Introduction: new methods in developmental science

This special issue of Developmental Science reflects a new era of methodological advances in tools for use in developmental science. These methods for imaging the structure and function of the brain will help provide insights into both new and classic developmental questions. Yet, other approaches (e.g. animal, computational and genetic methods) remain essential both for constraining the interpretation of data collected with imaging methods, and for informing general theories of behavioral and brain development. A parallel special issue of Developmental Psychobiology to be published this year highlights the importance of converging methodological approaches to the study of developmental science. By linking these two special issues, we hope to broaden the audience for both journals and encourage more cross talk among scientists using human and animal methods/models in the context of developmental science. The 12 papers in this issue cover traditional as well as contemporary methods for assessing functional localization and are based on the basic principles of magnetic resonance imaging (MRI), positron emission tomography (PET), electrophysiology and other techniques. Each of these methods can be contrasted in a number of ways (refer to Table 1). In the most general sense, the methods can be divided into those that provide functional information and those that provide structural information about the brain. Functional imaging methods allow one to measure changes in brain activity associated with simultaneous changes in behavior. For example, eventrelated potentials (ERP), functional magnetic resonance imaging (fMRI), magnetoencephalography (MEG), near infrared spectroscopy/optimal imaging (NIRS), positron emission tomography (PET) and single photon emission computed tomography (SPECT) all are methods used to measure subtle task-induced changes in signals from the brain. In contrast, methods like magnetic resonance imaging (MRI), magnetic resonance spectroscopy (MRS) and diffusion tensor imaging (DTI) are methods used to measure brain structure and chemistry. For example, MRI can be used to measure gross size or volume differences in brain regions while MRS can be used to measure the concentration of cerebral metabolites like N-acetylaspartate (NAA), creatine plus phosphocreatine (Cr) and choline-containing compounds (Cho) that have been related to neuronal loss or damage. Thus MRS can provide additional insight as to why an MRI-based measure of a brain structure may be smaller. Finally, DTI allows for measures of the regularity and myelination of fiber tracts and provides a more precise measure of myelination of fibers than traditional MRI measures of white matter volume. All three of these structural imaging methods can be correlated with behavior, but none involves simultaneous collection of behavior or the capability of measuring brain changes associated with trial-by-trial behavior. Table 1 provides a number of ways in which these methods may be distinguished with very rough qualitative rankings of each. Transcranial magnetic stimulation (TMS) is excluded from the table because it is a special method that looks at the effects of stimulation or of inactivation of a given brain region on task performance in order to infer the functions that require that brain region. This differs from the functional neuroimaging methods covered in this issue (e.g. PET, fMRI, MEG) that enable one to view activity throughout the brain. The article by Moll and colleagues in this issue provides a more specific description of this methodology. The methods are distinguished in Table 1 in terms of relative temporal and spatial resolution, depth of recording (i.e. superficial or deep structure recordings), relative invasiveness, expense, and ease in use with developmental populations. These rankings are very rough estimates and depend on a number of factors. For example, the depth of recording for ERP and MEG are typically thought to be at the more superficial cortical level than other imaging techniques but dipoles can be estimated for deeper structures and greater resolution can be achieved by combining these methodologies with high spatial resolution techniques such as fMRI. Likewise, in the table the temporal resolution of pharmocologic MRI (phMRI), a technique based on fMRI, is ranked low. This is because, even though fMRI has a temporal