Are you David P. Burke?

Claim your profile

Publications (3)1.71 Total impact

  • D P Burke · A M de Paor
    [Show abstract] [Hide abstract] ABSTRACT: We present an empirical model of the electroencephalogram (EEG) signal based on the construction of a stochastic limit cycle oscillator using Ito calculus. This formulation, where the noise influences actually interact with the dynamics, is substantially different from the usual definition of measurement noise. Analysis of model data is compared with actual EEG data using both traditional methods and modern techniques from nonlinear time series analysis. The model demonstrates visually displayed patterns and statistics that are similar to actual EEG data. In addition, the nonlinear mechanisms underlying the dynamics of the model do not manifest themselves in nonlinear time series analysis, paralleling the situation with real, non-pathological EEG data. This modeling exercise suggests that the EEG is optimally described by stochastic limit cycle behavior.
    No preview · Article · Nov 2004 · Biological Cybernetics
  • Source
    David P. Burke · Annraoi M. de Paor
    [Show abstract] [Hide abstract] ABSTRACT: This paper presents results of a non-linear study of the human electroencephalogram to establish the feasibility of extracting non-stationary information associated with internal or external events and stimuli. By invoking chaotic time series analysis techniques, short-term predictions are made on the attractor. Comparisons with the real evolution of the EEG could in principle yield stimulus-related information.
    Preview · Article · Jan 2001
  • Source
    Jane Courtney · David P Burke · Annraoi M De Paor
    [Show abstract] [Hide abstract] ABSTRACT: The standard method of human gait analysis in use in gait laboratories today invariably involves marker-based motion tracking systems. Although somewhat effective, these methods require accurate placement of awkward external markers. We report on an enhanced approach being researched and developed at the National Rehabilitation Hospital, Dublin based on marker-free motion tracking incorporating advanced digital image processing techniques. Introduction Many therapists and doctors rely on gait analysis to monitor efficacy in patient treatment, allowing the results of various types of gait therapy to be quantified and therefore more accurately compared and contrasted. The first recorded attempts to understand the process of animal movement date back to Aristotle circa 350 BC [1]. Aristotle formulated theories on the control of movement and even suggested a theoretical experiment to measure human gait: "…if a man were to walk alongside a wall (with a reed dipped in ink attached to his head) the line traced would not be straight but zigzag because it goes lower when he bends and higher when he stands upright…" Although science has advanced considerably since Aristotle's time, in particular by the now ubiquitous video recording methods, a new problem has arisen: To get an accurate trace of the patient's locomotion it is necessary to use some kind of markers on the joint centres. This is often achieved by externally attached markers, which reflect ultra-violet light into purposed built cameras. This method, however suffers from a number of issues. Accurate placement of markers, which must stay in position throughout data recording, is critical to the accuracy of the derived clinical indices. Also, markers can restrict the normal gait of a patient as can the feeling of self-consciousness while walking in under garments. Marker free gait analysis tries to alleviate some of these problems. It is still a relatively new concept and although there are a few basic commercially available systems e.g. [3], marker-free gait analysis is still very much in the development stages. We present a report on our findings from analysis and development of such a system in conjunction with the medical staff at the National Rehabilitation Hospital. We present the methods used, followed by results of the application. We conclude with a discussion on future goals. Image Processing Methods We have developed a system based on real-time analysis of acquired video data. This data manifests itself as a stream of matrices of triplets where each matrix location corresponds directly with a pixel location in the image and each triplet describes the relative intensity of the primary colours red, green, and blue. We have restricted ourselves to the problem in the sagittal plane and concentrated on describing the motion of the calf and thigh thus affording us accurate tracking of the knee. Briefly, Measurement in Biomedicine ● J. Courtney, D. P. Burke, A. M. de Paor 12 our approach involves a four step pre-processing method followed by a model fitting exercise. The initial steps employ a combination of background subtraction, thresholding, edge detection, and application of the Chamfer distance transform. We outline the concepts involved in each step and conclude with a description of our model-fitting algorithm. Background Subtraction Conceptually this is probably the simplest operation we perform: the background image is sampled and subtracted from each successive frame. This is similar to the classic blue-screening 1 method used in television production, although in our case we assume a very dark background colour.
    Full-text · Article · Jan 2001