Magnetic resonance electrical impedance tomography (MREIT) is a method for reconstructing a three-dimensional image of the conductivity distribution in a target volume using magnetic resonance (MR). In MREIT, currents are applied to the volume through surface electrodes and their effects on the MR induced magnetic fields are analyzed to produce the conductance image. However, current injection through surface electrodes poses technical problems such as the limitation on the safely applicable currents. In this paper, we present a new method called magnetic resonance driven electrical impedance tomography (MRDEIT), where the magnetic resonance in each voxel is used as the applied magnetic field source, and the resultant electromagnetic field is measured through surface electrodes or radio-frequency (RF) detectors placed near the surface. Because the applied magnetic field is at the RF frequency and eddy currents are the integral components in the method, a vector wave equation for the electric field is used as the basis of the analysis instead of a quasi-static approximation. Using computer simulations, it is shown that complex permittivity images can be reconstructed using MRDEIT, but that improvements in signal detection are necessary for detecting moderate complex permittivity changes.

A simple associationist neural network learns to factor abstract rules (i.e., grammars) from sequences of arbitrary input symbols by inventing abstract representations that accommodate unseen symbol sets as well as unseen but similar grammars. The neural network is shown to have the ability to transfer grammatical knowledge to both new symbol vocabularies and new grammars. Analysis of the state-space shows that the network learns generalized abstract structures of the input and is not simply memorizing the input strings. These representations are context sensitive, hierarchical, and based on the state variable of the finite-state machines that the neural network has learned. Generalization to new symbol sets or grammars arises from the spatial nature of the internal representations used by the network, allowing new symbol sets to be encoded close to symbol sets that have already been learned in the hidden unit space of the network. The results are counter to the arguments that learning algorithms based on weight adaptation after each exemplar presentation (such as the long term potentiation found in the mammalian nervous system) cannot in principle extract symbolic knowledge from positive examples as prescribed by prevailing human linguistic theory and evolutionary psychology.

It has been known that people, after being exposed to sentences generated by an artificial grammar, acquire implicit grammatical knowledge and are able to transfer the knowledge to inputs that are generated by a modified grammar. We show that a second order recurrent neural network is able to transfer grammatical knowledge from one language (generated by a Finite State Machine) to another language which differ both in vocabularies and syntax. Representa(cid:173) tion of the grammatical knowledge in the network is analyzed using linear discriminant analysis.

A self-organizing model of language comprehension at syntactic and semantic levels is presented. It has a two-stream architecture aimed at facilitating language acquisition. A head stream forms phrases that denote entities or events. Such phrases are called extended maximal projections (EMPs). An argument stream attaches an EMP as an argument of another EMP. The learning algorithm within each stream is unsupervised, consistent with the assumption that children can learn language without being corrected. The algorithm is based on distribution analysis of words and phrases, and feature extraction from the obtained distribution. The model was trained using 2000 sentences from a corpus that consists of parents' inputs to children, and was tested on 48 novel sentences that were randomly selected from the next 500 sentences. The model yielded correct syntactic structures for 44% of the test set and correct semantic memory structures for 38% of the test set. A basic assumption behind the network design was also tested using statistical analysis.

This paper presents the design and simulation results of a self(cid:173) organizing neural network which induces a grammar from exam(cid:173) ple sentences. Input sentences are generated from a simple phrase structure grammar including number agreement, verb transitiv(cid:173) ity, and recursive noun phrase construction rules. The network induces a grammar explicitly in the form of symbol categorization rules and phrase structure rules. 1 Purpose and related works The purpose of this research is to show that a self-organizing network with a certain structure can acquire syntactic knowledge from only positive (i.e. grammatical) data, without requiring any initial knowledge or external teachers that correct errors. There has been research on supervised neural network models of language acquisi(cid:173) tion tasks [Elman, 1991, Miikkulainen and Dyer, 1988, John and McClelland, 1988]. Unlike these supervised models, the current model self-organizes word and phrasal categories and phrase construction rules through mere exposure to input sentences, without any artificially defined task goals. There also have been self-organizing models of language acquisition tasks [Ritter and Kohonen, 1990, Scholtes, 1991]. Compared to these models, the current model acquires phrase structure rules in more explicit forms, and it learns wider and more structured contexts, as will be explained below. 2 Network Structure and Algorithm The design of the current network is motivated by the observation that humans have the ability to handle a frequently occurring sequence of symbols (chunk) as an unit of information [Grossberg, 1978, Mannes, 1993]. The network consists of two parts: classification networks and production networks (Figure 1). The classification networks categorize words and phrases, and the production networks 28