530 probes into adult cortical Learning plasticity Avi Karni* and Giuseppe Recent studies of the improvement Bertini? of perceptual as a function of training-perceptual new insights into the neuronal substrates of this type of skill learning in the adult brain. Issues such as the brain, when and under what conditions changes performance learning - have provided where in practice-related occur are under investigation. The results of these studies suggest that a behaviorally relevant degree of plasticity is retained in the adult cortex, even within early, low-level representations in sensory and motor processing streams. The acquisition and retention of skills may share many characteristics with the functional plasticity subserving early-life learning and development. While the specificity of learning provides localization constraints, an important clue to the nature of the underlying neuronal changes is the time course of learning. Addresses *Department of Neurobiology, Brain Research Building, Weizmann Institute of Science, Rehovot, 76100 Israel, and the Department of Neurology, The Haim Sheba Medical Center, Tel Hashomer, 52621 Israel; e-mail: bnkarniQweizmann.weizmann.ac.il tlaboratory of Brain and Cognition, National Institute of Mental Health, Building 49, Room 1860, 49 Convent Drive, Bethesda, Maryland 20892-4415, USA; e-mail: giuseppe@ln.nimh.nih.gov Current Opinion in Neurobiology 1997, 7~530-535 http://biomednet.com/elecref/0959436800700530 0 Current Biology Ltd ISSN 0959-4388 Abbreviation PET positron emission tomography Introduction It is a familiar and intuitive notion that ‘practice makes perfect’; the performance of a given task can dramatically improve with repetition and training. Perceptual learning refers to the robust gains in performance on basic perceptual tasks that are induced by sensory experience and are dependent on practice. It is reasonable to assume that such gains in performance reflect changes in the brain. Moreover, the fact that many perceptual skills, once acquired, are retained over long time intervals suggests that training can induce long-lasting neural changes. Perceptual learning has been the subject of psychophysical studies for over a century [l]. In recent years, the use of behavioral results to generate constraints and predictions on the nature of the neural mechanisms underlying the acquisition and retention of perceptual abilities has gained importance (e.g. [2-4,5*o,6,7,So,9-13]). These studies have helped generate new models of skill learning and memory in the adult brain [14-161. To study perceptual learning, na’ive subjects are trained to perform a task, using a specific set of conditions, until a stable level of performance is attained. Then, stimulus or paradigm variables are manipulated, one at a time, and performance is measured for each new task configuration in the same subjects, in order to define to what extent the newly acquired skill generalizes to untrained conditions. The time course of the improvement in performance (‘when’) has been used to probe the nature and dynamics of training-dependent functional changes in the brain and to gain insights into the possible neuronal mechanisms of plasticity that underlie this type of learning [10,16-20,21”,22-241. The lack of transfer (generalization) of the learning effects across stimulus and task conditions-an important characteristic of perceptual learning often unrecognized in everyday experience-has been used to generate constraints on the possible locus of the learning-related neural changes (‘where’) (reviewed in [5**,9,10]). The specificity of learning for a given parameter of the sensory input implies that only a discrete part (or subset of neurons) within a representational domain in the brain is affected by the experience. There is considerable anatomical and physiological evidence for a hierarchical organization of information processing in sensory systems such that many physical parameters of a sensory input are selectively represented only in low-level processing stages, whereas neurons in higher-order areas respond invariantly to these parameters. Therefore, a parsimonious interpretation of the specificity of perceptual learning is that only levels of representation in which a given parameter is differentially represented will undergo learning-dependent changes; at a level of processing in which neurons respond invariantly, one would expect learning to generalize for that particular parameter. This is not to say that the learning of many types of human skills does not occur in higher-level cortical representations; rather, the proposal is that the lowest level of representation wherein the relevant stimulus parameter is differentially encoded is a locus of learning [!?I. An alternative interpretation is that training reverses the representational hierarchy within a processing stream, perhaps as attention can sharpen and reduce the effective receptive-field size of higher-level cortical neurons. Learning may modify higher-level representations so as to acquire differential responses to parameters that are ‘normally’ the domain of lower-level representations. Such selective changes in higher-order cortical areas could still account for the specificity of learning [9,25,26]. This im- Learning perceptual skills: adult cortical plasticity Karni and Bertini plies, however, that experience-dependent developmental changes and adult skill learning go in opposing directions in terms of the hierarchy within sensory processing streams. The lowest-level plasticity hypothesis maintains a correspondence between developmental plasticity and adult perceptual learning. It is supported by electrophysiological studies of perceptual learning in adult monkeys and cats, and it leads to explicit testable predictions as to the localization of perceptual learning that are accessible to direct electrophysiological and imaging study ([5**,8*,12,26-33,34*,35**,36]; SJ Bolanowski et a/., SOG Neurosci Abstr 1995, 21:1443). This review focuses on human visual skill learning. A theoretical framework Underlying much of current research in perceptual skill learning is the notion that a behaviorally relevant degree of plasticity is retained in the adult cortex [50*,8*,9,11,12,16,32]. A recent paper [P] proposes a general framework for interpreting perceptual as well as motor skill learning. The proposal is that skills are mediated by discrete, potentially long-lasting, experience-driven changes in those sensory or motor representations that are critical for the performance of the task [15]. The locus of these changes is determined by a double constraint: the nature of the training experience (specificity in terms of physical attributes of the training experience and their corresponding neural representations) and the specific requirements of the task to be performed that select those attributes that are critical to performance (task relevancy). The hypothesis is that a conservative principle is at work: learning-dependent changes will occur at the earliest level of processing in which an explicit (in terms of differential neuronal responses) representation of those stimulus parameters that are critical for the performance of the given task is available (the minimal level hypothesis) [5**,8*,12,32,37]. The time course of skill learning, on the other hand, may reflect the properties of a limited repertoire of basic (cellular) mechanisms of neuronal plasticity throughout the adult cortex and would be very similar irrespective of the locus of the learning-dependent changes [P,17]. Related to this conjecture is the proposal that adult skills are subserved by neuronal mechanisms similar to those underlying the shaping and functional maturation of the cortical sensory and motor systems during early life and development [5**,9,12,24,32]. When: ‘fast’ and ‘slow’ learning Several studies have examined the time course of learning in different paradigms [ 17,19,20,34’,38’,39-42]. Karni and Sagi [ 171 have described two stages in the acquisition of improved perception. First, a fast, within session improvement that can be induced by a limited number of trials on a time scale of minutes. Second, slowly evolving, incremental performance gains, triggered by practice but 531 taking hours to become effective. Thus, most gains in performance occurred not during training, but a minimum of 6-8 hours afterwards (latent ‘consolidation’ phase) [ 17,181. Improvements in performance could be observed over the course of S-10 training sessions spaced 1 to 3 days apart. The skill was then retained for months and years. The hypothesis is that fast learning may reflect the setting up of a task-specific processing routine for solving the perceptual problem, while slow learning reflects an ongoing, perhaps structural, modification of basic representations within the processing system [ 171. One can speculate that possible neuronal substrates for fast and slow learning may be rapid receptive field modulation [9,43] and transcription-dependent synaptic consolidation [24], respectively. Both fast and slow learning were found in several recent studies of visual tasks, including Vernier acuity [ZO], masked contrast detection [41], stereofusion [44], and in monocular (although not binocular) pop-out detection [38’]. A latent phase lasting several hours before the effects of training became effective was found in learning an orientation discrimination task [ 191. Incremental, between sessions, performance gains over the course of several days were noted in tactile discrimination learning [34*]. Results from two recent studies of motor learning suggest both a fast and a slow learning phase, with latent consolidation similar to that found in perceptual learning [21”,45]. Fast learning followed by slow learning may, therefore, be the rule for skill learning. The variability in the time course of learning may depend on the design of the learning paradigm rather than on the sensory (or motor) function being explored. Gains in performance across multiple training sessions, ‘slow’ learning, were found in tasks previously known to induce only ‘fast’ learning. This was accomplished by keeping the task difficulty above but near threshold, by avoiding floor effects in the measurement of performance and by giving sufficient training in each session [2,6,22,46,47]. Conversely, careful testing of na’ive subjects at the very beginning of training may reveal fast learning effects in ‘slow’ learning paradigms [42,47,48]. Where: specificity and localization Specificity of learning to basic stimulus parameters has been observed for a variety of visual tasks, and several recent studies have focused on this aspect. For example, the ability to discriminate between stimuli presented at slightly different orientations improves over a period of 23 weeks [19]. The gains are restricted to the location in the visual field where stimuli are presented repeatedly, with new learning needed even when the stimulus is shifted slightly from a trained location. Schoups et a/. [19] found that learning is also highly specific for the orientation of the stimulus, but transferred between eyes (in three of four subjects). On the basis of this pattern of 532 Sensory systems specificity, the authors suggest a localization of the learning effects to visual areas as early as Vl or V2 [19]. 19975). Training on a visual search task in which subjects were asked to identify the ‘odd’ element in an array of distracters resulted in an enduring improvement in performance, with performance becoming independent of the array size [40]. Learning, not surprisingly, was not monocular but was also not specific for the particular stimulus array used in training, a result that suggests that the gains in performance may relate to a better search strategy that is not dependent on the basic features of the stimuli. Training subjects on texture pop-out tasks induces large gains in sensitivity to local, specific orientation gradients [49,50]. In a recent study, electrophysiological recordings from an adult monkey trained on a texture target discrimination task (as in [49]) suggest that the detection of orientation gradients in the visual input by both Vl and V2 neurons may be a locus of the learning, with Vl neurons differentially responding to texture gradients before V2 neurons (G Bertini zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA eta/., Sot Neurosci Abstr 1995, Variability: task and subject factors 21:276; G Bertini et a/., Sot Neurosci Abstr 1996, 22:1614). There is a significant variability in the specificity of the The ability to detect a single element pop-out target can, learning effects both between studies that make use of however, demonstrate some unexpected specificities as similar paradigms and between subjects in the same study well as generalizations of learning. Ahissar and Hochstein [6,19,34*,38*,39,51,57,58]. Moreover, one can readily see [38*] propose that the neural substrates of this learning in the raw published data that in almost all instances in may be located in V2 or up to V4, where top-down which genuine re-1earnGcould be demonstrated for a attentional control is available. new stimulus configuration (specific learning), some of the initial gains do generalize (see, for example, [49]). Matters Visual hyperacuity-discriminating features of the stimuare further complicated by the fact that different authors lus finer than the grain (spacing) of retinal receptors-has consider a similar degree of transfer as either specific or long been used as a tool to study perceptual learning. nonspecific learning [27,34’]. The ability to judge Vernier offsets improves with practice, and has been found to be specific for the orientation of the stimuli, their location in the visual field, and, in many cases, to the trained eye, implicating low-level visual processing stages as a locus of learning [20,22,43,48,51]. Recent studies support the conjecture that a gain in perceptual sensitivity to the basic features of the stimuli underlies learning. The finding of task-specific learning makes it unlikely that improvements in hyperacuity are attributable to a more precise localization of the stimuli or to better fixation or accommodation [52*]. Recordings from occipital cortex provide a neurophysiological correlate of practice-dependent gains in performance on a hyperacuity taskshorter component latencies and larger amplitudes of evoked responses for trained stimuli [53]. A psychophysical study [26] suggests a narrowing of the orientation tuning curves after training. This result supports the possibility of training-dependent adaptation of specific, oriented spatial filters (considered ‘hard-wired’ in the adult) encoding the stimulus patterns. A role for attentional mechanisms in the ‘sharpening’ of the tuning functions, however, is possible [26]. Not all paradigms of perceptual skill learning implicate early stages of sensory processing as the neuronal substrates for perceptual gains. Indeed, one would predict otherwise whenever the relevant aspects of a stimulus are represented at higher levels within the processing stream. In a tactile discrimination task, practice-dependent performance gains transferred to the untrained hand [34*], suggesting that learning may be subserved by neural changes at higher-level, nonlateralized representations. Similarly, training on novel object recognition would not be expected to show retinotopy but rather view specificity ([54-561; E Kobatake et al., Sot Neurosci Abstr 1993, A possible explanation for the variability is that even subtle differences in experimental conditions may introduce significantly different constraints on the processing of the sensory information (A Karni et a/., Sot Neurosci Abstr 1996, 22:1123). Task performance, at least initially, must include a choice of strategies. This can be conceptualized as the selection of the minimum sufficient representations of the sensory information that allows correct performance on a given task [59]. Such a choice can make some representations of the stimulus irrelevant. A recent PET study suggests that even simple orientated gratings produce different activation patterns in extrastriate cortex, depending on the nature of the task [60]. Consequently, a given task can be solved at different levels of representation and using different strategies. Thus, whereas an earlier study found significant transfer of learning to orthogonal orientations in an orientation discrimination task [57], a more recent study found no transfer [19]. In a tactile discrimination task, practice-dependent performance gains transferred to the untrained hand [34’], whereas in a slightly different paradigm, tactile learning was lateralized (SJ Bolanowski et al., Sot Neurosci Abstr 1995, 21:1443). In a recent study [39], no monocularity was found contrary to a previous study reporting significant monocularity in three subjects training in a similar texture target discrimination task [49]. In the later study, however, the authors failed to make explicit that learning was also not retinotopicthe learning effects transferred even across hemifields in all but one subject, which suggests that learning occurred at a nonretinotopic visual representation ([39]; AA Schoups, GA Orban, Perception 1995, 248383). One does not expect Learning monocular present. learning when no retinotopy of learning is Even within the same study, ‘loose’ behavioral constraints may result in different subjects learning to perform ‘different tasks’ ([38*,51]; A Karni et a/., Sot Neurom’ A&r 1996, .22:1123). The challenge is to fine-tune the paradigms so as to reduce variability and to identify the crucial variables of the task that, when manipulated, can produce substantial differences in the quality (specificity) of the learning. perceptual skills: adult cortical plasticity Karni and Bertini 533 retention of skills are mediated by a limited repertoire of neuronal mechanisms of experience-dependent plasticity subserving discrete changes within basic sensory and motor representations in the adult cortex. Insights gained from perceptual skill learning paradigms are of major relevance to understanding the neurobiology of learning and memory, as well as to generating new strategies for the rehabilitation of skills in individuals with neurological deficits [35”,67*,68,69”,70**,71]. Acknowledgements The authors thank P DeWeerd, B Jagadeesh and LG comments on the manuscript and helpful discussions. Ungerleider for Task relevancy An important issue concerns the minimal necessary conditions that must be satisfied for adult perceptual learning to occur. Is repeated exposure, per se, sufficient to trigger learning? There is evidence that functional plasticity is gated by nonsensory mechanisms during development and adulthood in animals [61-64], as well as during adult human perceptual learning [5”,49,50,65,66]. For example, subjects presented with stimuli containing two texture targets in two different locations of the visual field, but instructed to discriminate only one, did not improve their performance for the irrelevant target [49,66]. Training on evaluating the brightness of line-elements did not improve subjects’ ability to judge their orientation [65]. Learning to detect a pop-out target in a matrix of background elements did not generalize to the ability to judge the global shape of the stimulus array [SO]. Evidently, nonsensory (top-down) mechanisms control adult perceptual learning. Nevertheless, what is learned may not be determined explicitly; subjects practicing a texture target discrimination task improved on the detection of specific orientation gradients but not on shape discrimination ([5**,49,66]; G Bertini et a/., Sot Neurosci Abstr 1995, 21:276; G Bertini et al., Sot Neurosci Abstr 1996, 22:1614). That learning is controlled by top-down mechanisms does not necessarily entail that explicit performance feedback is required. Trial-by-trial signaling of response correctness can be of some help in some conditions [4], but it has often been reported that perceptual learning can occur without it [47-49]. It is possible that subjects make use of an ‘internal’ feedback signal, whereby reinforcement is provided by a degree of confidence of ‘being right’ [25]. Whereas explicit performance feedback is not necessary, initial training with above-threshold stimuli may be required to trigger and sustain the changes in the specific sensory channels recruited to solve the task [5”,6,17]. Conclusions There is compelling behavioral and neurophysiological evidence suggesting that the specifics of a given task determine the type of representation and level of processing in which learning effects will occur in the adult brain. A reasonable working hypothesis is that the acquisition and References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as: . .. of special interest of outstanding interest 1. Gibson EJ: Principles of Perceptual Learning and Development. New York: Appleton-Century-Crofts; 1969. 2. Ramachandran VS. Braddick 0: Orientation-specific stereopsis. Perception 1973, 2:371-376. 3. Fiorentini A, Berardi N: Learning in grating waveform discrimination: specificity for orientation and spatial frequency. Vision Res 1981, 21 :1149-l 158. 4. Ball K, Sekuler R: Direction-specific improvement discrimination. Vision Res 1967, 27:953-965. learning in in motion 5. .. Kami A: The acquisition of perceptual and motor skills: a memory system in the adult human cortex. Cogn Brain Res 1996, 539-46. . A theory tor a memory system for sklll iearnmg in the adult cortex is presented. The main conjecture (minimal level hypothesis) is that the plastic changes underlying the acquisition and retention of a skill occur at the earliest levels within the processing streams capable of representing those stimulus parameters that are relevant for successful behavioral performance. A parallel between developmental and adult cortical plasticity is discussed, as well as the issues of behavioral ‘gating’ of the experience-dependent functional changes and of the significance of their time course. It is proposed that basic principles regulating perceptual leaming may not be different from those underlying the improvements in performance of motor tasks. 6. O’Toole AJ, Kersten DJ: learning to see random-dot stereograms. Perception 1992, 21~227-243. 7. Nothdurft HC: The conspicuousness contrast Spat Vis 1993, 7~341-363. of orientation and motion 6. Gaffan D: Associative and perceptual learning and the concept of memory systems. Cogn Brain Res 1996, 5:69-60. The author reviews recent developments in memory research, focusing in particular on experimental models of nonhuman primate memory impairments. He proposes that plasticity is a general feature of neural systems. The special role played by different memory systems arise from the multimodal input patterns, rather than from intrinsic properties of specific structures. Thus, the neurobiological mechanisms responsible for associative memory should be similar to those underlying plasticity in other domains, such as perceptual systems. Two experiments are described, whose results support a ‘physiological’ hypothesis of an expansion of the cortical representation of the stimuli used during training of perceptual tasks. 9. Gilbert CD: Early perceptual 1994, 91 :1195-l 197. learning. Proc Nat/ Acad Sci USA 10. Sagi D, Tanne D: Perceptual Neurobiol 1994, 4:195-l 99. learning: learning to see. Gun Opin 11. Gilbert CD: Learning. Neuronal dynamics and perceptual learning. Curr Biol 1994, 4:627-629. 12. Merzenich MM, Sameshima K: Cortical plasticity and memory. Curr Opin Neurobiol 1993, 3:167-l 96. 13. Ramachandran VS: Behavioral and magnetoencephalographic correlates of plasticity in the adult human brain. Proc Nat/ Acad Sci USA 1993. 90:10413-10420. 534 Sensory 14. Mishkin M, Malamut B, Bachevalier J: Memories and habits: two neural systems. In The Neurobiology of Learning and Memory. Edited by Lynch G, McGaugh JL, Weinberger NM. New York: Guilford Press; 1984:65-88. 15. Squire LR: Memory and Brain. New York: Oxford University Press; 1987. 16. Salmon DP, Butters N: Neurobiology of skill and habit learning. Curr Opin Neurobiol 1995, 5:184-l 90. 17. Kami A, Sagi D: The time course of learning a visual skill. Nature 1993, 365:250-252. 36. Ungerleider LG: Functional brain imaging studies of cortical mechanisms for memory. Science 1995, 270:769-775. 16. Karni A, Tanne D, Rubenstein BS, Askenasy JJ, Sagi D: Dependence on REM sleep of overnight improvement perceptual skill. Science 1994, 265:679-682. 37. Gibson EJ: Improvement in perceptual judgement as a function of controlled practice or training. Psycho/ Bull 1953, 50:401431. systems Schoups AA, Vogels R, Orban GA: Human perceptual learning in identifying the oblique orientation: retinotopy, orientation specificity and monocularity. J Physiol (Land) 1995, 483:797810. 20. Fahle M: Human pattern recognition: parallel processing perceptual learning. Perception 1994, 23:41 l-427. and 21. Brashers-Krug T, Shadmehr R, Bizzi E: Consolidation in human .. motor memory. Nature 1996, 382:252-254. This study documents a consolidation phase of several hours duration in a paradigm of human motor skill learning. The data support the conjecture that similar neuronal mechanisms underlie memory consolidation in different brain systems. 22. Poggio T, Fahle M, Edelman S: Fast perceptual hyperacuity. Science 1992, 256:1018-l 021. 23. Aronson L, Rosenhouse J, Podoshin L, Rosenhouse G, Zanutto SB: Pitch perception in patients with a multi-channel cochlear implant using various pulses width. Med Prog Technol 1994, 20~43-51. 24. Dudai Y: Consolidation: fragility on the road to the engram. Neuron 1996, 17~367-370. 25. Mollon JD, Danilova MV: Three remarks Spat Vis 1996, IO:51 -58. 26. Saarinen J, Levi DM: Perceptual learning in vernier acuity: what is learned? Vision Res 1995, 35:519-527. 27. Recanzone GH, Jenkins WM, Hradek GT, Merzenich MM: Progressive improvement in discriminative abilities in adult owl monkeys performing a tactile frequency discrimination task. J Neurophysiol 1992, 67:1015-l 030. 28. Recanzone GH, Merzenich MM, Jenkins WM, Grajski KA, Dinse HR: Topographic reorganization of the hand representation in cortical area 3b owl monkeys trained in a frequency-discrimination task. J Neorophysiol 1992, 67:10311056. 30. 31. 38. . Ahissar M, Hochstein S: learning pop-out detection: specificities to stimulus characteristics. Vision Res 1996, 36:3487-3500. The authors examined the specificity of learning a simple pop-out detection task. Although learning was specific for target position and orientation, generalizations were found for expanded stimuli, mirror image orientations and transferred across midline to homologous positions. Intermediate cortical levels of visual processing are suggested as a locus of learning. 39. Schoups AA, Orban GA: Interocular transfer in perceptual learning of a pop-out discrimination task Proc Nat/ Acad Sci USA 1996, 93~7358-7362. 40. Sireteanu R, Rettenbach R: Perceptual learning in visual search: fast, enduring, but non-specific Vision Res 1995, 35:20372043. 41. Polat U, Sagi D: Spatial interactions in human vision: from near to far via experience-dependent cascades of connections. Proc Nat/ Acad Sci USA 1994, 91 :1206-l 209. 42. Vaina LM, Sundareswaran V, Harris JG: Learning to ignore: psychophysics and computational modeling of fast learning of direction in noisy motion stimuli. Brain Res Cogn Brain Res 1995, 2:155-l 63. 43. Kapadia MK, Gilbert CD, Westheimer G: A quantitative measure for short-term cortical plasticity in human vision. J Neurosci 1994, 14~451-457. 44. Kovacs I, Julesz B: Maturational windows and adult cortical plasticity. In Maturational Windows and Adult Cortical Plasticity. Edited by Julesz B, Kovacs I. Reading: Addison-Wesley; 1995:X1XXXVI. 45. Karni A, Meyer G, Jezzard P, Adams MM, Turner R, Ungerleider LG: Functional MRI evidence for adult motor cortex plasticity during motor skill learning. Nature 1995, 377:155158. 46. Fendick M, Westheimer G: Effects of practice and the separation of test targets on fovea1 and peripheral stereoacuity. Vision Res 1983, 23:145-l 50. 47. Fahle M, Edelman S: Long-term learning in vernier acuity: effects of stimulus orientation, range and of feedback. Vision Res 1993, 331397-412. 48. McKee SP, Westheimer G: Improvement in vernier acuity with practice. Percept Psychophys 1978, 24:258-262. 49. Karni A, Sagi D: Where practice makes perfect in texture discrimination: evidence for primary visual cortex plasticity. Proc Nat/ Acad Sci USA 1991, 88:4966-4970. 50. Ahissar M, Hochstein S: Attentional control of early perceptual learning. Proc Nat/ Acad Sci USA 1993, SO:571 8-5722. 51. Beard BL, Levi DM, Reich LN: Perceptual vision. Vision Res 1995, 35:1679-l 690. learning in visual on perceptual learning. Recanzone GH, Merzenich MM, Jenkins WM: Frequency discrimination training engaging a restricted skin surface results in an emergence of a cutaneous response zone in cortical area 3a. J Neurophysiol 1992, 67:1057-l 070. Recanzone GH, Merzenich MM, Schreiner CE: Changes in the distributed temporal response properties of SI cortical neurons reflect improvements in performance on a temporally based tactile discrimination task. J Neurophysiol 1992, 67:1071-1091. Recanzone GH, Schreiner CE, Merzenich MM: Plasticity in the frequency representation of primary auditory cortex following discrimination training in adult owl monkeys. J Neurosci 1993, 13:87-l 03. 32. Singer W: Development and plasticity of cortical processing architectures. Science 1995, 270~758-764. 33. Pascual-Leone A, Torres F: Plasticity of the sensorimotor cortex representation of the reading finger in Braille readers. Brain 1993, 116:39-52. 34. . Sadato N, PascuaCLeone A, Grafman J, lbanez V, Deiber MP, Dold G, Hallett M: Activation of the primary visual cortex by Braille readina in blind subiects. Nature 1996. 380:526-528. The possibility that the human visual cortex is capable of reorganizing to process nonvisual, sensorimotor information is explored using PET. Blind subjects showed activation of primary and secondary visual cortical areas while performing a variety of tactile discrimination tasks, whereas normal controls showed no activation. A Braille reading task by the blind produced the strongest activation. of a 19. 29. 35. .. Sathian K, Zangaladze A: Tactile learning is task specific but transfers between fingers. Percept Psychophys 1997, 59:119126. Subjects were trained to discriminate differences between gratings, by scanning the surface with one fingerpad. The roughness of the gratings was manipulated by varying either the groove or ridge width. The ability to discriminate differences in the groove width does not transfer to ridge width discrimination. Learning, however, transferred from the trained to the untrained fingers suggesting SII and higher-order areas as a locus of plasticity. 52. . learning in parafoveal Fahle M, Morgan M: No transfer of perceptual learning between similar stimuli in the same retinal position. Gun Biol 1996, 6:292-297. Two different hyperacuity tasks (a three-dot Vernier and a three-dot bisection) were used. Stimuli in the two conditions are almost identical, yet the improvement in performance observed in either of the tasks did not transfer to the other. zyxwvutsrqpon 53. Skrandies W, Lang G, Jedynak A: Sensory thresholds and neurophysiological correlates of human perceptual learning. Spat Vis 1996, 9~475-489. 54. Moses Y, Ullman S, Edelman S: Generalization to novel images in upright and inverted faces. Perception 1996, 25:443-461. Learning perceptual skills: adult cortical 55. Logothetis NK, Pauls J: Psychophysical and physiological evidence for viewer-centered object representations in the primate. Cereb Cortex 1995, 5:270-200. 66. 56. Bulthoff HH, Edelman SY, Tan MJ: How are three-dimensional objects represented in the brain? Cereb Cortex 1995. 5:247260. 67. . 57. Vogels R, Orban GA: The effect of practice on the oblique effect in line orientation judgments. \/ision Res 1905, 25:16791607. 50. Fahle M, Henke-Fahle S: Interobserver variance in perceptual performance and learning. Invest Ophthalmol Vis Sci 1996, 37069077. 60. 59. Ullman S: Sequence seeking and counter streams: a computational model for bidirectional information flow in the visual cortex. Cereb Cortex 1995, 51-l 1. 69. .. 60. Orban GA, DuPont P, Vogels R, Bormans G, Mortelmans L: Human brain activity related to orientation discrimination tasks. fur J Neurosci 1997, 9:246-259. 61. Singer W: Central control of developmental plasttcity in the mammalian visual cortex. Vision Res 1905, 25:309-396. 62. Singer W, Tretter F, Yinon U: Evidence for long-term functional plasticity in the visual cortex of adult cats. J Physiol (Land) 1982, 3241239-240. 63. Pasternak T, Movshon JA, Merigan WH: Creation of direction selectivity in adult strobe-reared cats. Nature 1961, 292034. 036. 64. Juliano SL, Ma W, Eslin D: Cholinergic depletion prevents expansion of topographic maps in somatosensory cortex. Proc Nat/ Acad Sci USA 1991, 80:700-704. 65. Shiu LP, Pashler H: Improvement in line orientation discrimination is retinally local but dependent on cognitive set Percept Psychophys 1992, 52:502-500. plasticity Karni and Bertini 535 Karni A, Sagi D: A memory system in the adult visual cortex. In Maturational Windows and Adult Cortical Plasticity. Edited by Julesz B, Kovacs 1. Reading: Addison-Wesley; 1995:95-l 09. Levi DM, Polat U: Neural Proc Nat/ Acad Sci USA An important demonstration that improvements of the weak eye degree of monocular learning is between eyes were found. plasticity in adults with amblyopia. 1996, 93:6030-6034. perceptual training can result in long-term in adults with amblyopia. Although some indicated by the data, significant transfers Nudo RJ, Wise BM, SiFuentes F, Milliken GW: Neural substrates for the effects of rehabilitative training on motor recovery efter ischemic infarct Science 1996, 272:1791-l 794. Merzenich MM, Jenkins WM, Johnston P, Schreiner C, Miller SL, Tallal P: Temporal processing deficits of language-learning impaired children ameliorated by training. Science 1996, 271~77-01. A theoretical foundation for the effects of perceptual training in temporal processing of sensory information on language skills (see [70”]). 70. .. Tallal P, Miller SL, Bedi G, Byma G, Wang X, Nagarajan SS, Schreiner C, Jenkins WM, Merzenich MM: Language comprehension in language-learning impaired children improved with acoustically modified speech. Science 1996, 271:01-04. On the basis of previous behavioral data suggesting temporal processing deficits-the discrimination of fast changes in sensory input -as a substrate of learning disabilities, the authors devised learning paradigms, games, to improve perceptual processing in children with language/learning disabilities using methodological approaches that were shown to be effective for inducing adult perceptual learning. An intriguing and important finding is the transfer of the training effects to real life performance (see [69’*1). 71. Taub E, Cargo JE: Increasing behavioral plasticity following CNS damage in monkeys and man. In Maturational Windows and Adult Cortical Plasticity. Edited by Julesz B, Kovacs I. Reading: Addison-Wesley; 1995:201-215.