
- A voltage that appears equally on both inputs of a differential amplifier, such as 60 Hz hum picked up by the whole body.Related: Common-mode rejection ratio (CMRR), Differential measurement, Driven-right-leg circuitUsed in: Project B: One channel of you, Why so small?, Why the hum?, Instrumentation amplifiers, properly, KiCad to JLCPCB, The characterization report
- Differential measurement#
- Measuring the difference between two electrodes rather than one electrode against ground, so anything common to both cancels.Related: Common-mode signal, Common-mode rejection ratio (CMRR), Instrumentation amplifierUsed in: Why so small?
- EEG technologist#
- The person who applies electrodes and runs clinical EEGs, with a certification path of its own.Used in: Alpha and the back of the head, The artifact bestiary, The clinic
- Electrocardiography (ECG)#
- Recording the heart's electrical activity, about a millivolt on the chest; it leaks into EEG as a small regular spike.Related: ArtifactUsed in: Why so small?, The artifact bestiary, Electrical safety, with the numbers
- Electrooculography (EOG)#
- Recording eye position from the standing voltage across the eyeball, which moves with the eyes; also the main source of blink artifacts in EEG.Related: Blink artifact, Electroencephalography (EEG)Used in: Project D: Eyes as a joystick, Why so small?, The artifact bestiary, Artifact removal, three ways, Project A: A speller for someone with ALS, Project F: An adaptive game controller, The clinic
- Epoch#
- A fixed-length slice of the recording locked to an event, such as one second starting 200 milliseconds before each stimulus.Related: Trial averaging, Baseline correctionUsed in: Python for signals, Project F: The software pipeline, Digital filters, and what they do to time, Artifact removal, three ways, Lab Streaming Layer and PsychoPy, Project B: The P300 speller, Project E: The sleep band, Project F: Reproduce a paper, Event-related potentials and oscillations, Deep learning for EEG, Designing an experiment that can fail
- Event-related desynchronization (ERD)#
- The drop in rhythmic power over a cortical area when it becomes active, such as mu weakening when you imagine moving your hand.Related: Sensorimotor rhythm (mu), Motor imageryUsed in: Alpha and the back of the head, Project F: The software pipeline, Event-related potentials and oscillations
- Event-related potential (ERP)#
- The small, stereotyped voltage waveform the brain produces in response to an event, invisible in a single trial and revealed by averaging many.Related: P300, N170, Error-related negativity (ERN), Trial averagingUsed in: Why so small?, The artifact bestiary, Digital filters, and what they do to time, Artifact removal, three ways, Spatial filters and montages, Lab Streaming Layer and PsychoPy, Project D: Your own ERPs, Project H: Play, Event-related potentials and oscillations, Classification done honestly, Designing an experiment that can fail, Statistics for small experiments, The reading track
- Fast Fourier transform (FFT)#
- The algorithm that breaks a stretch of signal into the sinusoids that would add up to it, giving amplitude and phase at each frequency.Related: Spectrum, Power spectral density (PSD), Window functionUsed in: What the spectrum did
- Ground electrode#
- The electrode that ties the body to the amplifier's zero and, in a good design, is driven to cancel common-mode hum.Related: Driven-right-leg circuit, Reference electrodeUsed in: Why so small?
- Ground loop#
- Two paths to ground that form a loop through which stray currents flow, injecting hum into whatever is between them.Related: Electrical isolationUsed in: Why the hum?
- Hodgkin–Huxley model#
- The 1952 set of four differential equations for the squid axon that reproduces the action potential from voltage-dependent sodium and potassium channels.Related: Action potential, Leaky integrate-and-fire neuronUsed in: What was that sound?, Project A: Neurons from the equations up, Project D: Neuromatch Academy, Neuron biophysics, The math track
- Independent component analysis (ICA)#
- A method that unmixes multichannel data into statistically independent sources, so blinks and heartbeat can be identified and removed.Related: ArtifactUsed in: The artifact bestiary, Project F: The software pipeline, Artifact removal, three ways, Spatial filters and montages, Classification done honestly
- Ion channel#
- A protein pore in the membrane that lets specific ions through, often only when the voltage or a chemical tells it to open.Related: Action potential, Hodgkin–Huxley modelUsed in: What was that sound?, A reading order for Kandel
- Membrane potential#
- The voltage across a cell's membrane, about −70 millivolts at rest, set by ion concentrations and which channels are open.Related: Resting potential, Action potential, Ion channelUsed in: What was that sound?, Neuron biophysics, Synapses and plasticity, A reading order for Kandel, Neuromodulation, mechanisms and evidence
- Motor imagery#
- Imagining a movement without making it, which changes rhythms over motor cortex in a way a classifier can sometimes detect.Related: Sensorimotor rhythm (mu), Event-related desynchronization (ERD), BCI inefficiencyUsed in: Project E: The zero-dollar version, Alpha and the back of the head, Project F: The software pipeline, Spatial filters and montages, Project C: Motor imagery, Project F: Reproduce a paper, Project H: Play, Event-related potentials and oscillations, Classification done honestly, Deep learning for EEG, Project B: Control for a real user
- Nyquist frequency#
- Half the sampling rate; anything in the signal above it cannot be represented and will appear as something false.Related: Aliasing, Anti-alias filterUsed in: What the spectrum did, Analog to digital
- Power spectral density (PSD)#
- The spectrum expressed as power per hertz, the standard way to quantify how strong alpha is.Related: Welch's method, SpectrumUsed in: What the spectrum did
- Reference electrode#
- The electrode every other channel is measured against; there is no neutral spot on the head, so the choice shapes every trace.Related: Montage, Common average referenceUsed in: Project B: One channel of you, Why so small?, The artifact bestiary, Project A: The ADS1299 board, Project D: The phantom head, Spatial filters and montages, Human factors and use error
- Refractory period#
- The millisecond or so after a spike during which the neuron cannot fire again because its sodium channels are inactivated; it caps firing rates at a few hundred per second.Related: Action potential, Ion channelUsed in: What was that sound?, Project A: Neurons from the equations up, Project E: Large-scale neural data
- Resting potential#
- The steady membrane voltage of a neuron that is not firing, typically −60 to −70 millivolts.Related: Membrane potentialUsed in: What was that sound?, Neuron biophysics
- Sensorimotor rhythm (mu)#
- An 8 to 13 Hz rhythm over motor cortex that weakens when you move, or imagine moving, the corresponding body part.Related: Motor imagery, Event-related desynchronization (ERD)Used in: Alpha and the back of the head, Project C: Motor imagery
- Sleep technologist#
- The person who runs overnight sleep studies and scores them; a certification path an undergraduate can complete and work nights.Related: Polysomnography (PSG)Used in: Alpha and the back of the head, Sleep neuroscience, The clinic
- Source localization#
- Estimating where in the brain scalp signals came from, using a head model and assumptions such as smoothness or sparsity.Related: Inverse problemUsed in: Project E: The zero-dollar version, The math track
- Spectrogram#
- A spectrum computed over sliding windows and plotted against time, showing how frequency content changes.Related: Wavelet transform, SpectrumUsed in: What the spectrum did, Project E: The sleep band, Event-related potentials and oscillations, Project E: Large-scale neural data
- Spectrum#
- A plot of how much of a signal lives at each frequency.Related: Fast Fourier transform (FFT), Power spectral density (PSD)Used in: Project B: One channel of you, Project E: The zero-dollar version, Tool: The Web Serial plotter, What the spectrum did, Alpha and the back of the head, Python for signals, Analog to digital, Noise, and where the microvolt goes, The characterization report, Project A: The SSVEP speller, Event-related potentials and oscillations, Project A: Neurons from the equations up, Sensory systems as prosthesis targets, Privacy, neuroethics, and the abandoned implant, Project A: Electrode fabrication
- Spike sorting#
- Deciding which spikes on an electrode came from which neuron, by clustering their waveform shapes.Related: Spike, NeuropixelsUsed in: What was that sound?, Project E: Large-scale neural data
- Thalamocortical loop#
- The two-way circuit between the thalamus and the cortex whose rhythmic activity produces alpha, sleep spindles, and other EEG rhythms.Related: Alpha rhythm, Sleep spindleUsed in: Alpha and the back of the head
- Topographic map#
- The head-shaped colour plot of voltage or power at each electrode, interpolated between them.Related: MontageUsed in: Project E: The zero-dollar version, Project F: The software pipeline, Spatial filters and montages, Project B: The forward and inverse problems
- Welch's method#
- Estimating a spectrum by averaging the spectra of many overlapping short windows, trading frequency detail for a less noisy estimate.Related: Power spectral density (PSD), Window functionUsed in: What the spectrum did, Python for signals
- Window function#
- A smooth taper, such as a Hann window, applied to a segment before an FFT so its abrupt edges do not smear the spectrum.Related: Fast Fourier transform (FFT), Welch's methodUsed in: What the spectrum did
Phase 2: Build the instrument
- 1/f noise#
- Noise that grows at low frequencies, present in amplifiers and electrodes, and the reason sub-hertz EEG is hard.Related: Noise floorUsed in: Instrumentation amplifiers, properly, Noise, and where the microvolt goes
- Artifact subspace reconstruction (ASR)#
- An automatic method that detects short high-variance bursts and reconstructs them from clean data; common in real-time pipelines.Related: ArtifactUsed in: Artifact removal, three ways
- Biomedical equipment technician (BMET)#
- The hospital staff who maintain and repair medical devices and know every way they fail in real use.Used in: Electrical safety, with the numbers, The clinic
- BrainFlow#
- A library that talks to OpenBCI and most consumer EEG headsets with one API, in Python and other languages.Related: OpenBCI, Lab Streaming Layer (LSL)Used in: Project A: The ADS1299 board, Lab Streaming Layer and PsychoPy
- Butterworth filter#
- The IIR filter with the flattest possible passband, the default choice when you do not have a reason to choose something else.Related: IIR filterUsed in: Digital filters, and what they do to time
- Common spatial patterns (CSP)#
- Spatial filters that maximize the variance of one class while minimizing the other, the classic feature extractor for motor imagery.Related: Motor imagery, ClassifierUsed in: Spatial filters and montages, Project C: Motor imagery, Classification done honestly, Deep learning for EEG
- Current dipole#
- The simplest model of a patch of active cortex: a tiny current source and sink close together, with a strength and a direction.Related: Volume conduction, Forward problemUsed in: Project D: The phantom head, Spatial filters and montages, Where oscillations come from, The math track
- Delta-sigma ADC#
- A converter that samples very fast with a crude 1-bit decision and averages, achieving 24-bit resolution at modest rates; what the ADS1299 uses.Related: Analog-to-digital converter (ADC), ADS1299Used in: Project A: The ADS1299 board, Analog to digital
- Dry electrode#
- An electrode used without gel, faster to apply but with higher, less stable impedance and more motion noise; the consumer-headset compromise.Related: Electrode impedance, Active electrodeUsed in: Project C: Make your own electrodes, Project E: The eight-channel EMG armband, The electrode-electrolyte interface, Noise, and where the microvolt goes
- Electrical isolation#
- A barrier, such as a battery, an optocoupler, or an isolated USB link, with no conductive path from the mains to the patient.Related: Leakage currentUsed in: Electrical safety, with the numbers
- ESP32#
- A cheap microcontroller with Wi-Fi and Bluetooth, good for streaming data from a sensor board to a laptop.Related: RP2040, FirmwareUsed in: Project A: The ADS1299 board, Electrical safety, with the numbers
- FIR filter#
- A filter that outputs a weighted sum of recent inputs only; symmetric ones delay every frequency equally, which preserves waveform shape.Related: IIR filter, Phase distortionUsed in: Project F: The software pipeline, Digital filters, and what they do to time
- Group delay#
- How long a filter delays the envelope of a signal at each frequency; constant for symmetric FIR filters.Related: Phase distortionUsed in: Digital filters, and what they do to time
- Half-cell potential#
- The DC voltage that appears wherever a metal meets an electrolyte; two electrodes with different half-cell potentials produce an offset that can swamp the signal.Related: Silver/silver-chloride (Ag/AgCl), ElectrodeUsed in: Project C: Make your own electrodes, The electrode-electrolyte interface
- High-pass filter#
- Removes slow drifts and DC offsets, passing everything above a cutoff; set it too high and you erase slow brain activity.Related: FilterUsed in: Project F: The software pipeline
- IIR filter#
- A filter that feeds its output back into itself; cheap and sharp, but it delays different frequencies by different amounts.Related: FIR filter, Butterworth filter, Phase distortionUsed in: Digital filters, and what they do to time
- KiCad#
- Free, open-source software for drawing schematics and laying out printed circuit boards.Related: Printed circuit board (PCB), JLCPCBUsed in: KiCad to JLCPCB
- Lab Streaming Layer (LSL)#
- A protocol that timestamps and synchronizes streams from many devices and stimulus programs on a network; every BCI lab uses it.Related: PsychoPy, BrainFlowUsed in: Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Lab Streaming Layer and PsychoPy, Project A: The SSVEP speller, Project B: The P300 speller, Project C: Motor imagery, Project D: Your own ERPs, Designing an experiment that can fail
- Linear discriminant analysis (LDA)#
- A simple, robust classifier that draws a straight boundary between classes; the workhorse of EEG decoding for two decades.Related: ClassifierUsed in: Project E: The eight-channel EMG armband, Project B: The P300 speller, Project C: Motor imagery, Classification done honestly, Deep learning for EEG, Project E: Large-scale neural data, Project C: Silent speech
- Low-pass filter#
- Removes fast content above a cutoff; the anti-alias filter is one, and so is the smoothing you apply to a noisy trace.Related: Filter, Anti-alias filterUsed in: Analog to digital
- P300#
- A positive bump about 300 milliseconds after a rare or meaningful stimulus, the basis of the classic BCI speller.Related: Event-related potential (ERP), Oddball paradigmUsed in: Project B: Reverse-engineer a bought board, Digital filters, and what they do to time, Lab Streaming Layer and PsychoPy, Project B: The P300 speller, Project D: Your own ERPs, Event-related potentials and oscillations, Classification done honestly, Deep learning for EEG, Designing an experiment that can fail, Privacy, neuroethics, and the abandoned implant
- Phase distortion#
- Different frequencies being delayed by different amounts, which smears and shifts waveforms such as an ERP peak.Related: Zero-phase filtering, Group delayUsed in: Digital filters, and what they do to time
- PsychoPy#
- Free Python software for presenting stimuli with precise timing and logging responses.Related: Lab Streaming Layer (LSL)Used in: Lab Streaming Layer and PsychoPy, Project A: The SSVEP speller, Project B: The P300 speller, Project C: Motor imagery, Project D: Your own ERPs, Designing an experiment that can fail
- Quantization#
- Rounding each sample to one of a finite set of levels; a 10-bit ADC has 1024 of them, a 24-bit one about 16 million.Related: Analog-to-digital converter (ADC), Resolution (bits)Used in: Analog to digital
- Signal-to-noise ratio (SNR)#
- How big the thing you want is compared to everything else, often in decibels; the number every design decision is trying to raise.Related: Trial averaging, Noise floorUsed in: Noise, and where the microvolt goes, Project E: Large-scale neural data, Project E: Intracortical data
- Silent speech interface#
- Decoding intended words from muscle activity of the face and throat while the person mouths or subvocalizes without sound.Related: Electromyography (EMG)Used in: Project E: The eight-channel EMG armband, Project C: Silent speech
- Silver/silver-chloride (Ag/AgCl)#
- The standard electrode material for biopotentials, because its chemistry produces a stable, low-noise, low-drift interface with salty gel.Related: Electrode, Half-cell potentialUsed in: Project C: Make your own electrodes, Project D: The phantom head, The electrode-electrolyte interface, Project C: Other windows on the brain, Project D: Stimulation, studied properly, Stimulation safety and electrochemistry
- SPI#
- A simple four-wire bus for talking to chips like the ADS1299 from a microcontroller.Related: FirmwareUsed in: Project A: The ADS1299 board, Project B: Reverse-engineer a bought board, Firmware and SPI for a biopotential chip, Project B: Build a lab a rig
- Steady-state visual evoked potential (SSVEP)#
- When you stare at something flickering at a fixed rate, visual cortex oscillates at that rate, and the frequency can be read straight out of the EEG spectrum.Related: Canonical correlation analysis (CCA), Brain-computer interface (BCI)Used in: Project B: Reverse-engineer a bought board, Lab Streaming Layer and PsychoPy, Project A: The SSVEP speller, Project B: The P300 speller, Project H: Play, Event-related potentials and oscillations, Classification done honestly, Deep learning for EEG, A reading order for Kandel, Project A: A speller for someone with ALS, Project B: Control for a real user, Project C: Other windows on the brain
- Surface Laplacian#
- Subtracting the average of neighbouring electrodes from each electrode, which sharpens spatial detail and suppresses distant sources.Related: Montage, Volume conductionUsed in: Spatial filters and montages
- Thermal (Johnson) noise#
- The random voltage every resistor produces just by being warm; about 13 nanovolts per root hertz for 10 kilohms, which matters at microvolt levels.Related: Noise floorUsed in: Noise, and where the microvolt goes
- Volume conduction#
- The spreading of currents from a source through the conductive tissues of the head, which is why one source shows up on many electrodes and why scalp EEG is blurry.Related: Current dipole, Forward problem, Surface LaplacianUsed in: Project D: The phantom head, Event-related potentials and oscillations, Project B: The forward and inverse problems
- Zero-phase filtering#
- Running a filter forward and then backward through recorded data so the delays cancel; impossible in real time.Related: Phase distortionUsed in: Digital filters, and what they do to time
Phase 3: Read intent
- Baseline correction#
- Subtracting the mean of the pre-stimulus interval from each epoch so trials line up.Related: EpochUsed in: Project D: Your own ERPs
- Bayesian estimation#
- Updating a belief about a hidden quantity by combining what you expected with what you measured, in proportion to their reliabilities.Related: Kalman filterUsed in: Bayesian decoding and the Kalman filter, The math track
- BCI Competition IV#
- A 2008 set of benchmark EEG datasets, especially dataset 2a (four-class motor imagery), still the most-cited test bed.Related: Motor imagery, MOABBUsed in: Project F: Reproduce a paper, Deep learning for EEG
- Braindecode#
- A PyTorch library of deep-learning models for EEG, including EEGNet and ShallowConvNet, with dataset loaders.Related: EEGNet, MOABBUsed in: Project C: Motor imagery, Project E: The sleep band, Project F: Reproduce a paper, Deep learning for EEG
- Canonical correlation analysis (CCA)#
- Finding the combination of channels that correlates best with reference sinusoids at each candidate frequency; the standard SSVEP decoder.Related: Steady-state visual evoked potential (SSVEP)Used in: Project A: The SSVEP speller
- Closed loop#
- A system in which the output feeds back to influence the input, such as a user adjusting their brain activity while watching the cursor it controls.Related: Brain-computer interface (BCI), NeurofeedbackUsed in: Bayesian decoding and the Kalman filter, Project E: Haptic feedback for a prosthetic hand, Neuromodulation, mechanisms and evidence
- DANDI archive#
- A public repository of neurophysiology data in NWB format, including human intracortical BCI recordings from published papers.Related: Neurodata Without Borders (NWB)Used in: Project F: Reproduce a paper, Project E: Intracortical data
- Data leakage#
- Information from the test set sneaking into training, for example by shuffling trials from one session across both; it inflates accuracy and is the most common flaw in student BCI papers.Related: Cross-validationUsed in: Classification done honestly
- EEGNet#
- A compact convolutional neural network for EEG, published in 2018, that became the standard deep-learning baseline because it is small enough to train on a laptop.Related: BraindecodeUsed in: Project C: Motor imagery, Project F: Reproduce a paper, Deep learning for EEG
- Error-related negativity (ERN)#
- A sharp negative wave about 50 to 100 milliseconds after you make a mistake, before you consciously notice.Related: Event-related potential (ERP)Used in: Project D: Your own ERPs, Event-related potentials and oscillations
- Hypnogram#
- The staircase plot of sleep stage against time across a night, scored in 30-second epochs from EEG, EOG, and EMG.Related: Polysomnography (PSG), Sleep-EDFUsed in: Project E: The sleep band, Sleep neuroscience, Project C: Other windows on the brain
- Kalman filter#
- An estimator that blends a prediction from a motion model with a noisy measurement, weighting each by how much it trusts it; the decoder behind most intracortical cursor control.Related: Decoder, Bayesian estimationUsed in: Project F: Reproduce a paper, Bayesian decoding and the Kalman filter, Project E: Intracortical data, The math track
- MOABB#
- Mother of All BCI Benchmarks: a library that runs classifiers across many public EEG datasets the same way so results are comparable.Related: Braindecode, BCI Competition IVUsed in: Project F: Reproduce a paper, Classification done honestly, Deep learning for EEG
- N170#
- A negative deflection about 170 milliseconds after seeing a face, over the back of the head on the right.Related: Event-related potential (ERP)Used in: Project D: Your own ERPs, Event-related potentials and oscillations, Designing an experiment that can fail
- Neurofeedback#
- Showing someone a measure of their own brain activity in real time so they can try to change it; its clinical claims range from plausible to unsupported.Related: Closed loopUsed in: Project G: Neurofeedback, and how to test whether it does anything, Statistics for small experiments, Synapses and plasticity
- Oddball paradigm#
- An experiment in which a rare target is mixed in among frequent standards; the rare one produces a P300.Related: P300Used in: Project D: Your own ERPs
- Overfitting#
- A model that has memorized its training data, including its noise, and fails on anything new.Related: Cross-validationUsed in: Deep learning for EEG
- Population vector#
- A way to read out intended movement direction by adding up the preferred directions of many motor cortex neurons, weighted by how fast each is firing.Related: Decoder, Rate codeUsed in: Bayesian decoding and the Kalman filter, Project E: Large-scale neural data, Neural coding, Motor cortex and the systems BCIs target
- Riemannian geometry classifiers#
- Methods that treat each trial's covariance matrix as a point on a curved space and classify by distance there; the strongest simple baseline for EEG today.Related: Classifier, Common spatial patterns (CSP)Used in: Project C: Motor imagery, Project F: Reproduce a paper, Classification done honestly
- Seizure detection#
- Automatically flagging seizures in EEG, a real clinical need and a good first machine-learning project on the TUH corpus.Related: TUH EEG Corpus, EpilepsyUsed in: Deep learning for EEG
- Sleep spindle#
- A half-second burst of 11 to 16 Hz activity in light sleep, generated in the thalamus.Related: Hypnogram, Thalamocortical loopUsed in: Project E: The sleep band, Neuron biophysics, Where oscillations come from, Sleep neuroscience
- Sleep-EDF#
- A public PhysioNet dataset of whole-night sleep recordings with expert stage labels, the standard training set for automatic sleep staging.Related: Hypnogram, PhysioNetUsed in: Project E: The sleep band, Sleep neuroscience
- Slow wave#
- A large, slow (under 1 Hz) oscillation of deep sleep during which most cortical neurons fall silent and then fire together.Related: Delta waves, HypnogramUsed in: Project E: The sleep band, Event-related potentials and oscillations, Where oscillations come from, Sleep neuroscience
- Transfer learning#
- Using data from other people or other days to shorten calibration for a new user.Related: CalibrationUsed in: Classification done honestly
- TUH EEG Corpus#
- Temple University Hospital's archive of tens of thousands of clinical EEGs, with a labelled seizure subset.Related: Seizure detectionUsed in: Project F: Reproduce a paper, Deep learning for EEG
- Validation#
- Testing that the device meets the user's need in real use; did you specify the right thing.Related: VerificationUsed in: Deep learning for EEG, Design controls and the design history file, Human factors and use error, Regulatory pathways
- Wavelet transform#
- A time-frequency analysis that uses short wiggles at high frequencies and long ones at low, matching the trade-off to the frequency.Related: SpectrogramUsed in: Event-related potentials and oscillations
Phase 4: The brain behind the signal
- Brian2#
- A Python simulator for spiking neural networks in which you write the equations directly.Related: Leaky integrate-and-fire neuronUsed in: Project A: Neurons from the equations up, Where oscillations come from
- Forward problem#
- Given a source in the head, compute the voltages on the scalp; it has a unique answer.Related: Inverse problem, Current dipoleUsed in: Project B: The forward and inverse problems
- Inverse problem#
- Given voltages on the scalp, find the sources; it has infinitely many answers, so every solution leans on assumptions.Related: Forward problem, Source localizationUsed in: Project B: The forward and inverse problems
- Izhikevich model#
- A two-equation neuron model that reproduces most known firing patterns at a tiny fraction of Hodgkin–Huxley's cost.Related: Leaky integrate-and-fire neuronUsed in: Project A: Neurons from the equations up, Neuron biophysics
- Leaky integrate-and-fire neuron#
- The simplest useful neuron model: a leaky capacitor that fires and resets when it reaches a threshold.Related: Hodgkin–Huxley model, Izhikevich modelUsed in: Project A: Neurons from the equations up
- Local field potential (LFP)#
- The slow, summed electrical activity of many neurons near an electrode, below a few hundred hertz, as opposed to individual spikes.Related: Electroencephalography (EEG), Extracellular recordingUsed in: Project D: Neuromatch Academy, Project E: Large-scale neural data, Where oscillations come from, Project E: Intracortical data, Electrode materials and the foreign body response
- Neuroethics#
- The study of the ethical questions raised by measuring and changing brains, from privacy of neural data to who owns an implant.Related: Neurorights, Neural data privacyUsed in: Neural coding, Privacy, neuroethics, and the abandoned implant
- Neuromatch Academy#
- An online, three-week summer school in computational neuroscience with a large alumni network, on a country-scaled tuition.Used in: Project D: Neuromatch Academy, Community, The calendar
- Neuropixels#
- Silicon probes with hundreds to thousands of recording sites along a thin shank, the current standard for large-scale animal recording.Related: Spike sortingUsed in: Project E: Large-scale neural data, Electrode materials and the foreign body response, The industry landscape
- Neurotransmitter#
- A chemical released at a synapse that changes the next neuron's membrane potential.Related: SynapseUsed in: Synapses and plasticity
- Phosphene#
- A spot of light perceived when the visual system is stimulated electrically rather than by light.Related: Retinal prosthesisUsed in: Project C: Prosthesis simulators, Sensory systems as prosthesis targets, Neuromodulation, mechanisms and evidence, The Utah Array, a history
- Rate code#
- The idea that a neuron carries information in how many spikes per second it fires.Related: Temporal code, Population vectorUsed in: Neural coding, Sensory systems as prosthesis targets
- Retinal prosthesis#
- An electrode array on or under the retina that produces spots of light, phosphenes, in people blinded by retinal disease.Related: PhospheneUsed in: Project C: Prosthesis simulators, Sensory systems as prosthesis targets, The industry landscape
- Sensory substitution#
- Delivering information from one sense through another, such as an image as a soundscape, and letting the brain learn to use it.Used in: Synapses and plasticity, Project D: Sensory substitution
- Spike-timing-dependent plasticity (STDP)#
- A rule in which a synapse strengthens if the input spike arrives just before the output spike and weakens if it arrives just after.Related: Synaptic plasticityUsed in: Synapses and plasticity
- Stroke#
- Loss of blood supply to part of the brain; rehabilitation after it is the largest clinical use of neurotechnology today.Used in: Synapses and plasticity, Motor cortex and the systems BCIs target, Neuromodulation, mechanisms and evidence, The industry landscape, The clinic
- Synapse#
- The junction where one neuron signals to another, usually by releasing a chemical that opens channels on the next cell.Related: Synaptic plasticity, NeurotransmitterUsed in: Project A: Neurons from the equations up, Neuron biophysics, Synapses and plasticity, Where oscillations come from, Motor cortex and the systems BCIs target, Sleep neuroscience, A reading order for Kandel, Neuromodulation, mechanisms and evidence
- Synaptic plasticity#
- The strengthening or weakening of synapses with use, the physical basis of learning.Related: Spike-timing-dependent plasticity (STDP), SynapseUsed in: Synapses and plasticity
- Temporal code#
- The idea that the precise timing of spikes, not just their count, carries information.Related: Rate codeUsed in: Neural coding
- Theta rhythm#
- Roughly 4 to 8 Hz activity, prominent in drowsiness, some memory tasks, and strongly in the rodent hippocampus.Related: Alpha rhythmUsed in: Neural coding
Phase 5: Build for a person
- 510(k)#
- The FDA pathway that clears a device by showing it is substantially equivalent to one already on the market, its predicate.Related: Predicate device, FDA device classesUsed in: Regulatory pathways
- Breakthrough Device designation#
- An FDA program that speeds review for devices treating serious conditions with no good alternative; most invasive BCI companies have it.Used in: Regulatory pathways, The Utah Array, a history
- De Novo#
- The FDA pathway for a low-to-moderate-risk device that has no predicate.Related: 510(k)Used in: Regulatory pathways
- IEC 62366#
- The standard for usability engineering of medical devices, aimed at preventing use errors.Related: Human factors, Use errorUsed in: Risk management and the FMEA, Human factors and use error
- Investigational device exemption (IDE)#
- FDA permission to use an unapproved device in a clinical study.Used in: Regulatory pathways, The Utah Array, a history, The industry landscape
- ISO 13485#
- The quality management standard for medical device companies; the reason everything is written down.Used in: Design controls and the design history file
- Need statement#
- One sentence, "a way to [do X] for [whom] that [outcome]," with no solution in it.Related: Needs findingUsed in: Project A: A speller for someone with ALS, Project F: An adaptive game controller, Competitions: Bench to Bedside, DEBUT, and RESNA, Needs finding, Design controls and the design history file
- Neural data privacy#
- The question of who may collect, store, and sell recordings of brain activity, now addressed by laws in several US states.Related: NeuroethicsUsed in: Privacy, neuroethics, and the abandoned implant
- Neurorights#
- Proposed rights to mental privacy, identity, and agency, already written into Chile's constitution and debated elsewhere.Related: NeuroethicsUsed in: Privacy, neuroethics, and the abandoned implant
- Predicate device#
- The already-legal device a 510(k) submission compares itself to.Related: 510(k)Used in: Regulatory pathways
- Premarket approval (PMA)#
- The strictest FDA pathway, requiring clinical evidence of safety and effectiveness; the route for Class III implants.Related: FDA device classesUsed in: Regulatory pathways, The Utah Array, a history, The industry landscape
- Requirements#
- Testable statements of what the device must do, such as "input-referred noise below 1 microvolt RMS from 0.5 to 40 Hz."Related: VerificationUsed in: Project A: A speller for someone with ALS, Project E: Haptic feedback for a prosthetic hand, Project F: An adaptive game controller, Design controls and the design history file, Privacy, neuroethics, and the abandoned implant, Project B: Build a lab a rig, Packaging, power, and telemetry
- RP2040#
- The Raspberry Pi microcontroller, cheap and well documented, with fast SPI and USB.Related: ESP32Used in: Project F: An adaptive game controller, Project B: Build a lab a rig
- Use error#
- A mistake made while using a device that the design invited, as opposed to random user carelessness.Related: Human factorsUsed in: Risk management and the FMEA, Human factors and use error
Phase 6: Beyond the scalp
- Augmentative and alternative communication (AAC)#
- Any tool that helps someone who cannot speak communicate, from picture boards to eye-gaze spellers to BCIs.Related: Amyotrophic lateral sclerosis (ALS)Used in: The industry landscape
- Charge balance#
- Every stimulation pulse followed by an equal and opposite one, so no net charge accumulates and no electrochemistry runs away.Related: Charge densityUsed in: Project D: Stimulation, studied properly, Stimulation safety and electrochemistry
- Charge density#
- The charge injected per pulse divided by electrode area; keeping it below safe limits prevents electrode damage and tissue injury.Related: Shannon safety limit, Charge balanceUsed in: Project A: Electrode fabrication, Project D: Stimulation, studied properly, Stimulation safety and electrochemistry
- Electrochemical impedance spectroscopy (EIS)#
- Measuring an electrode's impedance across many frequencies to characterize its interface, standard in electrode development.Related: Electrode impedance, PEDOT:PSSUsed in: Project A: Electrode fabrication, Project D: Stimulation, studied properly, Electrode materials and the foreign body response
- Neurodata Without Borders (NWB)#
- The standard file format for neurophysiology data, so recordings from different labs can be read the same way.Related: DANDI archiveUsed in: Project E: Intracortical data
- PEDOT:PSS#
- A conductive polymer coating that drops a microelectrode's impedance by an order of magnitude and improves charge injection for stimulation.Related: Electrochemical impedance spectroscopy (EIS), Charge densityUsed in: Project A: Electrode fabrication, Stimulation safety and electrochemistry, Electrode materials and the foreign body response
- Shannon safety limit#
- An empirical curve relating charge per phase and charge density below which stimulation did not damage tissue in animal studies.Related: Charge densityUsed in: Stimulation safety and electrochemistry
Threads
- Epilepsy monitoring unit (EMU)#
- The hospital ward where patients are recorded on video-EEG for days to capture seizures before surgery.Related: EpilepsyUsed in: The clinic
- HIPAA#
- The US law governing health information privacy; it shapes how clinical data can be stored and shared.Used in: The clinic
- Polysomnography (PSG)#
- The overnight sleep study: EEG, EOG, chin EMG, breathing, oxygen, and heart rate recorded together and scored in 30-second epochs.Related: Hypnogram, Sleep technologistUsed in: The clinic
Other terms
- BCI inefficiency#
- The finding that around one in five people cannot control a motor-imagery BCI even after training; once called "BCI illiteracy," a label the field is retiring.Related: Motor imagery
- Beta rhythm#
- Roughly 13 to 30 Hz activity, prominent over motor areas and associated with active concentration or holding still.Related: Alpha rhythm, Sensorimotor rhythm (mu)
- Blink artifact#
- The large, slow deflection an eye blink produces in frontal channels, often ten times the size of the brain signal.Related: Electrooculography (EOG), Artifact
- Common average reference#
- Re-referencing every channel to the mean of all channels, which approximates a neutral reference when electrodes cover the head well.Related: Montage
- Conductive gel or paste#
- Salty gel between electrode and skin that lowers impedance, stabilizes the chemistry, and is the reason research EEG takes half an hour to set up.Related: Electrode impedance, Dry electrode
- Design for manufacturability (DFM)#
- Designing so the thing can actually be made reliably and cheaply, not just once on your bench.
- ECoG grid#
- A flexible sheet of electrodes laid on the brain's surface, used clinically for epilepsy and in research for speech decoding.Related: Electrocorticography (ECoG)
- Gamma activity#
- Fast activity above about 30 Hz, tied to local processing; weak and easily confused with muscle noise on the scalp.Related: Electromyography (EMG), Artifact
- Informed consent#
- A participant's agreement to take part after understanding the risks, in language they can follow; a process, not a signature.Related: Institutional review board (IRB)
- Intraoperative neuromonitoring (IONM)#
- Recording evoked potentials and EMG during surgery to warn the surgeon before a nerve or the spinal cord is damaged.Related: Event-related potential (ERP)
- Jig or fixture#
- A purpose-built holder or tool that makes a repeated lab task faster or more repeatable; labs are full of ones nobody has built yet.
- Locked-in syndrome#
- Full awareness with almost no voluntary movement, sometimes only the eyes; communication depends entirely on assistive technology.Related: Amyotrophic lateral sclerosis (ALS)
- Receptive field#
- The set of stimuli, such as a patch of the visual field or a spot on the skin, that makes a particular neuron respond.
- Resolution (bits)#
- How finely an ADC divides its range; each extra bit halves the step size.Related: Quantization, Delta-sigma ADC
- Shielding#
- Surrounding a signal wire with a grounded conductor so electric fields from the room couple into the shield instead of the signal.Related: Common-mode signal
- Spinal cord injury (SCI)#
- Damage to the cord that cuts communication between brain and body below the injury; the population for motor BCIs and spinal stimulation.
- Support vector machine (SVM)#
- A classifier that finds the boundary with the widest margin between classes; strong on small datasets.Related: Classifier
- Trial averaging#
- Adding many time-locked epochs together so the consistent response grows and the random noise shrinks by the square root of the number of trials.Related: Event-related potential (ERP), Signal-to-noise ratio (SNR)
- Utah Slanted Electrode Array (USEA)#
- A version of the Utah Array with needles of graded lengths, designed to reach different depths of a peripheral nerve.Related: Utah Array
- Vocoder (noise vocoder)#
- A simulation of what a cochlear implant delivers: the loudness envelope of each frequency band, applied to noise in that band.Related: Cochlear implant
#
A
B
C
D
E
- ECoG grid
- EEG technologist
- EEGNet
- Electrical isolation
- Electrocardiography (ECG)
- Electrochemical impedance spectroscopy (EIS)
- Electrocorticography (ECoG)
- Electrode
- Electrode impedance
- Electroencephalography (EEG)
- Electromyography (EMG)
- Electrooculography (EOG)
- Epilepsy
- Epilepsy monitoring unit (EMU)
- Epoch
- Error-related negativity (ERN)
- ESP32
- Event-related desynchronization (ERD)
- Event-related potential (ERP)
- Extracellular recording
F
G
H
I
- IEC 60601
- IEC 62366
- IIR filter
- Independent component analysis (ICA)
- Informed consent
- Input bias current
- Input impedance
- Institutional review board (IRB)
- Instrumentation amplifier
- Intracortical recording
- Intraoperative neuromonitoring (IONM)
- Inverse problem
- Investigational device exemption (IDE)
- Ion channel
- ISO 13485
- ISO 14971
- Izhikevich model
J
K
L
M
N
O
P
Q
R
S
- Sampling rate
- Seizure detection
- Sensorimotor rhythm (mu)
- Sensory substitution
- Shannon safety limit
- Shielding
- Signal-to-noise ratio (SNR)
- Silent speech interface
- Silver/silver-chloride (Ag/AgCl)
- Sleep spindle
- Sleep technologist
- Sleep-EDF
- Slow wave
- Source localization
- Spectrogram
- Spectrum
- SPI
- Spike
- Spike sorting
- Spike-timing-dependent plasticity (STDP)
- SpikerBox
- Spinal cord injury (SCI)
- Spinal cord stimulation (SCS)
- Steady-state visual evoked potential (SSVEP)
- Stentrode
- Stroke
- Support vector machine (SVM)
- Surface Laplacian
- Synapse
- Synaptic plasticity