π Neuroscience Landscape
Key Reading
- Principles of Neural Science, McGraw Hill / Medical; 6th edition (March 29, 2021)
Labs
| Penn Image Computing & Science
| Brain Dynamics | Lee Lab
| Perceptual Reality | Gallant Lab | Theunissen Lab | Bouchard Lab
Companies
Organizations
Software
Labs
- NITRC - large list of neuroimaging tools
- LEAD-DBS - model of deep brain stimulation as population level
- VERTEX - model of electrical stimulation at neuronal level
- SpikeInterface - a unified framework for spike sorting
- MICrONS - machine intelligence from cortical networks, IARPA
- SpikeGLX - used in neuropixels, Allen Institute
- SMART - spatial registration tool, BICCN
- HistoloZee - spatial registration tool, MRI visualization, histology, BICCN
- cnpkg - cortical network simulator, optimized with CUDA, Seung Lab, MIT
- fimpy - parallelizable processing of fluorescent imaging data, HDF5
- ... many more software is available in labs and organizations links
Enterprise
Hardware
Basics
- brain is not a randomly connected mesh, there is a lot of high and low level structure
- grey matter β surface layer, cortex, mostly neurons, ~6 layers of neurons; white matter β insides, connections between neurons, moslty axons
- current representation β vector at layer in network, what is being processed or thinked now; abeyant representation β stored knowledge, weights at layer of network, what can be thinked or recalled
- local coding β concept of "grandmother" is handled by a single neuron; distributed coding β concept of "grandmother" is distributed on many neurons, vector coding
- connectomics β given brain is a graph, what are its properties, how does it differ in animals, how it comes into being, how it changes with time Status of mapping: C elegans worm, Drosophilia fruit fly β fully mapped; pigeon, cat, monkey, human β good, but not full. A lot of structure of connectome is encoded in genes
- plasticity β how neurons change their connection weights: structural plasticity β changing what is connected; functional plasticity β changing strength of connection
- time is important, as a lot of processing is oscilating potentiation; many recurrent loops; different types of neurons
- neurons are either inhibitory or excitatory
- brain development follows phases β neural plate, progenitor cells, neural fates, migrataion, synapse path finding, synapse connection forming, cells and synapse pruning
- action potential β propagation of neuron membrane depolarization and subsequent re-polarization to original state by activation of Na+ and subseqent K+ ion channles spread over membrane
- ion channles have different conductance-voltage dependenance, which leads them to open-close at different conditions. this makes action potential posssible. some ion channels activated mechanically, some chemically, some via photons, some by voltage.
Courses
Additional Reading
- Fundamental Neuroscience, Squire, Academic Press
- Connectome: How the Brain's Wiring Makes Us Who We Are, Sebastian Seung, PhD, prof at MIT, Director of R&D at Samsung, 2013
- The Computational Brain, Patricia S. Churchland, Terrence J. Sejnowski, MIT Press
- Changing Connectomes, Marcus Kaiser, MIT Press
- Dynamic Patterns: The Self-Organization of Brain and Behavior, MIT Press
- Neuroscience of Mathematical Cognitive Development, Rhonda Douglas Brown, Springer
- Chasing Men on Fire, Stephen G Waxman, MIT Press
- The Ego Tunnel, Thomas Metzinger
- The Neocortex, MIT Press, 2019
- Emergent Brain Dynamics: Prebirth to Adolescence, MIT Press, 2018
- Translational Neuroscience: Toward New Therapies, MIT Press, 2015