BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Engineered neural networks as self-organised computational substra
 tes in the healthy and lesioned CNS - Ioanna Sandvig - Norwegian Universit
 y of Science and Technology
DTSTART:20181025T110000Z
DTEND:20181025T120000Z
UID:TALK108652@talks.cam.ac.uk
CONTACT:Dr Romina Vuono
DESCRIPTION:Neural networks demonstrate self-organising behaviour undersco
 red by\nemergence/morphogenesis and self-organised criticality (SoC). Emer
 gence\ninvolves cell behaviours driven by local cell-cell interactions and
  the\nspontaneous appearance of a highly ordered structure or function as 
 a whole\,\nnot simply explained by the sum of the elements’ complexity. 
 At the same time\,\nSoC represents a universal characteristic of neural sy
 stems\, being a\nspontaneous dynamic state established in networks of mode
 rate complexity. In\nthese networks\, cascades of spontaneous activity are
  typically characterised by\npower-law distributions and rich\, stable spa
 tiotemporal patterns (neuronal\navalanches). Thus SoC plays a functional r
 ole in neural computation by\nputatively maximising information transmissi
 on\, the number of stable patterns\,\ninformation capacity\, and the range
  of usable inputs in a neural system.\n\nA developed/mature neural network
  is expected to reach a fine balance of neural\nexcitation and inhibition\
 , consistent with normal function. Intrinsic or\nextrinsic perturbations t
 o the network  (e.g.evolving disease-related\npathology\; age-associated a
 nd/or epigenetic changes/ DNA damage) will result in\nremodelling of struc
 tural and functional connectomes. The associated changes\ncan be adaptive 
 or maladaptive in nature and involve an interplay between\nhomeostatic and
  Hebbian plasticity contingent on the inherent state of the\nneuron\, natu
 re of the perturbation\, and state of the microenvironment. A better\nunde
 rstanding of such processes can be instrumental for our ability to\nelucid
 ate fundamental mechanisms of neural network dynamics in the healthy and\n
 lesioned CNS. A highly promising perspective towards a better understandin
 g of\nneural network behaviour is the application of advanced in vitro and
  in silico\nmodelling tools incorporating deep learning principles.\n\nSpe
 cifically\, by applying advanced computational tools and machine learning\
 nprinciples\, we can identify patterns that can reveal principal component
 s\ninfluencing adaptive or maladaptive network responses as well as critic
 al\nvulnerability states. This is crucial for early diagnosis\, and timing
 /type of\nintervention (e.g. gene repair\, DNA repair\, or pharmacological
  treatment)\, as\nwell as assessment of their efficacy.  Fundamental resea
 rch questions which we\ncan address may thus include: can we associate and
 /or predict\nmorphology-activity relationships at the neural network and s
 ynaptic level with\nprogressive disease-related pathology? Do the most vul
 nerable neural networks\nself-regulate intrinsic responses to perturbation
 s in an attempt to preserve\nnormal function? If so\, to what extent can h
 omeostatic plasticity mask\nmanifestation of disease pathology at the prod
 romal disease state? Are we able\nto identify critical vulnerability state
 s that might instruct the timing and/or\nnature of appropriate interventio
 ns? Which aspects of complex neural network\ndynamics and at which level/s
 cale are relevant for restoration of functional\nconnectivity?\n\nIn this 
 talk\, I will present an overview of our relevant research activities at\n
 NTNU\, providing examples of how we apply morphogenetic neuroengineering\,
 \ncomputational tools and deep learning principles in an advanced modellin
 g\nplatform for the study and elucidation of evolving neural network dynam
 ics in\nhealthy and perturbed conditions.
LOCATION:James Fawcett Seminar Room\, van Geest Building
END:VEVENT
END:VCALENDAR