Senior Lecturer, School of Computer Science,
   University of Lincoln, Lincoln, U.K.

Faculty, Athens International Masters Program in
Neuroscience, Dept of Biology, University of Athens, Greece

Associate Editor, Cognitive Computation

Associate Editor, Scholarpedia

Email: vcutsuridis AT lincoln DOT ac DOT uk

Edited proceedings

Books series

Home Page
   • Theoretical Neuroscience
    • Systems level models
    • Neural Network Models
    • Microcircuits Models
    • Synaptic and single neuron models
   • Machine Learning
Activities, Grants
Talks, Seminars
Education, Career

Computational neuroscience  
Arguably one of the most difficult problems in science today is the brain-mind problem. Brains are collections of billions of interconnected cells, each of them being an individual machinery, which receives, processes and transmits information. The human brain generates complex patterns of behavior at different scales of organization based on the large amount of neural components that interact simultaneously in a rich number of parallel ways. In order to understand the inherent complexity of such a system then models from various levels of spatial and temporal complexity including synaptic and single cell models, microcircuit models, network models and systems models need to be developed.
Systems level models of action planning, action execution, neuromodulation in movement disorders  
A systems level model of action planning and execution under neuromodulatory (dopamine) control in movement disorders such as Parkinson’s disease (PD) bradykinesia. The model is multi-modular consisting of a module (basal ganglia) capable of selecting the most appropriate motor command in a given context, a module (cortex) for coordinating and executing the final motor commands, and a module (spinal cord) for guiding the arm to its final target and providing proprioceptive (feedback) input of the current state of the muscle and arm to higher cortical and lower spinal centers.

The model [1, 2] provides a unified theoretical framework for PD bradykinesia and it is capable of producing a wealth of empirical findings such as increased behavioural reaction and movement times, repetitive bursts of muscle activation, reduction in the size and rate of development of the first agonist burst of EMG activity, movement variability, etc.

Extensions to the original model incorporated experimental evidence regarding the effects of spindle feedback on cortical cell activities as well as examined the effects of DA depletion on spinal activities, particularly the reciprocal disynaptic Ia inhibition and its effects on the activities of reciprocally activated a-MNs [3]. An interesting prediction of the new model was that the reduced reciprocal disynaptic Ia inhibition in the DA depleted case did not lead to the co-contraction of antagonist motor units and hence to the observed rigidity. The causes of MN co-contraction ought to be searched more centrally, potentially in the microcircuit of the motor cortex and/or the basal ganglia.

The latter prompted the quantitative investigation of the origins of the experimentally observed repetitive and co-contractive pattern of muscle activation in Parkinson's disease [6]. Computer simulations showed that an oscillatory disrupted globus pallidus internal segment (GPi) response signal comprising at least two excitation–inhibition sequences as an input to a normally functioning cortico-spinal model of movement generation resulted in a repetitive, but not cocontractive agonist–antagonist pattern of muscle activation. A repetitive and co-contractive pattern of muscle activation resulted when also dopamine was depleted in the cortex. Additional dopamine depletion in the spinal cord sites resulted in a reduction of the size, duration and rate of change of the repetitive and co-contractive EMG bursts. These results have important consequences in the development of PD therapies such as dopamine replacement in cortex and spinal cord, which can alleviate some of the impairments of Parkinson’s Disease such as slowness of movement (bradykinesia) and rigidity.

  1. Cutsuridis V, Perantonis S. (2006). A Neural Model of Parkinson's Disease Bradykinesia. Neural Networks, 19(4): 354-374

  2. Cutsuridis V. (2006). Neural Model of Dopaminergic Control of Arm Movements in Parkinson’s Disease Bradykinesia. Lecture Notes in Computer Science (LNCS) 4131, Springer-Verlag, 583-591.

  3. Cutsuridis V. (2007). Does Abnormal Reciprocal Inhibition Lead to Co-contraction of Antagonist Muscles? A Modeling Study. International Journal of Neural Systems, 17(4): 319-327

  4. Cutsuridis V. (2010). Neural network modeling of voluntary single joint movement organization. II. Parkinson’s disease. In: Chaovalitwongse WA et al. (eds), Computational neuroscience, Springer-Verlag, 193-212

  5. Cutsuridis V. (2010). Neural network modeling of voluntary single joint movement organization. I. Normal conditions. In: Chaovalitwongse WA et al. (eds), Computational neuroscience, Springer-Verlag, 181-192

  6. Cutsuridis V. (2011). Origins of a repetitive and co-contractive pattern of muscle activation in Parkinson’s disease. Neural Networks, 24(6): 592-601

  7. Cutsuridis V. (2013). Bradykinesia Models. In: Jaeger D, Jung R (eds), Encyclopedia of Computational Neuroscience. Springer, USA

  8. Cutsuridis V. (2013). Bradykinesia Models of Parkinson’s Disease. Scholarpedia, 8(9):30937

  9. Cutsuridis V. (2018). Bradykinesia Models. In: Jaeger D, Jung R (eds), Encyclopedia of Computational Neuroscience, 2nd Edition. Springer, USA

  10. Cutsuridis V. (2019). Basal ganglio-thalamo-cortico-spino-muscular model of Parkinson’s disease bradykinesia. In: Cutsuridis V (ed), Multiscale Models of Brain Disorders. Springer-Nature, Switzerland

Neural network level models of decision making in the antisaccade task in health and disease  
Humans and animals are constantly facing the problem of having to choose from a variety of possible actions as they interact with the environment. Both external and internal cues have to be used to guide their selection of a single action from many possible alternatives. Which action to choose in a given context may have important biological consequences to their survival. Decision making is regarded as an accumulation process of evidence about the state of the world and the utility of possible outcomes. Cognitive and behavioural neuroscientists have begun to investigate the neural basis of decision making using various behavioural paradigms. The behavioural paradigm often used is the antisaccade task.

A neural network model of the superior colliculus in the antisaccade task, where decisions were formed via stochastic accumulating processes and contrast enhancement of decision signals. The model was successful at explaining why the response times in the antisaccade task are so long and variable and at predicting accurately the shapes of correct and error RT distributions as well as the response probabilities of a large 2006 sample of subjects. The model predicted that there was no need of a top-down inhibitory signal that prevented the error prosaccade from being expressed, thus allowing the correct antisaccade to be released. This finding challenged the currently accepted view of saccade generation in the antisaccade task, which required a top-down inhibitory signal to suppress the erroneous saccade after the correct saccade has been expressed.

  1. Cutsuridis V, Smyrnis N, Evdokimidis I, Perantonis S. (2007). A Neural Network Model of Decision Making in an Antisaccade Task by the Superior Colliculus. Neural Networks, 20(6): 690-704

  2. Cutsuridis, V. (2017). Behavioral and computational varieties of response inhibition in eye movements. Philos Trans R Soc Lond B, 372: 20160196.

Investigated the biophysical mechanisms underlying the generation of the accumulator-like activity of the decision signals. The cortico-superior colliculus model predicted that only the NaP, NMDA and AMPA currents can produce the observed variability in the accumulator-like activities of the cortical decision signals driving the superior colliculus module.

  1. Cutsuridis V, Kahramanoglou I, Perantonis S, Evdokimidis I, Smyrnis N. (2005). A Biophysical Neural Model of Decision Making in an Antisaccade Task Through Variable Climbing Activity. Lecture Notes in Computer Science (LNCS) 3696, Springer-Verlag, 205-210

  2. Cutsuridis V, Kahramanoglou I, Smyrnis N, Evdokimidis I, Perantonis S. (2007). A Neural Variable Integrator Model of Decision Making in an Antisaccade Task. Neurocomputing, 70(7-9): 1390-1402

  3. Cutsuridis V. (2010). Neural accumulator models of decision making in eye movements. Adv Exp Med Biol, 657: 61-72

Schizophrenia, Obsessive-Compulsive Disorder (OCD) and Huntigton Disease (HD) patients show deficits in antisaccade performance: increased variability in saccade response times and increased error rates. In collaboration with two psychiatric groups from Germany and Greece, I have been investigating the mechanisms that give rise to these antisaccade performance deficits.

This research has led a number of important discoveries on what goes wrong in the decision making processes in schizophrenia, OCD and HD:
  • Why is the antisaccade performance of schizophrenia, OCD and HD patients so poor?

  • Why are latencies more variable and errors greater in schizophrenia, OCD and HD patients than in controls?

  • Is the poor performance of schizophrenia, OCD and HD suffering patients performing the antisaccade task is due to a deficit in the top-down inhibitory control of the erroneous response?

  • Is there a need for an additional STOP decision signal (inhibitory and top-down in nature) to suppress (or inhibit) the erroneous response in the antisaccade task when the correct antisaccade has been expressed first?

  1. Kahramanoglou I, Perantonis S, Smyrnis N, Evdokimidis I, Cutsuridis V. (2008). Modeling the Effects of Dopamine on the Antisaccade Reaction Times (aSRT) of Schizophrenia Patients Lecture Notes in Computer Science (LNCS) 5164, (Springer-Verlag Berlin Heidelberg 2008), 290-299

  2. Cutsuridis V, Kumari V, Ettinger, U. (2014). Antisaccade performance in schizophrenia: A Neural Model of Decision Making in the Superior Colliculus. Front. Neurosci., 8:13. doi: 10.3389/fnins.2014.00013

  3. Cutsuridis V. (2015). Neural Competition via Lateral Inhibition between Decision Processes and Not a STOP Signal Accounts for the Antisaccade Performance in Healthy and Schizophrenia Subjects. Front. Neurosci., 9:5. doi: 10.3389/fnins.2015.00005.

  4. Cutsuridis V. (2017). A neural network model of antisaccade performance of healthy controls and obsessive-compulsive disorder patients. In: International Conference in Cognitive Modelling, Warwick, U.K.

  5. Cutsuridis V. (2017). Behavioral and computational varieties of response inhibition in eye movements. Philos Trans R Soc Lond B, 372: 20160196.

  6. Cutsuridis V, Dunn M, ..., Erichsen J. (2019). ECEM 2019, Alicante, Spain (abstract, talk).

  7. Cutsuridis V. (2019). Modelling cognitive processing of healthy controls and obsessive compulsive disorder subjects in the antisaccade task. In: Cutsuridis V (ed), Multiscale Models of Brain Disorders. Springer-Nature, Switzerland

Microcircuits Models of Memory Formation in Hippocampus in Health and Disease  
How new memories are formed without disrupting existing ones? What are the functional roles for the different classes of inhibitory interneurons in the storage and recall of memories? What is the recall performance of the CA1 microcircuit as a function of partial completion and input pattern presentation period? What temporal strategy of recall is used by the hippocampus?

  1. Cutsuridis V, Cobb S, Graham BP. (2008). Encoding and Retrieval in a CA1 Microcircuit Model of the Hippocampus. Lecture Notes in Computer Science (LNCS) 5164, (Springer-Verlag Berlin Heidelberg 2008), 238–247

  2. Graham BP, Cutsuridis V. (2009). Dynamical Information Processing in the CA1 Microcircuit of the Hippocampus. In: Heinke D, et al. (Eds.): Computational Modeling in behavioral neuroscience: Closing the gap between neurophysiology and behavior. London: Psychology Press, Taylor and Francis Group

  3. Cutsuridis V, Wenneckers T. (2009). Hippocampus, microcircuits and associative memory. Neural Networks, 22(8): 1120-8

  4. Cutsuridis V, Graham BP, Cobb S. (2010). Encoding and retrieval in the hippocampal CA1 microcircuit model. Hippocampus, 20(3): 423-446

  5. Graham BP, Cutsuridis V, Hunter R. (2010). Associative Memory Models of Hippocampal Areas CA1 and CA3. In: Cutsuridis V et al. (eds), Hippocampal Microcircuits: A Computational Modeller’s Resource Book. Springer, USA, 459-494

  6. Cutsuridis V. (2018). Computational Microcircuit Models of Associative Memory In Healthy and Diseased Hippocampus. In: Computational Models of Brain and Behavior, 1st Edition. Edited by Dr Ahmed A. Moustafa. © 2018 John Wiley & Sons, Ltd. Published 2018 by John Wiley & Sons, Ltd.

How is sequence learning of spatial memories achieved in the absence of recurrent connectivity? How different spatial representations are chunked together? How forward and reverse replay of behavioral sequences is accomplished? How phase and rate code of place cells is generated? What functional roles do the various types of inhibitory interneurons play in the encoding, theta retrieval, ripple forward and reverse replay processes of such sequences as well as in the generation of theta phase precession in the hippocampus?

  1. Cutsuridis V, Hasselmo M. (2010). Dynamics and function of a CA1 model of the hippocampus during theta and ripples. Lecture Notes in Computer Science (LNCS) 6352, Springer-Verlag Berlin Heidelberg, pp. 230-240, 2010

  2. Cutsuridis V*, Hasselmo M. (2011). Spatial memory sequence encoding and replay during modeled theta and ripple oscillations. Cognitive Computation, 3: 554-74.

  3. Cutsuridis V, Grahan B.P., Cobb S., Hasselmo M.E. (2011). Bio-inspired models of memory capacity, recall performance and theta phase precession. Proc. IJCNN, 2011 IEEE, pp. 3141-48

  4. Cutsuridis V, Hasselmo M. (2012). GABAergic modulation of gating, timing and theta phase precession of hippocampal neuronal activity during theta oscillations. Hippocampus, 22: 1597-1621

  5. Cutsuridis V. (2018). Models of Rate and Phase Coding of Place Cells in Hippocampal Microcircuits. In: Cutsuridis V, Graham BP, Cobb S, Vida I (eds). Hippocampal Microcircuits: A computational modeller’s resource book, 2nd edition, Springer, USA

In a collaboration with Yiannis Taxidis from CaLTech (USA) we constructed a region CA1 conceptual model of how dendritic and somatic inhibition may collectively contribute to the sharp wave ripple (SWR) generation.

SWRs are population oscillatory patterns in hippocampal LFPs during deep sleep and immobility, involved in the replay of memories acquired during wakefulness. Their exact generation mechanism is still unknown.

Previous computational studies suggested that fast perisomatic inhibition may generate the high frequency ripples (~200 Hz). Others have showed how replay of memories can be controlled by various classes of inhibitory interneurons targeting specific parts of pyramidal cells (PC) and firing at particular SWR phases. Optogenetic studies revealed new roles for interneuronal classes and rich dynamic interplays between them, shedding new light in their potential role in SWRs.

Our conceptual model suggests that sharp wave excitation and basket cell (BC; perisomatic inhibitory cells) recurrent inhibition synchronises BC spiking in ripple frequencies. This rhythm is imposed on bistratified cells (dendritic inhibitory cells), which prevent pyramidal bursting. Axo-axonic (axonic inhibitory cells) and stratum lacunosum/moleculare (distal dendritic inhibitory cells) interneurons are silenced by inhibitory inputs originating in the medial septum. PCs receiving rippling inhibition in both dendritic and perisomatic areas and excitation in their apical dendrites, exhibit sparse ripple phase-locked spiking.

  1. Cutsuridis V, Taxidis V. (2013). Deciphering the CA1 inhibitory circuits in sharp wave ripple complexes. Frontiers in Systems Neuroscience, 7:13, doi: 10.3389/fnsys.2013.00013

In an another collaboration with Motoharu Yoshida (Bochum, Germany) we showed how the interaction of ACh with intrinsic dynamic activity change in region CA1 of the hippocampus switches hippocampal function between encoding and consolidation.

Our toy network model of region CA3 of the hippocampus showed that a high-acetylcholine level during awaking condition supports slow propagation of activity which resembles firing pattern of place cells during running. On the other hand, a low-acetylcholine level during slow-wave-sleep condition supports fast propagating brief firing in the same cells which resembles firing of place cells during replay and pre-play.

These observations suggest that acetylcholine could switch encoding and consolidation dynamics through the modulation of the CAN current and synaptic conductance in the hippocampus.

  1. Saravanan V, Cui A, Gootjes-Dreesbach L, Cutsuridis V, Yoshida M. (2015). Transition between encoding and consolidation/replay dynamics via cholinergic modulation of CAN current: A modelling study. Hippocampus, 25(9): 1052-70. doi: 10.1002/hipo.22429

I've constructed a biophysically detailed model of the whole hippocampal formation to show that the experimentally reported long temporal delays in the DG, CA3 and CA1 hippocampal regions are due to theta modulated somatic and axonic inhibition.

  1. Cutsuridis V. (2015). A computational study on how theta modulated inhibition can account for the long temporal delays in the entorhinal-hippocampal loop. Neurobiology of Learning and Memory, 120: 69-83

In a recent collaboration with Ahmed Moustafa (Western University of Sydney) we started to investigate what breaks down at a network level in Alzheimer's disease.

  1. Moustafa AA, Hassan M, Hewedi D, Garami J, Alashwal H, Zaki N, Seo SY, Cutsuridis V, Angulo SL, Hewedi E, Natesh JY, Herzallah MM, Frydecka D, Misiak B, Salama M, Mohamed W, El Haj M, Hornbeger M. (2017). Genetic Underpinnings in Alzheimer’s disease – a review. Reviews in the Neurosciences, accepted for publication

  2. Cutsuridis V, Moustafa AA. (2017). Computational models of Alzheimer's disease. Scholarpedia, 12(1):32144.

  3. Cutsuridis, V., Moustafa, A. (2017). Multiscale Models of Pharmacological, Immunological and Neurostimulation Treatments in Alzheimer's Disease. Drug Discov Today: Dis Model,

  4. Cutsuridis V, Moustafa A. (2018). Computational model of pharmacological and immunological treatments of Alzheimer’s disease. In: Computational Models of Brain and Behavior, 1st Edition. Edited by Dr Ahmed A. Moustafa. © 2018 John Wiley & Sons, Ltd. Published 2018 by John Wiley & Sons, Ltd.

Synaptic and single neuron models of synaptic plasticity  
How inhibition affects spike timing-dependent plasticity? STDP is a refinement of Hebb's rule, where the timing of presynaptic and postsynaptic spiking determines the sign of plasticity.

STDP shape depends on dendritic location: (1) A symmetric STDP profile centered at 0 ms (largest LTP value) with two distinct LTD windows at about ±20ms in the proximal to the soma dendrite, and (2) an asymmetric one in the distal to the soma dendrite. Bicuculline application revealed that inhibition is responsible for the symmetry of the STDP curve.

Computer simulations led to a number of predictions, such as theta-burst inhibition and burst interspike intervals are the necessary conditions for the asymmetry-to-symmetry transition, the relative timing of inhibition with excitation is an important factor, inhibition can boost LTP in the synapse in the presence of triplets, etc.

  1. Cutsuridis V, Cobb S, Graham BP. (2008). A Ca2+ Dynamics Model of the STDP Symmetry-to-Asymmetry Transition in the CA1 Pyramidal Cell of the Hippocampus. Lecture Notes in Computer Science (LNCS) 5164, (Springer-Verlag Berlin Heidelberg 2008), 627-635

  2. Cutsuridis V, Cobb S, Graham BP. (2009). Modelling the STDP symmetry-to-asymmetry transition in the presence of GABAergic inhibition. Neural Network World, 19(5): 471-81

  3. Cutsuridis V, Cobb S, Graham BP. (2009). How bursts shape the STDP curve in the presence/absence of GABA inhibition. Lecture Notes in Computer Science (LNCS) 5768, Springer-Verlag, 229–238

  4. Cutsuridis V. (2010). Action Potential Bursts Modulate the NMDA-R Mediated Spike Timing Dependent Plasticity in a Biophysical Model. Lecture Notes in Computer Science (LNCS) 6352, Springer-Verlag Berlin Heidelberg, pp. 107–116, 2010

  5. Cutsuridis V. (2011). GABA inhibition modulates NMDA-R mediated spike timing dependent plasticity (STDP) in a biophysical model. Neural Networks, 24(1): 29-42.

  6. Cutsuridis V. (2012). Bursts shape the NMDA-R mediated spike timing dependent plasticity curve: Role of burst interspike interval and GABA inhibition. Cognitive Neurodynamics, 6(5): 421-441

  7. Cutsuridis V. (2013). Interaction of Inhibition and Triplets of Excitatory Spikes Modulates the NMDA-R Mediated Synaptic Plasticity in a Computational Model of Spike Timing Dependent Plasticity. Hippocampus, 23(1): 75-86

  8. Cutsuridis V. (2018). Simplified compartmental models of CA1 pyramidal cells of theta-modulated inhibition effects on spike timing-dependent plasticity. In: Cutsuridis V, Graham BP, Cobb S, Vida I (eds), Hippocampal Microcircuits: A computational modeller’s resource book, 2nd edition, Springer, USA

In an ongoing collaboration Dr. Ausra Saudergiene (Lithuania) we’ve investigated the modulatory effects of ACh on STDP in a detailed CA1 pyramidal neuron model. Our results showed that very low concentrations of ACh does not affect STDP, but moderate and high levels of ACh enhanced LTP and switched LTD to LTP matching the experimental data.

  1. Saudargiene A, Cutsuridis V. Modulatory effects of acetylcholine on spike timing-dependent plasticity in hippocampal CA1 pyramidal neuron. Under revision