Use this URL to cite or link to this record in EThOS:
Title: Processing and analysis of foetal phonocardiographic signals
Author: McDonnell, James Thomas Edward
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 1990
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
There has been renewed interest in foetal phonocardiography since the recent advent of wide-bandwidth (0.5-250 Hz) phono-transducers which yield higher signal-to-noise ratios than were hitherto obtainable. With this passive and non-passive transducer there is the potential to realize the goal of long-term continuous foetal heart monitoring. In order to achieve this goal, signal processing and analysis methods are required to automatically extract from the phonocardiogram (PhCG) the beat-to-beat temporal parameters of foetal cardiac function. The research detailed in this thesis investigates the temporal and spectral morphology of the princicpal heart sounds, and proposes a method which detects and identifies the first and second heart sounds in the PhCG on a bcat-to-beat basis. Examination of the princicpal heart sounds in 252 PhCGs, each of 4.1 seconds duration, from 19 subjects has found that wide variations exist in the temporal morphology of these sounds. It was also observed that the PhCG is susceptible to persistent contamination by many adventitious sounds. These latter sounds often coalesce with the principal heart sounds and obscure the end points of both. Second heart sounds were noted to be subject to sudden and severe attenuation, but were more morphologically stable than first heart sounds. The frequency analysis of the principal heart sounds was performed using Fourier transformation, a high-resolution parametric technique based on the Burg algorithm. In all the 741 principal heart sounds examined from 12 subjects, it was observed that the majority of the energy in their spectra is concentrated in the previously unmonitored 20-40 Hz band. In this sample set of first and second heart sounds, it was found that each set of values docs not produce a spectral signature with which it may be uniquely associated. Results arc presented to substantiate this observation. A procedural knowledge-based system (KBS) was developed to process and analyse the PhCG with the objective of detecting and identifying individual first and second heart sounds. In this the KBS has been successful. The KBS analyses PhCGs over a wide range of FHRs (80-220 bpm), irrespective of either short-term variabiity in FHR or contamination by adventitious sounds. Results are presented which illustrate the performance of this system over the spectrum of PhCG types.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available