The persistence of cortical plasticity in the adult can help explain functional recovery after stroke. Computer modeling tools developed to explain the process of early development of sensory systems can be extended to help us relate cortical plasticity to both behavior and to underlying molecular and cellular mechanisms. Computer modeling results suggest a two-phase recovery process, involving immediate alterations in activity patterns caused by the loss of the infarcted neurons ("dynamic plasticity"), followed by true plastic changes as the new activity alters synaptic weights between neurons. Recognition of these two phases suggests that timing of physiotherapy and pharmacotherapy may play an important role in their efficacy.Causal links are nonintuitive in complex systems. In the realm of weather, a rogue current in the mid-Pacific can throw off your vacation plans on the Atlantic seaboard. In the brain, a drug or other intervention can have farreaching and unexpected effects. The computer has proven a valuable tool for permitting predictions to be made in complex systems. This is done by making connections between different levels of organization ( 1 ). In the nervous system, these levels of organization can be identified anatomically and physiologically (Box 1) (2, 3). A major goal of computational neuroscience is to reach across these levels-for example, to explain and predict molecular effects at the network level or network effects at the behavioral level. Creation of these conceptual links is important in neurological disease because pharmacotherapeutic interventions are aimed at the molecular level, whereas clinical outcomes are sought at the functional, behavioral levels.In the early 1950s, computer-science pioneer A.M. Turing developed a series of equations that help explain how the leopard got its spots (4). He noted that a chemical process marked by a difference in reaction rates with diffusing reactants could produce a variety of striped and spotted designs that could be responsible for a wide variety of biological patterns (5). The processes of evoked activity and activity-dependent plasticity present in the central nervous system can be viewed as &dquo;reactions&dquo; that occur at different rates, resulting in the development of patterns such as ocular dominance and orientation columns (6). Given mathematical methods that can explain the development of patterns, it was natural to apply these methods to the problem of recreation of appropriate or inappropriate brain patterns after the disruption produced by a stroke.
Plasticity in the Adult BrainThe pioneering studies of Hubel and Wiesel introduced the concept of a critical period in brain development. They found that appropriate neural connections from the eye either formed during this period or did not form at all, leaving the animal visually impaired. Extrapolation from this result suggested the hypothesis that neural organization of primary cortex was largely fixed at some definable developmental stage, restricting adult plasticity to remote...