Page not found

Neuroscience PhD at Wake Forest University at Wake Forest University

Wake Forest University Graduate School » Neuroscience PhD at Wake Forest University

Emilio Salinas

Emilio Salinas
A neural network
My research involves the analysis and modeling of single neurons and neural networks using mathematical methods and computer simulations. I apply these techniques to study how the biophysical properties of neurons determine their electrical responses, how neurons interact to produce functioning neural circuits, and how large neural networks represent, store, and process information. Specific topics of interest are the dynamical properties of spiking networks, the mechanisms by which neurons represent and tranform sensory information, and the way this information is integrated and used to guide behavior. My most recent project centers around the mechanisms that determine perceptual processing speed during a choice task; that is, how fast can a perceptual judgement be made? By tracking how perceptual judgements unfold in time, we can dissociate perceptual and motor performance, and understand how these factors interact to determine a subject's choice behavior.


Tachometric curves revealing perceptual processing speed Tachometric curves revealing perceptual processing speed. Perceptual decision-making during fast choices In collaboration with Terry Stanford, we investigate the neural mechanisms that determine perceptual processing speed (Stanford et al. 2010; Shankar et al. 2011). How long does it take for the brain to establish, for example, whether a visual stimulus is red or green? What are the neural circuits and biophysical processes that place this limit? And how does perceptual performance relate to other brain mechanisms engaged during choice behavior? We will address several objectives in our ongoing and future research. (1) Designing and implementing new variants of our fast-choice task for probing various sensory features and modalities, and for engaging different brain areas. (2) Investigating how indirect factors such as reward, motivation, difficulty, experience, and stimulus statistics bias a subject's choice, and what are the neural correlates of such biases. (3) Comparing fast-choice performance in humans and monkeys. (4) Extending our modeling framework to include not only the new task variants just mentioned, but also results from other laboratories based on different tasks. Development of prefrontal cortex In collaboration with Christos Constantinidis and Terry Stanford, we are using oculomotor choice tasks to investigate how the activity of prefrontal cortical neurons changes during development. The idea is to monitor, in parallel, the behavior of subjects and the responses of PFC neurons around the transition from adolescence to adulthood. We will generate quantitative methods for analyzing the psychophysical and neural data, as well as models that will help us characterize the progressive changes expected to occur in the function of prefrontal circuits.


My overall interest is in computational neuroscience; i.e., understanding how neural circuits work from a computational perspective. What algorithms do they use to process information, and how are these related to the activity recorded experimentally from live cells? How do neurons encode and store sensory information about the world? How is this information retrieved and used to select an appropriate motor response? Below are descriptions, roughly in reverse chronological order, of the projects I have worked on. The accelerated rise-to-threshold model Perception, action and choice A crucial issue is how information about the sensory world is perceived and used to select an appropriate motor action. In the laboratory of Ranulfo Romo I studied mechanisms of perceptual decision-making in the context of tactile categorization and discrimination tasks (Romo and Salinas 1999, 2001, 2003). Later I became interested in the link between perception and action during visual search tasks, and started a fruitful collaboration with Terry Stanford. Our work is unique in that it focuses on fast choices. We can track with high temporal resolution the evolution of a perceptual judgement and its moment-to-moment impact on a subject's choice during a trial (Stanford et al., 2010; Salinas et al. 2010; Shankar et al. 2011). The model that we developed for this process reproduces both behavioral and neuronal data, and is a key tool for thinking mechanistically about the interaction between perception and movement when a quick response is needed. Optimal sensory representation of target distance in owls Optimal sensory representations Models based on the optimal transmission of information have explained many characteristics of early auditory and visual neurons (e.g., V1 orientation preferences). In more central areas, however, it is not obvious what criterion a neuron or a network may be attempting to optimize. I suggested that some sensory representations are constructed to drive the repertoire of output motor behaviors of an organism as efficiently as possible (Salinas 2006a). This novel approach explains the existence of monotonic tuning curves and other features of higher-order sensory responses that would otherwise seem rather mysterious. Part of the effort in this project consisted of devising ways to measure the efficiency of a sensory representation (Salinas and Bentley 2007). Arbitrary visuomotor remapping through gain modulation Functional significance of gain modulation Many types of neuron, under various stimulation and behavioral conditions, have been shown to vary the amplitude of their responses (i.e., their gain) without changing their selectivity (i.e., the order in which they prefer stimuli). For instance, the response of a neuron to an oriented bar may vary depending on whether the subject is paying attention to the bar or not, but the cell's preferred orientation remains constant. I found that this nonlinear mechanism for combining two sources of information (e.g., stimulus orientation and attentional location) has powerful computational properties and may play a crucial role in many functions (Salinas and Thier 2000; Salinas and Sejnowski 2001b), such as arbitrary visuomotor remapping (Salinas 2004a, 2004b), location-invariant object recognition (Salinas and Abbott 1997), and the generation of motor sequences (Salinas 2009), among others. Wave of activity in a recurrent network Biophysical bases of gain modulation A gain-modulated neuron can be thought of as implementing a multiplication of two types of input, one that provides its primary selectivity (e.g., visual stimulus location) and another that regulates its gain (e.g., attention). What is the biophysical basis of such nonlinear interaction? Several mechanisms have been proposed. Larry Abbott and I studied one based on recurrent activity (Salinas and Abbott 1996). Model cell driven by inputs with high or low variability Neural synchrony and correlations Each cortical neuron is driven by the spiking activity of other neurons that provide input to it, and it has long been speculated that whether those inputs arrive independently of each other (i.e., are uncorrelated) or not (i.e., are correlated) is crucial for determining the postsynaptic response. While in Terry Sejnowski's laboratory, I studied the biophysical conditions under which a neuron is sensitive to the synchrony or correlation structure of its input spike trains (Salinas and Sejnowski 2000, 2002). This work led to some interesting speculation (Salinas and Sejnowski 2001a) on the possible functional role of synchrony -- whether it can be used to regulate the communication between neurons rather than the quantities encoded by neuronal responses -- and was the precursor to many recent studies on this subject by other groups. Also, noise correlations are generally thought to be detrimental to the neural code, but this is not necessarily always the case, and some of Ranulfo Romo's data provide an example in which they may actually be beneficial (Romo et al. 2003). The compelled-saccade task Psychophysics and experimental design During my first postdoc, in the laboratory of Ranulfo Romo, I realized that to be able to analyze and interpret correctly the activity recorded from single neurons, it is absolutely crucial to start with the appropriate experimental design; i.e., with the right behavioral task (see Hernandez et al. 1997). Since then, I have always been heavily involved in task design, and in the analysis of the resulting psychophysical data, with my experimental collaborators -- and we have had some important successes (Romo et al. 1998; Stanford et al. 2010; Shankar et al. 2011). Responses of motor and visuomotor neurons in FEF Analysis of raw neuronal responses Understanding what features of a stimulus drive a particular sensory cell and what aspects of the responses (e.g., onset latency, mean firing rate, periodicity of the spikes) encode useful information depends critically on specialized analytical methods. More generally, such quantitative tools are key for relating neuronal activity to behavior. Over the years, I have implemented and developed many such methods based on statistical and probabilistic techniques, typically applying them to recordings from awake behaving monkeys (Salinas et al. 2000; Stanford et al. 2010). Translation invariance as a result of attentional gain modulation Coordinate transformations The position of an object is initially encoded in eye-centered coordinates, that is, according to where its image falls on the retina. But to reach for it, its location must be specified relative to the hand. How is the initial (eye-centered) visual information converted into a (hand-centered) motor command, and how is this correspondence learned? During my PhD work with Larry Abbott, we found important mathematical conditions for solving this problem (Salinas and Abbott 1995) that are also important in relation to attention and to object recognition (Salinas and Abbott 1997). Estimating wind direction using four decoding methods Decoding methods This is a very specific problem: how to reconstruct or read out the characteristics of a stimulus or physical variable given the observed activity of a population of responsive neurons. During my PhD work with Larry Abbott, we developed a more accurate version of the vector method that is still commonly used (Salinas and Abbott 1994).


A neural network for categorization Primary Research Articles Zhou X, Zhu D, Katsuki F, Qi XL, Lees CJ, Bennett AJ, Salinas E, Stanford TR, Constantinidis C (2014) Age-dependent changes in prefrontal intrinsic connectivity. Proceedings of the National Academy of Sciences, USA 111: 3853--3858. [pdf] Zhou X, Zhu D, Qi XL, Lees CJ, Bennett AJ, Salinas E, Stanford TR, Constantinidis C (2013) Working memory performance and neural activity in the prefrontal cortex of peri-pubertal monkeys. Journal of Neurophysiology 110:2648--2660. [pdf] Costello MG, Zhu D, Salinas E, Stanford TR (2013) Perceptual modulation of motor --- but not visual --- responses in the frontal eye field during an urgent-decision task. Journal of Neuroscience 33: 16394--16408. [pdf] Vázquez Y, Salinas E, Romo R (2013) Transformation of the neural code for tactile detection from thalamus to cortex. Proceedings of the National Academy of Sciences, USA 110: E2635--2644. [pdf] Katsuki F, Qi XL, Meyer T, Kostelic PM, Salinas E, Constantinidis C (2013) Differences in intrinsic functional organization between dorsolateral prefrontal and posterior parietal cortex. Cerebral Cortex [Epub ahead of print]. [pdf] Salinas E, Stanford TR (2013) The countermanding task revisited: fast stimulus detection is a key determinant of psychophysical performance. Journal of Neuroscience 33: 5668--5685. [pdf] Vázquez Y, Zainos A, Alvarez M, Salinas E, Romo R (2012) Neural coding and perceptual detection in the primate somatosensory thalamus. Proceedings of the National Academy of Sciences, USA109: 15006--15011. [pdf] Shankar S, Massoglia DP, Zhu D, Costello MG, Stanford TR, Salinas E (2011) Tracking the temporal evolution of a perceptual judgment using a compelled-response task. Journal of Neuroscience 31: 8406--8421. [pdf] Salinas E, Shankar S, Costello MG, Zhu D, Stanford TR (2010) Waiting is the hardest part: comparison of two computational strategies for performing a compelled-response task. Frontiers in Computational Neuroscience 4: 153. [pdf] Stanford TR, Shankar S, Massoglia DP, Costello MG, Salinas E (2010) Perceptual decision making in less than 30 milliseconds. Nature Neuroscience 13: 379--385. [pdf] Salinas E (2009) Rank-order-selective neurons form a temporal basis set for the generation of motor sequences. Journal of Neuroscience 29: 4369--4380. [pdf] Salinas E, Bentley NM (2007) A simple measure of the coding efficiency of a neuronal population. BioSystems 89: 16--23. [pdf] Salinas E (2006) How behavioral constraints may determine optimal sensory representations. PLoS Biology 4: e387. [pdf] Basalyga G, Salinas E (2006) When response variability increases neural network robustness to synaptic noise. Neural Computation 18: 1349--1379. [pdf] Salinas E (2005) A model of target selection based on goal-dependent modulation. Neurocomputing 65-66: 161--166. [pdf] Tiesinga PH, Fellous J-M, Salinas E, José JV, Sejnowski TJ (2004) Inhibitory synchrony as a mechanism for attentional gain modulation. Journal of Physiology (Paris) 98: 296--314. [pdf] Salinas E (2004) Context-dependent selection of visuomotor maps. BMC Neuroscience 5: 47. [pdf] Salinas E (2004) Fast remapping of sensory stimuli onto motor actions on the basis of contextual modulation. Journal of Neuroscience 24: 1113--1118. [pdf] Tiesinga PHE, Fellous J-M, Salinas E, José JV, Sejnowski TJ (2004) Synchronization as a mechanism for attentional gain modulation. Neurocomputing 58-60: 641--646. [pdf] Bentley NM, Salinas E (2004) Bistability in oscillatory cortical modules. Neurocomputing 58-60: 769--774. [pdf] Salinas E (2003) Background synaptic activity as a switch between dynamical states in a network. Neural Computation 15: 1439--1475. [pdf] Romo R, Hernández A, Zainos A, Salinas E (2003) Correlated neuronal discharges that increase coding efficiency during perceptual discrimination. Neuron 38: 649--657. [pdf] Salinas E (2003) Self-sustained activity in networks of gain modulated neurons. Neurocomputing 52-54: 913--918. [pdf] Salinas E, Sejnowski TJ (2002) Integrate-and-fire models driven by correlated stochastic inputs. Neural Computation 14: 2111--2155. [pdf] Salinas E, Sejnowski TJ (2000) Impact of correlated synaptic input on output firing rate and variability in simple neuronal models. Journal of Neuroscience 20: 6193--6209. [pdf] Salinas E, Hernández H, Zainos A, Romo R (2000) Periodicity and firing rate as candidate neural codes for the frequency of vibrotactile stimuli. Journal of Neuroscience 20: 5503--5515. [pdf] Salinas E, Abbott LF (2000) Do simple cells in primary visual cortex form a tight frame? Neural Computation 12: 131--335. [pdf] Romo R, Hernández A, Zainos A, Salinas E (1998) Somatosensory discrimination based on cortical microstimulation. Nature 392: 387--390. [pdf] Salinas E, Romo R (1998) Conversion of sensory signals into motor commands in primary motor cortex. Journal of Neuroscience 18: 499--511. [pdf] Hernández A, Salinas E, García R, Romo R (1997) Discrimination in the sense of flutter: new psychophysical measurements in monkeys. Journal of Neuroscience 17: 6391--6400. [pdf] Zainos A, Merchant H, Hernández A, Salinas E, Romo R (1997) Role of primary somatic sensory cortex in the categorization of tactile stimuli: effects of lesions. Experimental Brain Research 115: 357--360. [pdf] Salinas E, Abbott LF (1997) Invariant visual responses from attentional gain fields. Journal of Neurophysiology 77: 3267--3272. [pdf] Merchant H, Zainos A, Hernández A, Salinas E, Romo R (1997) Functional properties of primate putamen neurons during the categorization of tactile stimuli. Journal of Neurophysiology 77: 1132--1154. [pdf] Salinas E, Abbott LF (1996) A model of multiplicative neural responses in parietal cortex. Proceedings of the National Academy of Sciences, USA 93: 11956--11961. [pdf] Salinas E, Abbott LF (1995) Transfer of coded information from sensory to motor networks. Journal of Neuroscience 15: 6461--6474. [pdf] Salinas E, Abbott LF (1994) Vector reconstruction from firing rates. Journal of Computational Neuroscience 1: 89--107. [pdf] A neural network for categorization Invited Reviews Romo R, Salinas E (2003) Flutter discrimination: neural codes, perception, memory and decision-making. Nature Reviews Neuroscience 4: 203--218. [pdf] Romo R, Hernández A, Salinas E, Brody CD, Zainos A, Lemus L, de Lafuente V, Luna R (2002) From sensation to action. Behavioural Brain Research 135: 105--118. [pdf] Romo R, Hernández H, Zainos A, Brody C, Salinas E (2002) Exploring the cortical evidence of a sensory-discrimination process. Philosophical Transactions of the Royal Society of London B 357: 1039--1051. [pdf] Romo R, Salinas E, Hernández H, Zainos A, Lemus L, de la Fuente V, Luna R (2002) Neural codes for perception. Revista de Neurologia 34: 364--371. [pdf] Salinas E, Sejnowski TJ (2001) Gain modulation in the central nervous system: where behavior, neurophysiology and computation meet. The Neuroscientist 7: 430--440. [pdf] Salinas E, Sejnowski TJ (2001) Correlated neuronal activity and the flow of neural information. Nature Reviews Neuroscience 2: 539--550. [pdf] Romo R, Salinas E (2001) Touch and go: decision mechanisms in somatosensation. Annual Review of Neuroscience 24: 107--137. [pdf] Salinas E, Thier P (2000) Gain modulation: a major computational principle of the central nervous system. Neuron 27: 15--21. [pdf] Romo R, Salinas E (1999) Sensing and deciding in the somatosensory system. Current Opinion in Neurobiology 9: 487--493. [pdf] A neural network for categorization Commentaries Salinas E (2011) Prior and Prejudice (News and Views). Nature Neuroscience 14: 943--945. [pdf] Stanford TR, Salinas E (2010) Clocking perceptual processing speed: From chance to 75% correct in less than 30 milliseconds. Communicative and Integrative Biology 3: 1--3. [pdf] Salinas E (2009) Neuronal communication: a detailed balancing act (News and Views). Nature Neuroscience 12: 372--374. [pdf] Salinas E (2008) So many choices: what computational models reveal about decision-making mechanisms (Previews). Neuron 60: 946--949. [pdf] Salinas E, Romo R (2007) Molecules to remember. Cell 129: 245--247. [pdf] Salinas E (2006) Noisy neurons can certainly compute (News and Views). Nature Neuroscience 9: 1349--1350. [pdf] Salinas E, Romo R (2000) A chorus line (News and Views). Nature 404: 131--132. [pdf] A neural network for categorization Book Chapters Hauser CK, Salinas E (in press) Perceptual decision making. In: Encyclopedia of Computational Neuroscience: SpringerReference (, D Jaeger, R Jung, eds. (Berlin, Heidelberg: Springer-Verlag). [ ] Salinas E (in press) Decision making: overview. In: Encyclopedia of Computational Neuroscience: SpringerReference (, D Jaeger, R Jung, eds. (Berlin, Heidelberg: Springer-Verlag). [ ] Stanford TR, Salinas E (2012) The potential impact of multisensory integration on perceptual decision making. In: The New Handbook of Multisensory Processes, BE Stein, ed., pp. 495--508 (Cambridge, MA: MIT Press). [pdf] Salinas E, Bentley NM (2009) Gain modulation as a mechanism for switching reference frames, tasks and targets. In: Coherent Behavior in Neuronal Networks, K Josic, J Rubin, M Matías, R Romo, eds., pp. 121--142 (New York: Springer). [pdf] Salinas E (2009) Gain modulation. In: Encyclopedia of Neuroscience, LR Squire, ed., volume 4, pp. 485--490 (Oxford: Academic Press). [pdf] Bentley NM, Salinas E (2009) Neural coding of spatial representations. In: Encyclopedia of Neuroscience, LR Squire, ed., volume 6, pp. 117--122 (Oxford: Academic Press). [pdf] Salinas E, Sejnowski TJ (2004) Correlated neural activity: high- and low-level views. In: Computational Neuroscience: A Comprehensive Approach, J Feng, ed., pp. 341--374 (London: Chapman and Hall/CRC). [pdf] Salinas E, Abbott LF (2001) Coordinate transformations in the visual system: how to generate gain fields and what to compute with them. In: Principles of Neural Ensemble and Distributed Coding in the Central Nervous System, M Nicolelis, ed., vol 130, pp 175--190 (Amsterdam: Elsevier, Progress in Brain Research series). [pdf] Abbott LF, Chance FS, Salinas E (2001) Gain modulation: applications and mechanisms. In: Göttingen Meeting of the German Neuroscience Society, N Elsner, ed. (Stuttgart: Georg-Theime-Verlag). [pdf] Salinas E, Opris I, Zainos A, Hernández A, Romo R (2000) Motor and non-motor roles of the cortico-basal ganglia circuitry. In: Brain Dynamics and the Striatal Complex, R Miller, JR Wickens, eds., pp. 237--255 (Harwood Academic Publishers). [pdf] Salinas E, Romo R (1998) Neuronal representations in a categorization task: sensory to motor transformation. In: Computational Neuroscience: Trends in Research, 1998, J Bower, ed., pp. 599--604 (New York: Plenum). [pdf] Salinas E, Abbott LF (1997) Attentional gain modulation as a basis for translation invariance. In: Computational Neuroscience: Trends in Research, 1997, J Bower, ed., pp. 807--812 (New York: Plenum). [pdf] Salinas E, Abbott LF (1995) Decoding vectorial information from firing rates. In: The Neurobiology of Computation, J Bower, ed., pp. 299--304 (Norwell, MA: Kluwer Academic Publishers).