Compendium

Neuroscience Foundations — Biology Reference

29 min read

Neuroscience Foundations — Biology Reference

Neuroscience studies the nervous system across scales: ions and channels (µm, ms), single neurons and synapses (10 µm, 1 ms), microcircuits and columns (100 µm, 10 ms), areas and pathways (cm, 100 ms), whole brain and behavior (m, s–years). This note is the Tier 1 reference: cellular substrates, circuit anatomy, systems-level computation, plasticity and development, disease, recording technology, and the 2024–26 state of the art. Cross-link: cell-molecular-biology, bioinstrumentation, biochemistry-foundations, biomechanics, probability-fundamentals, transformer-architecture.

1. The neuron — structural and electrical unit

1.1 Anatomy

A typical mammalian neuron has four functional compartments:

  • Dendrites: branched receptive arbors. Receive thousands of synaptic inputs; passive and active integration via dendritic spikes (Na⁺, NMDA, Ca²⁺ plateau). Cortical pyramidal cells have apical (toward pia) and basal dendrites.
  • Soma (cell body): 10–30 µm. Contains nucleus, ER, Golgi, mitochondria. Integrates synaptic currents arriving at the axon hillock (high Na⁺ channel density, lowest threshold).
  • Axon: single output process, 1 µm–1 m. Conducts action potentials. Often myelinated (oligodendrocytes in CNS, Schwann cells in PNS).
  • Synapse (axon terminal / bouton): presynaptic vesicle pool, active zone, postsynaptic density. Chemical (most) or electrical (gap-junction, connexin36).

Glia outnumber neurons in many regions: astrocytes (K⁺ buffering, glutamate uptake via EAAT, tripartite synapse), oligodendrocytes (myelin), microglia (immune surveillance, synaptic pruning via C1q/C3), ependymal (CSF). Human brain ~86 billion neurons (Herculano-Houzel 2009), roughly equal glia.

1.2 Membrane biophysics

The neuronal membrane is a phospholipid bilayer (~5 nm, capacitance C_m ≈ 1 µF/cm²) studded with channels and pumps.

Nernst equation — equilibrium potential for ion X: E_X = (RT/zF) · ln([X]_out / [X]_in)

At 37 °C (T = 310 K), RT/F ≈ 26.7 mV. For typical mammalian neuron concentrations:

  • E_K ≈ −90 mV ([K⁺]_in 140 mM, [K⁺]_out 5 mM)
  • E_Na ≈ +60 mV ([Na⁺]_in 15 mM, [Na⁺]_out 145 mM)
  • E_Cl ≈ −65 mV
  • E_Ca ≈ +120 mV (sub-µM intracellular)

Goldman-Hodgkin-Katz (GHK) equation — resting potential weighted by permeabilities: V_m = (RT/F) · ln[(P_K[K⁺]_o + P_Na[Na⁺]_o + P_Cl[Cl⁻]_i) / (P_K[K⁺]_i + P_Na[Na⁺]_i + P_Cl[Cl⁻]_o)]

At rest P_K ≫ P_Na, so V_m ≈ −70 mV (close to E_K but depolarized by Na⁺ leak). The Na⁺/K⁺-ATPase (3 Na⁺ out / 2 K⁺ in per ATP, electrogenic, contributes ~−10 mV) maintains gradients; accounts for ~20% of basal metabolic rate brain-wide.

1.3 Action potential — Hodgkin & Huxley

Alan Hodgkin and Andrew Huxley (1952, J. Physiol., six-paper series on the squid giant axon) reconstructed the action potential using voltage-clamp and a kinetic model of Na⁺ and K⁺ conductances. Nobel Prize in Physiology or Medicine 1963 (with John Eccles).

The HH equations:

  • I_m = C_m · dV/dt + g̅_Na m³h (V − E_Na) + g̅_K n⁴ (V − E_K) + g̅_L (V − E_L)
  • Gating variables m, h, n follow first-order kinetics dx/dt = α_x(V)(1 − x) − β_x(V) x.

Phases of the AP (~1–2 ms duration, ~100 mV amplitude):

  1. Depolarization to threshold (~−55 mV) — synaptic current or pacemaker.
  2. Rising phase — voltage-gated Na⁺ channels (Na_V1.1–1.9, α-subunit with four homologous domains, S4 voltage sensors) open; m³ activation, h inactivation lags.
  3. Peak ~+30–40 mV.
  4. Repolarization — Na⁺ inactivation + delayed-rectifier K⁺ (K_V) opening.
  5. Hyperpolarization (afterhyperpolarization, AHP) — K⁺ overshoot toward E_K; Ca²⁺-activated K⁺ (SK, BK channels) prolongs.
  6. Refractory period — absolute (~1 ms, Na⁺ inactivation) and relative (~3–5 ms).

Propagation: longitudinal current depolarizes adjacent membrane. In myelinated axons, the AP jumps from node of Ranvier to node (saltatory conduction, ~120 m/s in human Aα fibers vs ~1 m/s unmyelinated). Internodal distance ~1 mm, ~100× axon diameter.

1.4 Ion-channel families

  • Voltage-gated Na⁺ (Na_V): APs; tetrodotoxin (TTX) blocks; saxitoxin; local anesthetics (lidocaine).
  • Voltage-gated K⁺ (K_V): K_V1–12, including delayed-rectifier, A-type (K_V4, transient), M-current (K_V7), inward-rectifier (Kir).
  • Voltage-gated Ca²⁺ (Ca_V): L-type (Ca_V1, dihydropyridine-sensitive, muscle + dendrites), N/P/Q-type (Ca_V2, presynaptic transmitter release), T-type (Ca_V3, thalamic burst firing).
  • Ligand-gated: ionotropic glutamate (AMPA, NMDA, kainate), GABA-A, glycine, nicotinic ACh, 5-HT3, P2X.
  • Leak (background): K2P / TASK / TREK two-pore-domain K⁺ — set resting potential.
  • HCN (h-current): hyperpolarization-activated cyclic nucleotide-gated; pacemaker in cortex, thalamus, sinoatrial node.
  • TRP: temperature, mechanical, chemical sensors (TRPV1 capsaicin/heat — Julius Nobel 2021; TRPM8 cold/menthol; PIEZO1/2 mechanical — Patapoutian Nobel 2021).

2. Synaptic transmission

2.1 Presynaptic release

Arrival of AP → opening of P/Q-type Ca²⁺ channels at active zone → local Ca²⁺ rises 10–100 µM in microdomains → triggers fusion of synaptic vesicles with plasma membrane (exocytosis) within ~100 µs.

SNARE complex mediates vesicle fusion:

  • v-SNARE on vesicle: synaptobrevin/VAMP.
  • t-SNAREs on plasma membrane: syntaxin-1 and SNAP-25.
  • Coiled-coil zippering pulls membranes together.
  • Synaptotagmin-1 is the Ca²⁺ sensor (C2A/C2B domains).
  • Munc18, Munc13, RIM, complexin coordinate priming and clamping.

James Rothman, Randy Schekman, Thomas Südhof — Nobel Prize in Physiology or Medicine 2013 for discoveries of machinery regulating vesicle traffic, a major transport system in our cells. Tetanus and botulinum toxins cleave SNAREs (BoNT/A on SNAP-25, BoNT/B on VAMP, TeNT on VAMP) — basis for therapeutic Botox.

Quantal release (Katz, Nobel 1970): vesicles release in discrete packets (~5,000 glutamate molecules), producing miniature EPSCs (mEPSCs) of ~20 pA.

2.2 Postsynaptic response

Neurotransmitter binds receptors in postsynaptic density (PSD-95 scaffold at glutamatergic; gephyrin at GABAergic/glycinergic).

  • Ionotropic receptors: ligand-gated ion channels, fast (<1 ms latency, ms decay). E.g., AMPA, NMDA, GABA-A, nicotinic.
  • Metabotropic receptors: GPCRs (7TM), slow (10 ms–s), via G_s/G_i/G_q → cAMP, IP3/DAG, β-arrestin. E.g., mGluR1–8, GABA-B, dopamine D1–D5, muscarinic ACh, all monoamine.

Excitatory postsynaptic potential (EPSP) ~0.1–1 mV; many summed (temporal + spatial) drive soma past threshold. Inhibitory PSP (IPSP) usually Cl⁻ influx via GABA-A (more negative than threshold in mature neuron) or K⁺ efflux via GABA-B.

2.3 Termination and recycling

Removal of transmitter:

  • Reuptake by SLC transporters: SERT (serotonin — SSRI target), DAT (dopamine — cocaine/amphetamine), NET (norepinephrine), GAT (GABA), EAAT (glutamate, astrocytic glutamine cycle).
  • Enzymatic degradation: acetylcholinesterase (AChE — nerve agents/organophosphates inhibit; donepezil for Alzheimer’s), MAO-A/B (monoamines), COMT.
  • Diffusion out of cleft.

Vesicle recycling: clathrin-mediated endocytosis or kiss-and-run; refilled by vesicular transporters VGLUT1–3 (glutamate), VGAT (GABA/glycine), VMAT1/2 (monoamines), VAChT (ACh).

3. Neurotransmitters

TransmitterReceptorsFunctionDrugs
GlutamateAMPA (GluA1–4), NMDA (GluN1+2A–D, Mg²⁺ block, Ca²⁺-permeable), kainate, mGluR1–8Major excitatory; LTP; ~80% of cortical synapsesKetamine, memantine (NMDA antagonists)
GABAGABA-A (Cl⁻ ionotropic, α1–6β1–3γ1–3 subunit combinations), GABA-B (GPCR, K⁺/Ca²⁺)Major inhibitoryBenzodiazepines (α subunit), barbiturates, baclofen (GABA-B), Z-drugs
DopamineD1, D5 (G_s ↑cAMP); D2, D3, D4 (G_i)Reward, motor (basal ganglia), prefrontalL-DOPA, antipsychotics (D2 antagonists), amphetamines
Serotonin (5-HT)5-HT1–7 (most GPCR, 5-HT3 ionotropic)Mood, sleep, appetiteSSRIs (fluoxetine/Prozac), tricyclics, MAOIs, triptans, psychedelics (5-HT2A agonists)
AcetylcholineNicotinic (ionotropic, α/β subunits), muscarinic M1–M5 (GPCR)NMJ, autonomic, attention, REM, memoryCurare, nicotine, atropine, donepezil, sarin
Norepinephrineα1, α2, β1–3 (GPCR)Arousal, sympathetic, attention; locus coeruleusβ-blockers, SNRIs, clonidine
HistamineH1–H4Wakefulness; tuberomammillary nucleusAntihistamines (sedation via H1)
GlycineGlyR (Cl⁻ ionotropic)Spinal inhibitionStrychnine blocks → tetanic spasms
NeuropeptidesMany GPCRsCo-released; slow modulation. Opioids (β-endorphin, enkephalin, dynorphin → µ/δ/κ); substance P, NPY, CRF, oxytocin, vasopressin, orexin/hypocretin (sleep), CCKOpioids morphine/fentanyl, suvorexant
EndocannabinoidsCB1 (presynaptic, brain), CB2 (peripheral immune)Retrograde messengers; 2-AG, anandamideTHC, CBD, rimonabant

4. Synaptic plasticity and learning

4.1 Hebbian rule

Donald Hebb (1949): “When an axon of cell A is near enough to excite cell B and repeatedly or persistently takes part in firing it, some growth process or metabolic change takes place in one or both cells such that A’s efficiency, as one of the cells firing B, is increased.” Compressed: “cells that fire together, wire together.”

4.2 Long-term potentiation (LTP) and depression (LTD)

Discovered in rabbit hippocampus by Tim Bliss and Terje Lømo (1973): brief high-frequency stimulation produces synaptic strengthening lasting hours to days. Now a cornerstone of memory’s cellular basis.

Mechanism at CA1 Schaffer-collateral synapse (canonical):

  1. Strong correlated activity depolarizes postsynaptic membrane.
  2. NMDA receptors detect coincidence: glutamate-bound AND postsynaptic depolarized (releases Mg²⁺ block) → Ca²⁺ influx.
  3. Ca²⁺ activates CaMKII (autophosphorylation makes it persistent), PKA, PKC.
  4. Early-phase LTP: AMPA-receptor phosphorylation + lateral insertion into PSD → larger EPSCs.
  5. Late-phase LTP (>3 h): CREB-dependent gene transcription, BDNF, structural changes (spine head enlargement, new spines).

LTD: low-frequency stimulation → moderate Ca²⁺ → PP1/calcineurin → AMPA dephosphorylation and endocytosis.

4.3 STDP

Spike-timing-dependent plasticity (Markram, Bi & Poo 1997–98): if presynaptic spike precedes postsynaptic by <20 ms → LTP; reverse order → LTD. Asymmetric exponential window; depends on dendritic backpropagating APs. Provides causal Hebbian rule.

4.4 Homeostatic plasticity

Synaptic scaling (Turrigiano): all synapses scaled to keep average firing rate stable over days. Intrinsic excitability also tuned. Prevents runaway excitation.

4.5 Structural plasticity

Adult neurogenesis: dentate gyrus (granule cells, ~700/day in adult human — Spalding 2013 ¹⁴C dating), subventricular zone → olfactory bulb (rodents). Spine turnover continual; learning produces new spines that persist.

5. Brain anatomy

5.1 Cerebral cortex

~2.5 mm thick, ~2,500 cm² unfolded in humans, ~16 billion neurons. Six-layered (neocortex):

  • L1: molecular, mostly dendrites, few neurons.
  • L2/3: pyramidal cells, cortico-cortical projections.
  • L4: granular, input from thalamus (especially V1, S1).
  • L5: large pyramidal (Betz cells in M1), subcortical output → spinal cord, striatum, brainstem.
  • L6: cortico-thalamic feedback.

Minicolumn (~30 µm, ~100 neurons spanning all layers, Mountcastle 1957): proposed canonical unit; cortical column ~500 µm with shared receptive field. Brodmann (1909) parcellated 52 cytoarchitectonic areas; modern multi-modal parcellation (Glasser HCP 2016) identifies 180 areas per hemisphere.

Functional subdivisions:

  • Frontal lobe: motor (M1, area 4), premotor + SMA (6), prefrontal (executive, 9, 10, 46), Broca’s (44/45 — language production).
  • Parietal: somatosensory (S1, 1–3), posterior parietal (visuo-motor, attention, 5/7).
  • Temporal: auditory (A1, 41/42), Wernicke’s (22 — comprehension), MT/MST (motion), fusiform face area (FFA), hippocampus medially.
  • Occipital: V1 (17), V2, V4 (color), V3, dorsal/ventral streams (“where”/“what” — Ungerleider & Mishkin 1982).

5.2 Subcortical

  • Basal ganglia: striatum (caudate + putamen; receives cortical glutamate, dopamine from SNc), globus pallidus (internal/external), subthalamic nucleus, substantia nigra (pars compacta SNc, pars reticulata SNr). Direct (D1, “go”) and indirect (D2, “no-go”) pathways; hyperdirect from cortex → STN. Action selection, motor gating, habit. Parkinson disease: SNc dopamine neuron loss.
  • Thalamus: relay for nearly all sensory (except olfaction) and motor information to cortex. Specific nuclei: LGN (vision), MGN (audition), VPL/VPM (somatosensation), VA/VL (motor). Non-specific: pulvinar (attention), mediodorsal (PFC), reticular nucleus (gates).
  • Hippocampus (CA1, CA3, dentate gyrus) + entorhinal cortex: episodic memory, spatial navigation. Edward Tolman (1948) proposed cognitive maps. John O’Keefe discovered place cells (1971); May-Britt + Edvard Moser discovered grid cells in MEC (2005). Nobel Prize 2014. Adjacent cell types: head-direction, border, speed, time cells.
  • Amygdala: emotional valence, fear conditioning (LeDoux), social processing. Lateral nucleus receives sensory; central nucleus drives autonomic via PAG.
  • Hypothalamus: ~4 g; homeostasis. Nuclei: SCN (circadian), PVN (CRH→ACTH, oxytocin, vasopressin), ARC (leptin sensing, appetite), VMH (satiety), LH (orexin/MCH, wakefulness/appetite). Controls pituitary (hypophyseal portal); HPA axis.
  • Cerebellum: ~50 billion granule cells (more than rest of brain); coordination, timing, motor learning, increasing recognition of cognitive roles. Cortex three layers (molecular, Purkinje, granular). Climbing fibers (inferior olive, 1:1, complex spikes) and mossy fibers (parallel fibers → Purkinje); deep nuclei (dentate, interposed, fastigial) output. Marr-Albus-Ito theory: LTD at parallel-fiber–Purkinje synapses underpins supervised motor learning.
  • Brainstem: midbrain (superior/inferior colliculi, PAG, VTA, SNc), pons (locus coeruleus, raphe, pontine reticular), medulla (NTS, DMV, vagal, respiratory/cardiac control). Cranial nerves III–XII.
  • Spinal cord: dorsal horn (sensory in), ventral horn (motor out), tracts (corticospinal, spinothalamic, dorsal column–medial lemniscus, etc.).

6. Sensory systems

6.1 Visual

  • Retina: photoreceptors (rods, ~120 M; cones — S/M/L, ~6 M peaked in fovea) → horizontal/amacrine → bipolar → retinal ganglion cells (RGCs). RGCs encode On/Off center-surround. ~1.2 M optic nerve axons.
  • Optic chiasm: nasal fibers cross → contralateral.
  • LGN (thalamus): magnocellular (motion, low spatial freq), parvocellular (color, high res), koniocellular.
  • V1: orientation columns, ocular dominance columns, retinotopy. David Hubel and Torsten Wiesel (1962+; Nobel 1981) — simple/complex cells, critical periods.
  • Higher areas: V2, V3, V4 (color, Zeki), MT/V5 (motion). Dorsal (“where/how”, parietal) vs ventral (“what”, temporal — IT, FFA, PPA).

6.2 Auditory

  • Cochlea: 3,500 inner hair cells (transduction via MET channels — TMC1/2 + tip links), 12,000 outer (electromotile via prestin). Tonotopic: base = high freq, apex = low.
  • VIII nerve → cochlear nucleus → superior olive (ITD/ILD localization) → inferior colliculus → MGN → A1. Preserves tonotopy. Belt + parabelt areas.

6.3 Somatosensory

  • Mechanoreceptors: Merkel (slow-adapting, edges), Meissner (fast, flutter), Pacinian (vibration), Ruffini (stretch). PIEZO2 essential (Patapoutian 2021).
  • Pain/temperature: free nerve endings, TRPV1 (heat/capsaicin), TRPM8 (cold), Aδ + C fibers.
  • Dorsal column–medial lemniscus (touch/proprioception): cross at medulla. Spinothalamic (pain/temp): cross at spinal level.
  • VPL/VPM thalamus → S1 (3a/3b/1/2). Penfield homunculus.

6.4 Olfactory

  • ~400 functional olfactory receptors (Buck & Axel, Nobel 2004) on bipolar OSNs in olfactory epithelium → olfactory bulb glomeruli (one receptor → one or few glomeruli) → piriform cortex (direct, no thalamic relay).

6.5 Gustatory and vestibular

  • Taste: sweet (T1R2/3), umami (T1R1/3), bitter (T2R), salt (ENaC), sour (OTOP1) → VII/IX/X → NTS → VPM → insula.
  • Vestibular: hair cells in semicircular canals (rotation), utricle/saccule (linear acceleration/gravity) → VIII → vestibular nuclei → cerebellum, oculomotor (VOR).

7. Motor system

  • Primary motor cortex (M1): Betz cells (giant L5 pyramidal) → corticospinal tract → α motor neurons in ventral horn → NMJ (ACh, nicotinic) → muscle.
  • Premotor + SMA: planning, sequencing. Mirror neurons (Rizzolatti, F5).
  • Cerebellum: error correction, internal models, timing.
  • Basal ganglia: action selection, habit, vigor (dopamine-dependent).
  • Spinal cord: central pattern generators (locomotion), reflexes (stretch — muscle spindles Ia, Golgi tendon Ib).
  • α-motor neuron + its muscle fibers = motor unit; fine control (eye, hand) has small units, gross (back) large.

8. Memory systems

Larry Squire and Endel Tulving taxonomy:

  • Declarative (explicit, conscious):
    • Episodic (events, time/place — Tulving 1972) — hippocampus, MTL.
    • Semantic (facts, concepts) — anterior temporal lobe (hub-and-spoke model, Patterson & Lambon Ralph).
  • Non-declarative (implicit):
    • Procedural (skills) — basal ganglia, cerebellum.
    • Priming — neocortex.
    • Classical conditioning — cerebellum (motor), amygdala (emotional).
    • Habituation/sensitization — reflex pathways (Aplysia, Kandel Nobel 2000).

Working memory (Baddeley & Hitch 1974): central executive + phonological loop + visuospatial sketchpad + episodic buffer. PFC, parietal sustain neural activity during delay.

Consolidation: hippocampus replays recent experience during sharp-wave ripples (sleep, quiet wake) → gradually transferred to neocortex (systems consolidation, Marr-McClelland complementary learning systems). Synaptic consolidation requires protein synthesis (within hours; cycloheximide blocks). Reconsolidation: retrieval can re-open consolidation window.

Patient H.M. (Henry Molaison, bilateral MTL resection 1953, Brenda Milner) — selective anterograde amnesia with intact working + procedural memory.

9. Reward, decision-making, executive function

9.1 Mesolimbic dopamine and reward prediction error

VTA dopamine neurons project to NAc (mesolimbic), PFC (mesocortical), and dorsal striatum (nigrostriatal from SNc). Wolfram Schultz (1997, Science) showed dopamine neurons encode a reward prediction error (RPE): δ = reward − expected reward.

  • Unexpected reward → phasic burst.
  • Predicted reward → no response (signal transfers to cue).
  • Predicted reward omitted → pause below baseline.

Maps cleanly onto temporal-difference learning (Sutton & Barto). Foundational link between neuroscience and reinforcement learning.

9.2 Decision theory

  • Expected utility: EU(a) = Σ p(s|a) · u(s) (von Neumann–Morgenstern).
  • Prospect theory (Kahneman & Tversky, Nobel 2002 econ.): reference-dependence, loss aversion, probability weighting.
  • Neural correlates: vmPFC (valuation), striatum (anticipated reward), insula (risk/loss), dlPFC (control over choice).

9.3 Executive function and prefrontal cortex

Patricia Goldman-Rakic: persistent delay-period firing in dlPFC (area 46) supports working memory and goal maintenance. Multiple PFC subdivisions: dlPFC (working memory, planning), vmPFC (value), OFC (outcome value, model-based), ACC (conflict, effort), FPC/area 10 (meta-cognition, multitasking — Koechlin).

Cognitive control = ability to override habitual responses (Stroop, flanker, Go/No-Go). Conflict monitoring (ACC) → control allocation (dlPFC) — Botvinick.

10. Language

  • Broca’s area (Broca 1861, left IFG 44/45): production. Lesion → non-fluent (Broca’s) aphasia: effortful, agrammatic, preserved comprehension.
  • Wernicke’s area (Wernicke 1874, left STG 22): comprehension. Lesion → fluent but meaningless (Wernicke’s) aphasia, poor comprehension.
  • Arcuate fasciculus: connects them. Disconnection → conduction aphasia.
  • Modern view (Hickok & Poeppel dual-stream): dorsal (sound → articulation) + ventral (sound → meaning).

11. Attention and consciousness

  • Posner cuing: alerting, orienting, executive networks.
  • Dorsal attention network (FEF, IPS — top-down) vs ventral attention network (TPJ, VFC — bottom-up salience). Corbetta & Shulman 2002.
  • Salience network: anterior insula + dorsal ACC (Seeley, Menon).
  • Theories of consciousness: Global Workspace (Dehaene, Baars), Integrated Information Theory (IIT, Tononi, Φ), Higher-Order Theories, predictive processing.

12. Circadian and sleep

  • Suprachiasmatic nucleus (SCN): ~20,000 neurons, master clock. Period ~24.2 h; entrained by light via retinohypothalamic tract (intrinsically photosensitive RGCs, melanopsin).
  • Molecular clock: TTFL with BMAL1/CLOCK transcription factors driving PER1/2/3 and CRY1/2, which inhibit them. Jeffrey Hall, Michael Rosbash, Michael Young — Nobel Prize 2017 for discoveries of molecular mechanisms controlling circadian rhythm (period gene in Drosophila).
  • Sleep stages (EEG): wake (β/α), N1, N2 (sleep spindles 11–16 Hz, K-complexes), N3 (slow-wave sleep, δ <4 Hz), REM (rapid eye movement, theta + ponto-geniculo-occipital waves, atonia). ~90-min cycles.
  • Functions: memory consolidation (replay during SWS + REM), glymphatic clearance of waste including Aβ (Nedergaard 2013), synaptic homeostasis (Tononi-Cirelli SHY hypothesis).

13. Development

Neurulation begins ~3 weeks post-conception (human): ectoderm folds → neural tube. Defects → spina bifida, anencephaly (folate-preventable).

Stages:

  1. Neurogenesis: ventricular zone proliferation; symmetric → asymmetric divisions (Notch, Pax6).
  2. Migration: radial glia scaffold for cortical excitatory neurons (inside-out, L6 first → L2/3 last). Lissencephaly (LIS1, DCX) — migration failure.
  3. Axon guidance: growth cone responds to netrins, slits, semaphorins, ephrins (attractive/repulsive). Sperry chemoaffinity (Nobel 1981).
  4. Synapse formation: synaptogenesis peaks postnatal year 2–3 in human cortex; ~50% later eliminated.
  5. Pruning: microglia (C1q/C3 complement), activity-dependent. Excess in ASD-associated lines; deficient in schizophrenia (C4A — Sekar 2016).
  6. Myelination: postnatal to ~30 years (PFC last).
  7. Critical periods: e.g. ocular dominance V1 (Hubel-Wiesel monocular deprivation), language acquisition. Closure tied to PV interneuron maturation, perineuronal nets, GABA/Otx2.

14. Aging and neurodegeneration

14.1 Healthy aging

Brain volume decreases ~0.2%/year after 30, accelerating after 60. White-matter integrity declines (FA on DTI). Episodic memory + processing speed decline; semantic + crystallized intelligence preserved. Cognitive reserve mitigates.

14.2 Alzheimer’s disease

  • ~55 M people worldwide (2024); leading cause of dementia.
  • Pathology: extracellular amyloid-β (Aβ40/42) plaques (from APP cleavage by β- + γ-secretase) and intracellular hyperphosphorylated tau neurofibrillary tangles.
  • Amyloid cascade hypothesis (Hardy & Higgins 1992): Aβ → tau → neurodegeneration. Genetic support: APP, PSEN1/2 mutations cause familial early-onset.
  • 2024–26 therapeutics: lecanemab (Eisai/Biogen, Leqembi, anti-Aβ protofibrils — FDA full approval Jul 2023, modest cognitive slowing ~27% on CDR-SB, ARIA-E/H side effects); donanemab (Eli Lilly, Kisunla — FDA approved Jul 2024). Both modest effect; debate ongoing. AHEAD 3-45 preclinical trial enrolling.

14.3 Parkinson’s disease

  • ~10 M worldwide. SNc dopamine neuron loss → bradykinesia, rigidity, resting tremor, postural instability.
  • α-synuclein aggregates (Lewy bodies). Genetic: SNCA, LRRK2, GBA, PRKN.
  • L-DOPA (1960s, Cotzias) — dopamine precursor crosses BBB; gold-standard symptom relief but motor fluctuations.
  • DBS of STN/GPi for advanced PD (Benabid 1990s).
  • 2024–26: PASADENA, SPARK trials of anti-α-synuclein antibodies (prasinezumab — phase 2 mixed). Levodopa-carbidopa enteral suspension (Vyalev/foslevodopa, AbbVie, FDA approved Oct 2024).

14.4 Other neurodegenerative diseases

  • Huntington’s disease: autosomal dominant; CAG trinucleotide-repeat expansion in HTT (>36 repeats); striatal MSN loss; chorea + cognitive + psychiatric. Tominersen (Roche, ASO) phase 3 failed 2021; GENERATION HD2 lower-dose ongoing 2024–26.
  • ALS (amyotrophic lateral sclerosis): motor-neuron degeneration; ~10% familial — SOD1, C9orf72 (also FTD), TDP-43, FUS. Tofersen (Qalsody, Biogen ASO targeting SOD1, FDA approved Apr 2023) for SOD1 ALS.
  • Frontotemporal dementia (FTD): tau, TDP-43, FUS subtypes; C9orf72 also. Personality, language variants (PNFA, semantic, behavioral).
  • Prion diseases: misfolded PrPSc templates conversion. Creutzfeldt-Jakob (sporadic + familial + variant from BSE), GSS, FFI, kuru. Stanley Prusiner Nobel 1997.
  • Multiple sclerosis: autoimmune demyelination; relapsing-remitting and progressive forms. Disease-modifying therapies: ocrelizumab (anti-CD20), natalizumab, fingolimod, cladribine.

15. Recording techniques

15.1 Single-cell electrophysiology

  • Patch-clamp (Neher & Sakmann, Nobel 1991): cell-attached, whole-cell, perforated, outside-out, inside-out configurations. Resolution: single channel currents (pA).
  • Sharp electrode: high-resistance, intracellular recording, less disturbed cytoplasm.
  • Slice preparation: 300 µm brain slices, ACSF perfusion; standard for circuit and plasticity work.

15.2 Extracellular multi-electrode

  • Utah array (Blackrock Neurotech, originally Normann group Utah ~1990s): 96-channel silicon microelectrodes, ~4×4 mm, ~1.5 mm shafts. Workhorse for BrainGate clinical trials (Donoghue, Hochberg) since 2004 enabling tetraplegic patients to control cursors and robotic arms.
  • Neuropixels (IMEC / Allen Institute / HHMI / UCL collaboration). Neuropixels 1.0 (2017): single 1 cm CMOS probe, 960 sites, 384 simultaneously recorded; transformative for systems neuroscience. Neuropixels 2.0 (2021): 4-shank, 5,120 sites, smaller, chronic. Neuropixels Ultra (2023–24) for axonal/dendritic resolution; Neuropixels-Opto (2024) for optogenetic targeting.
  • Spike sorting: Kilosort (Pachitariu) GPU-based — handles 1,000+ neurons per probe.

15.3 Optical methods

  • Two-photon calcium imaging (Denk, Strickler, Webb 1990 Science): Ti:Sapph laser, ~1 mm depth, ~µm resolution. GCaMP (Nakai 2001; GCaMP6/7/8 Looger/Janelia; jGCaMP8 2023 — fast, kinetics matched to spikes) genetically encoded Ca²⁺ indicator.
  • Three-photon imaging (Xu 2013, Wang/Xu 2024 mouse cortex L6 + hippocampus through intact skull).
  • Light-sheet (SPIM): whole-zebrafish-brain functional imaging (Ahrens 2013, ~100,000 neurons simultaneously).
  • Voltage imaging: ASAP, Voltron, JEDI-2P (St-Pierre, Schreiter), Positron — millisecond resolution; mechanistic challenges (signal/noise, photobleaching) being addressed.
  • Mesoscope (Sofroniew 2016 Janelia 2p-RAM): cm-scale FOV, multi-region.
  • Miniscopes (UCLA, Inscopix nVista/nVoke): head-mounted ~2 g for freely moving rodents.

15.4 Whole-brain, non-invasive

  • EEG: scalp electrodes; ms temporal, cm spatial (volume conduction). Frequency bands: δ (0.5–4 Hz), θ (4–8), α (8–13), β (13–30), γ (30–100+). ERPs: N100, P200, N400 (semantic), P300 (oddball, attention).
  • MEG: SQUID magnetometers (helium-cooled) or newer optically-pumped (OPM) at room temperature (QuSpin, Cerca) — wearable MEG. Same ms temporal as EEG but better source localization (skull transparent to magnetic fields).
  • fMRI BOLD: Seiji Ogawa 1990 — blood-oxygen-level-dependent contrast; deoxyhemoglobin paramagnetic. Indirect (hemodynamic, ~6 s lag); ~mm spatial; 1.5/3/7 T. Logothetis (2001 Nature) — BOLD correlates with LFP more than spikes. 7T human (Siemens Terra, GE) routine in research; 11.7T Iseult INSERM (Saclay, first human images 2024).
  • PET: radiotracers (18F-FDG glucose; 11C-raclopride D2; 18F-AV-1451 tau; PiB Aβ). Worse spatial than fMRI but specific.
  • fNIRS: near-infrared, hemodynamic, portable.
  • TMS: transcranial magnetic stimulation (Barker 1985); non-invasive cortical perturbation. rTMS / iTBS approved depression (OCD, smoking cessation extensions). MagVenture, Brainsway deep TMS.
  • tDCS / tACS: transcranial direct/alternating current; modest effects.
  • Focused ultrasound (FUS): emerging non-invasive deep neuromodulation; BBB opening for drug delivery.

15.5 Intracranial in humans

  • ECoG (electrocorticography): subdural grid/strip for epilepsy localization. Higher SNR + bandwidth than scalp EEG; basis of speech decoding (Chang UCSF — 2023 Nature, near-natural-speech BCI for ALS patient).
  • sEEG (stereoEEG): depth electrodes; standard for epilepsy and increasingly research.
  • Deep brain stimulation (DBS): Medtronic Activa, Abbott Infinity, Boston Scientific Vercise. Targets: STN/GPi (PD), VIM (essential tremor), GPi (dystonia), ANT (epilepsy), SCG (depression — investigational), NAc (OCD).

16. Omics and connectomics

16.1 Single-cell and spatial transcriptomics

  • scRNA-seq + snRNA-seq: 10x Genomics, Drop-seq, SMART-seq.
  • BICCN (NIH BRAIN Initiative Cell Census Network): mouse cortex cell-type atlas (2021 Nature package, 17 papers, ~5,000 transcriptomic types). BICAN (Cell Atlas Network) extending to whole human brain — 2024 first-pass atlas (>3,000 cell types).
  • Spatial transcriptomics: MERFISH (Zhuang), Visium, Stereo-seq (BGI), Xenium (10x), seqFISH+. 2024–26: subcellular resolution and protein co-profiling.
  • HuBMAP — Human BioMolecular Atlas Program (NIH) — whole-body, including CNS.

16.2 Connectomics

  • C. elegans: complete connectome of all 302 neurons (hermaphrodite — White, Southgate, Thomson, Brenner 1986 Phil. Trans. B); male added (Cook 2019); MIND-ful re-analyses ongoing.
  • Drosophila: hemibrain (~25,000 neurons, ~20 M synapses — Janelia + Google 2020, Scheffer); full adult fly brain (FlyWire, Princeton + Cambridge 2024 Nature — all ~140,000 neurons with synaptic connectivity).
  • Mouse: MICrONS (Allen + BCM + Princeton 2025 Nature package) — 1 mm³ of mouse V1, ~200,000 neurons, ~500 M synapses, with co-registered functional Ca²⁺ imaging from same neurons. Allen Institute “pinhead” (Apr 2024) — 1 mm³ cortex sample.
  • Human: ~1 mm³ temporal cortex (Lichtman + Google 2024 Science) ~57,000 cells, ~150 M synapses. BICAN + EBRAINS scaling.
  • Methods: serial-section electron microscopy (ssEM, ATUM), MultiSEM (Zeiss), expansion microscopy. Segmentation by deep learning (flood-fill networks — Januszewski).

16.3 Behavior

  • DeepLabCut (Mathis 2018) — markerless pose estimation; standard for behavioral tracking. SLEAP (Pereira/Murthy), B-SOiD, VAME for behavioral syllable discovery.

17. Brain-computer interfaces (BCIs) — 2024–26 state of the art

17.1 Invasive intracortical

  • Neuralink (Musk 2016). N1 implant: 1,024 electrodes on 64 flexible threads, hermetic skull-flush package, inductive charging. R1 surgical robot performs threading. First human implant January 28 2024 (patient P1, Noland Arbaugh — quadriplegic, cervical SCI). By 2025 mid-year ~5 patients implanted (PRIME study) — cursor control, gaming, CAD use; some threads retracted in P1, mitigated algorithmically. Convoy trial (companion brainstem-arm signal) and CONVOY+ extension planning per filings.
  • Precision Neuroscience (“Layer 7” cortical interface): ultra-thin (~30 µm) microelectrode arrays placed on surface (epicortical/sub-pial) — minimally invasive vs penetrating. FDA 510(k) clearance Apr 2025 for short-term recording (up to 30 days). 4,096 electrodes per array; aiming multiple arrays.
  • Paradromics Connexus: 421 microelectrodes per module; first human implant 2025 in epilepsy patient (intraoperative use); chronic trial pending.
  • Blackrock Neurotech Utah array (legacy clinical, 96-channel) — used by BrainGate consortium since 2004; ~50+ patients cumulative. New Neuralace flexible array (acquired 2023) advancing.
  • Onward (BSCI): epidural spinal stimulation for SCI walking restoration (Courtine + Bloch STIMO); brain-spine interface (BSI, Lausanne 2023 Nature) — Utah-array M1 decoder + ARC-IM spinal stim; ongoing PRIMO trial.

17.2 Less-invasive

  • Synchron Stentrode: 16 electrodes on a self-expanding nitinol stent delivered via internal jugular vein into superior sagittal sinus — no craniotomy. SWITCH and COMMAND trials (US/Australia) since 2019. As of 2024–25: ~10 patients implanted, smartphone control, integration with Apple Vision Pro accessibility prototype (Sep 2024).
  • Motif Neurotech, Inbrain Neuroelectronics (graphene): emerging.

17.3 Non-invasive

  • OpenBCI Galea, Kernel Flow (TD-fNIRS), Cognixion.
  • Diffuse and TD-NIRS, dry-electrode EEG (Muse, Neurable), OPM-MEG (Cerca) wearable.
  • Meta / Reality Labs sEMG wristband (CTRL-Labs acquisition) for peripheral interface.

17.4 Clinical results 2024–26

  • Speech BCIs: Chang lab UCSF — multi-articulator decoding ~78 WPM (Willett 2023 Nature) and avatar speech + facial expressions for stroke patient (Metzger 2023 Nature); 2024–25 extended to bilingual decoding and naturalistic-speech LLM-coupled outputs.
  • BrainGate consortium ongoing; high-performance handwriting BCI (Willett 2021, 90 cpm) now coupled with LLM decoders.
  • Tetraplegic robotic-arm + grasp control (Hochberg, Donoghue).

18. Pharmacology

ClassExamplesMechanism
SSRIsFluoxetine (Prozac), sertraline, escitalopramBlock SERT
SNRIsVenlafaxine, duloxetineBlock SERT + NET
TCAsAmitriptylineBlock SERT/NET; antimuscarinic, antihistamine
MAOIsPhenelzine, selegilineInhibit MAO-A/B
Atypical antidepressantsBupropion (DAT/NET), mirtazapine (α2, 5-HT2/3)Mixed
Typical antipsychoticsHaloperidol, chlorpromazineD2 antagonism
Atypical antipsychoticsRisperidone, olanzapine, clozapine, lurasidone, aripiprazole (partial D2)D2 + 5-HT2A
BenzodiazepinesDiazepam, lorazepam, alprazolamPositive allosteric mod GABA-A α1/2/3/5
Z-drugsZolpidem, eszopicloneGABA-A α1-selective
Mood stabilizersLithium, valproate, lamotrigine, carbamazepineMultiple (GSK3β, Na channels)
StimulantsMethylphenidate, amphetamine, modafinilDAT/NET inhibition or reversal
Cognitive enhancers (AD)Donepezil, rivastigmine, galantamine, memantineAChE inhibitors, NMDA antagonist
Anti-Aβ mAbsLecanemab (2023 FDA), donanemab (2024 FDA)Aβ protofibril / plaque
Anti-PDL-DOPA + carbidopa, dopamine agonists, MAO-B inh., COMT inh., apomorphineDopamine replacement / preservation
OpioidsMorphine, oxycodone, fentanyl, naloxoneµ/δ/κ agonists/antagonist
CannabinoidsTHC, CBD, nabiximols (Sativex)CB1/CB2
Ketamine / esketamineSpravato (J&J, FDA 2019 for TRD)NMDA antagonist; rapid antidepressant
PsychedelicsPsilocybin (Compass COMP360 phase 3), LSD (MindMed MM-120 LSD for GAD phase 3), DMT, mescaline5-HT2A agonists
MDMALykos (MAPS) MDMA-assisted therapy5-HT/DA/NE releaser

18.1 Psychedelic and MDMA developments (2024–26)

  • MAPS / Lykos Therapeutics filed NDA for MDMA-assisted therapy for PTSD; FDA Advisory Committee June 4, 2024 voted against approval citing trial design/blinding/efficacy concerns; FDA Complete Response Letter (CRL) August 9, 2024 required additional phase 3 trial. Re-pivot ongoing 2025–26.
  • Compass Pathways COMP360 psilocybin: phase 3 COMP005 (single-dose monotherapy for TRD) results expected mid-2025; COMP006 (multi-dose) into 2026.
  • MindMed MM-120 (LSD) for GAD — phase 3 trials Voyage + Panorama 2024–26.
  • atai Life Sciences, Cybin (CYB003 deuterated psilocybin analog).
  • Esketamine (Spravato) monotherapy expanded indication FDA Jan 2025.

19. Computational and cognitive neuroscience

19.1 Neural codes

  • Rate coding: firing rate over windows. Adrian (1926, Nobel 1932).
  • Temporal coding: precise spike times, phase relative to oscillation, ISI patterns.
  • Population coding: vector decoding (Georgopoulos M1 reaching).
  • Sparse coding: V1 simple cells (Olshausen & Field 1996) — Gabor-like filters from natural-image priors.
  • Predictive coding (Rao & Ballard 1999, Friston free-energy): cortex codes prediction errors.

19.2 Models

  • Hodgkin-Huxley (biophysical, 4 variables).
  • Leaky integrate-and-fire: τ dV/dt = −(V − V_rest) + R · I; fires at threshold.
  • Izhikevich (2003): two equations, captures 20+ firing patterns efficiently.
  • Networks: Hopfield (1982) attractor; Wilson-Cowan; reservoir computing; spiking neural networks.
  • Cortical microcircuit models: Blue Brain (Markram 2015 — neocortical column ~31,000 neurons reconstructed); Allen V1 model (Billeh 2020).
  • Whole-brain models: Virtual Brain (TVB), Human Brain Project legacy (concluded 2023). EBRAINS continues platform. SOTA whole-mouse-brain simulation efforts (Jülich + Allen) running ~1:1 scale neuron-by-neuron simulation on exascale (JUPITER, Frontier).
  • Neural ODEs (Chen 2018) and continuous-time models bringing biological plausibility to deep learning.

19.3 Brain-inspired AI and interpretability

  • Transformer attention (Vaswani 2017) inspired in name by neural attention but functionally distinct; see transformer-architecture. Active dialogue: are LLMs reasonable models of cortex? (Schrimpf Brain-Score; Goldstein/Hasson aligned representations in podcast listening — Nature 2022, 2024 updates).
  • Mechanistic interpretability (Anthropic, Olah lab): circuit tracing, feature decomposition. Sparse autoencoders on Claude (2024 Templeton scaling monosemanticity), Crosscoders (2024), Transcoders (2024); circuit tracing in production-scale models (Lindsey, Anthropic 2025). Parallels with neuroscience: features ↔ tuning curves, circuits ↔ canonical microcircuits.
  • NeuroAI white papers (Zador, Yamins, Macke 2023) — bidirectional benchmark and dialogue.
  • 2024–26 trend: AI for brain reading — diffusion-model image reconstruction from fMRI (MindEye2 2024); LLM-aligned ECoG speech decoding; foundation models for neural data (POYO Brunton/Azabou 2023; Neuroformer 2024; EEG-GPT 2024–25).

20. 2024–26 landmarks

  • MICrONS (Apr 2025 Nature package): 1 mm³ mouse V1 connectome + function; demonstrates feasibility of structure-function correlation at scale.
  • FlyWire whole adult fly brain connectome (Oct 2024 Nature): all neurons + synapses, ~140k neurons. Cell-type taxonomy, neurotransmitter prediction, connectome motif analysis.
  • Human cortex connectome 1 mm³ (Shapson-Coe et al. Lichtman + Google 2024 Science).
  • BICAN human whole-brain cell atlas first-pass (2024 Science package).
  • Neuralink first 5 human implants and longitudinal data; Synchron smartphone control; Precision 510(k); Paradromics first human.
  • Lecanemab + donanemab approvals — first disease-modifying AD therapies (modest).
  • MDMA-PTSD Lykos CRL: psychedelic regulatory speed-bump.
  • Cortical organoids and assembloids (Pașca, Lancaster) — 1+ year culture; growing functional realism; transplantation into rodent cortex (Pașca 2022) with integration; ethical debates.
  • AI brain reading: image, speech, semantic decoding from fMRI (Tang/Huth 2023; Scotti MindEye 2024); cross-subject decoders; closer brain-to-text systems with LLMs.
  • BRAIN Initiative 2.0: NIH BRAIN extends to BICAN, BICCN, Connects, Armamentarium for tools, CONNECTS for ultra-high-res connectomics.
  • EBRAINS, EU Brain Conduction Network, Australian Brain Project, China Brain Project — international coordination.

21. Cross-references

  • cell-molecular-biology — membranes, ion gradients, ATPases, vesicle trafficking, signal transduction.
  • genetics-and-genomics — channelopathies, neurodevelopmental disorders, GWAS for psychiatric traits.
  • biochemistry-foundations — biochemistry of neurotransmitters, enzyme kinetics (AChE), thermodynamics of ATP-driven pumps.
  • bioinstrumentation — EEG/ECG/EMG electrodes, amplifiers, signal-conditioning, safety, IEC standards.
  • biomechanics — neuromuscular coupling, motor units, prosthetic control.
  • probability-fundamentals — Poisson spike statistics, Bayesian inference in perception, RL theory.
  • transformer-architecture — attention mechanism (named after but distinct from neural attention); brain-inspired AI; interpretability parallels.

22. Citations and key references

Foundational textbooks

  • Kandel ER, Koester JD, Mack SH, Siegelbaum SA (eds). Principles of Neural Science, 6th ed., McGraw-Hill, 2021.
  • Bear MF, Connors BW, Paradiso MA. Neuroscience: Exploring the Brain, 5th ed., Jones & Bartlett, 2024.
  • Purves D et al. Neuroscience, 6th ed., Sinauer/Oxford, 2017.
  • Squire LR et al. Fundamental Neuroscience, 4th ed., Academic Press, 2012.
  • Dayan P, Abbott LF. Theoretical Neuroscience, MIT Press, 2001.

Foundational papers and Nobels

  • Hodgkin AL, Huxley AF. A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. 117:500–544 (1952). Nobel Prize Physiology or Medicine 1963 (with Eccles).
  • Hubel DH, Wiesel TN. Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. J. Physiol. 160:106–154 (1962). Nobel Prize 1981.
  • O’Keefe J, Dostrovsky J. The hippocampus as a spatial map. Brain Res. 34:171–175 (1971). Hafting/Fyhn/Molden/Moser/Moser. Microstructure of a spatial map in the entorhinal cortex. Nature 436:801–806 (2005). Nobel Prize 2014 (O’Keefe, Moser & Moser).
  • Rothman JE, Schekman R, Südhof TC — Nobel Prize 2013 for vesicle traffic machinery.
  • Bliss TVP, Lømo T. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit. J. Physiol. 232:331–356 (1973).
  • Schultz W, Dayan P, Montague PR. A neural substrate of prediction and reward. Science 275:1593–1599 (1997).
  • Neher E, Sakmann B — Nobel Prize 1991 for patch clamp.
  • Hall JC, Rosbash M, Young MW — Nobel Prize 2017 for circadian clock.
  • Buck L, Axel R — Nobel Prize 2004 for olfactory receptors.
  • Julius D, Patapoutian A — Nobel Prize 2021 for TRP/PIEZO.
  • Kandel ER — Nobel Prize 2000 for memory storage (Aplysia, LTP).
  • Carlsson, Greengard, Kandel — Nobel Prize 2000 (signal transduction in nervous system).
  • Prusiner SB — Nobel Prize 1997 for prions.
  • Sperry RW — Nobel Prize 1981 for chemoaffinity.
  • Eccles JC — Nobel Prize 1963 for IPSPs.
  • Granit, Hartline, Wald — 1967 vision.
  • von Békésy — 1961 cochlear traveling wave.

Connectomics and recording

  • White JG, Southgate E, Thomson JN, Brenner S. The structure of the nervous system of the nematode Caenorhabditis elegans. Phil. Trans. R. Soc. B 314:1–340 (1986).
  • Scheffer LK et al. (Janelia + Google). A connectome and analysis of the adult Drosophila central brain. eLife 9:e57443 (2020). (Hemibrain.)
  • Dorkenwald S et al. (FlyWire). Neuronal wiring diagram of an adult brain. Nature (2024). (~140k neurons.)
  • MICrONS Consortium. Functional connectomics spanning multiple areas of mouse visual cortex. Nature (Apr 2025).
  • Shapson-Coe A et al. (Lichtman, Google). A petavoxel fragment of human cerebral cortex reconstructed at nanoscale resolution. Science 384 (2024).
  • Steinmetz NA, Aydın Ç, Lebedeva A et al. Neuropixels 2.0: A miniaturized high-density probe for stable, long-term brain recordings. Science 372 (2021).
  • Jun JJ et al. Fully integrated silicon probes for high-density recording of neural activity. Nature 551:232–236 (2017). (Neuropixels 1.0.)
  • Denk W, Strickler JH, Webb WW. Two-photon laser scanning fluorescence microscopy. Science 248:73–76 (1990).
  • Ogawa S et al. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. PNAS 87 (1990).
  • Logothetis NK et al. Neurophysiological investigation of the basis of the fMRI signal. Nature 412:150–157 (2001).
  • Willett FR et al. A high-performance speech neuroprosthesis. Nature 620 (2023).
  • Metzger SL et al. A high-performance neuroprosthesis for speech decoding and avatar control. Nature 620 (2023).
  • Mathis A et al. DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nat. Neurosci. 21 (2018).

2024–26 BCI and clinical

  • Neuralink Corp. PRIME study FDA IDE 2023; clinicaltrials.gov NCT06429735; first implant Jan 28 2024.
  • Synchron — SWITCH (Australia) and COMMAND (USA) Stentrode trials, results presentations 2023–25.
  • van Heuvel et al. (Onward/Lausanne). Walking naturally after spinal cord injury using a brain-spine interface. Nature 618:126–133 (2023).
  • Lecanemab — van Dyck CH et al. Lecanemab in early Alzheimer’s disease. NEJM 388:9–21 (2023). Donanemab — Sims JR et al. JAMA 330:512–527 (2023). FDA full approvals 2023/2024.
  • Lykos (MAPS PBC) — MAPP1 and MAPP2 MDMA-AT phase 3; FDA AdComm + CRL Aug 2024.

Computational and AI

  • Hodgkin & Huxley 1952; Izhikevich 2003 IEEE Trans Neural Netw; Rao & Ballard 1999 Nat Neurosci; Friston 2010 Nat Rev Neurosci (free energy).
  • Yamins DLK, DiCarlo JJ. Using goal-driven deep learning models to understand sensory cortex. Nat. Neurosci. 19 (2016).
  • Schrimpf M et al. Brain-Score benchmark and CORnet. bioRxiv and updates.
  • Templeton A et al. (Anthropic). Scaling Monosemanticity (2024). Lindsey J et al. (Anthropic). Circuit tracing on Claude 3.5 Haiku (2025).
  • Zador A, Yamins D, Macke J et al. Catalyzing next-generation AI through NeuroAI. Nat. Commun. 14 (2023).

End of reference. SI units used throughout. All cited Nobel laureates listed with year; all probes, technologies, and BCI companies named with current (2024–26) trial status where applicable.

Graph View