Granular neural networks for numerical-linguistic data fusion andknowledge discovery

We present a neural-networks-based knowledge discovery and data mining (KDDM) methodology based on granular computing, neural computing, fuzzy computing, linguistic computing, and pattern recognition. The major issues include 1) how to make neural networks process both numerical and linguistic data in a database, 2) how to convert fuzzy linguistic data into related numerical features, 3) how to use neural networks to do numerical-linguistic data fusion, 4) how to use neural networks to discover granular knowledge from numerical-linguistic databases, and 5) how to use discovered granular knowledge to predict missing data. In order to answer the above concerns, a granular neural network (GNN) is designed to deal with numerical-linguistic data fusion and granular knowledge discovery in numerical-linguistic databases. From a data granulation point of view the GNN can process granular data in a database. From a data fusion point of view, the GNN makes decisions based on different kinds of granular data. From a KDDM point of view the GNN is able to learn internal granular relations between numerical-linguistic inputs and outputs, and predict new relations in a database. The GNN is also capable of greatly compressing low-level granular data to high-level granular knowledge with some compression error and a data compression rate. To do KDDM in huge databases, parallel GNN and distributed GNN will be investigated in the future.     

Euclidean controllable realizations of linear hereditary systems

A realization problem for a class of linear hereditary systems is formulated and state space representations are constructed from inputoutput relations. The resulting hereditary models are shown to be Euclidean controllable, and comparisons are made between these hereditary realizations and other recently developed theories.This research was supported in part by the Air Force Flight Dynamics Laboratory under grant AFOSR-77-3221.     

A habituating blood glucose control strategy for the critically ill

In the intensive care unit patients benefit from being fed and from having well controlled glucose levels. Insulin and glucose infusion serves as manipulated inputs to regulate blood glucose, while glucose infusion serves as a sole nutritional input. In this paper, a model predictive control strategy, based on simultaneously manipulating glucose and insulin infusion, is developed to improve blood glucose regulation in intensive care unit patients. In the short term, glucose infusion is used for tighter glucose control, particularly for disturbance rejection, while, in the long-term (24 h period), glucose infusion is used to meet nutritional needs. The “habituating control” algorithm is proposed and tested against a model predictive control (MPC) strategy that only manipulates insulin. The simulation results indicate that the Habituating MPC strategy outperforms the single input–single output MPC by providing faster setpoint tracking and tighter glucose control for a patient population, and producing less glucose variability while rejecting disturbances in insulin infusion and insulin sensitivity.     

A study on the rate of degradation of the bioabsorbable polymer polyglycolic acid (PGA)

As part of a study to characterise bioabsorbable scaffolds for tissue engineering an investigation has been conducted into the rate of degradation of polyglycolic acid (PGA). This is one of the most commonly used bioabsorbable materials and has been used in sutures since the 60s and more recently in cell scaffolds, drug delivery devices and bone fixation pins. This study looks at the influence that surface-to-volume ratio i.e. thickness of material, has on degradation. By degrading various thicknesses of PGA in a buffer saline solution over 24 days and testing their properties at regular intervals, a knowledge of how surface-to-volume ratio affects degradation was developed. Properties such as weight loss, crystallinity, molecular weight and structural integrity were measured. Results showed that rate of mass loss was dependent on sample thickness but crystallinity, melting point and molecular weight were independent of thickness.     

Adherence to antiretroviral therapy and its impact on clinical outcome in HIV-infected patients.

We analyse data on patient adherence to prescribed regimens and surrogate markers of clinical outcome for 168 human immunodeficiency virus infected patients treated with antiretroviral therapy. Data on patient adherence consisted of dose-timing measurements collected for an average of 12 months per patient via electronic monitoring of bottle opening events. We first discuss how such data can be presented to highlight suboptimal adherence patterns and between-patient differences, before introducing two novel methods by which such data can be statistically modelled. Correlations between adherence and subsequent measures of viral load and CD4+T-cell counts are then evaluated. We show that summary measures of short-term adherence, which incorporate pharmacokinetic and pharmacodynamic data on the monitored regimen, predict suboptimal trends in viral load and CD4+T-cell counts better than measures based on adherence data alone.     

Robust sound onset detection using leaky integrate-and-fire neurons with depressing synapses

A biologically inspired technique for detecting onsets in sound is presented. Outputs from a cochlea-like filter are spike coded, in a way similar to the auditory nerve (AN). These AN-like spikes are presented to a leaky integrate-and-fire neuron through a depressing synapse. Onsets are detected with essentially zero latency relative to these AN spikes. Onset detection results for a tone burst, musical sounds and the DARPA/NIST TIMIT speech corpus are presented.     

A DSP-based mixed-signal waveform generator

As it currently stands, waveform generators are being employed in many diverse areas, such as radar applications, communication systems, simulation, and testing. There is considerable interest in exploring the beacon of opportunity posed by using digital signal processor (DSP)-based systems to replace costly and inadequate conventional waveform generators. The availability of high performance DSPs allows for the design of highly flexible equipment capable of real-time processing. The arbitrary waveform generator introduced in this paper consists of a DSP, a FIFO memory, a video palette, and a PC interface. A program running on the PC creates a digital representation of a waveform according to user parameters and sends it to the DSP's RAM. The DSP algorithmically converts the digital representation to a point-by-point representation of the desired analog signal. The data points are sent to the FIFO buffer, which latches the data to a video palette that serves a three-channel digital-to-analog converter. The system employs a high-speed serial port interface to communicate with a PC, to allow for readily programmed signals. In addition to the production of traditional signals, such as sinusoids and squarewaves, a mathematical tool known as splines is employed to aid in the development of customized arbitrary waveforms. Finally, a picture of our new device is presented in the paper.     

Image processing of electron micrographs of deformed filaments

Electron micrographs of filaments are frequently unsuitable for image processing because they are not perfectly straight. Even minor departures from straightness result in appreciable deterioration of the Fourier transform and analysis of periodicities and subsequent image processing becomes difficult. A procedure is described for recovering the projection of a straight filament from that of an imperfectly straight filament.     

Integrated Computational Materials Engineering (ICME) Approach to Design of Novel Microstructures for Ti-Alloys

In this overview, we present integrated CAPHAD and phase-field modeling with critical experiments to explore a newly discovered, nonconventional, solid–solid phase transformation pathway based on the so-called pseudo-spinodal mechanism. We show that this new transformation pathway offers a new design strategy for Ti alloys with extremely fine and uniform α + β microstructures that could potentially have highly attractive balances of mechanical properties. To broaden the processing window for such a mechanism to operate, we also explore a different nonconventional transformation pathway that involves precursory phase separation. In addition, the variant selection process during the β → α transformation leading to macrozones is investigated and the results could shed light on how to control processing conditions to avoid or reduce microtexture at both the individual β grain level and the overall polycrystalline sample level.     

Non-targeted analysis of tea by hydrophilic interaction liquid chromatography and high resolution mass spectrometry

Tea is the second most consumed beverage in the world and its consumption has been associated with numerous potential health benefits. Factors such as fermentation methods, geographical origin and season can affect the primary and secondary metabolite composition of tea. In this study, a hydrophilic interaction liquid chromatography (HILIC) method coupled to high resolution mass spectrometry in both positive and negative ionisation modes was developed and optimised. The method when combined with principal component analysis to analyse three different types of tea, successfully distinguished samples into different categories, and provided evidence of the metabolites which differed between them. The accurate mass and high resolution attributes of the mass spectrometric data were utilised and relative quantification data were extracted post-data acquisition on 18 amino acids, showing significant differences in amino acid concentrations between tea types and countries. This study highlights the potential of HILIC chromatography combined with non-targeted mass spectrometric methods to provide a comprehensive understanding of polar metabolites in plant extracts.