 |
||
          |
This line of research includes 2 projects underway and 1
with preliminary data from M.Sc. study for which support is being sought. Generally this
line of research is aimed at adding value to existing instruments associated with
anaesthesia or developing new measurement systems that can be integrated into future
decision support systems. One Ph.D. project on improving blood gas machine reliability is
now complete. The methodologies for considering electrode responses and their
classification using PCA and a novelty detection algorithm for detecting calibration
failure named ADDaM have been established. Another project on analysis of
physiological information logged by the Recall system at the Manchester Royal Infirmary is
entering its final year. Artefact rejection algorithms have now been shown to given
suitable performance. A project based on preliminary data on the classification of chest
sounds is proposed for the detection of respiratory changes during anaesthesia and other
areas of work including respiratory medicine, health and safety and sleep apnoea. The
approach to be used is a dual hybrid system with pre-classifiers supervised by a
diagnostic neural classifier to optimise the use of pathological data and separates the
problem of feature extraction for different sounds from the classification problem.
For more information see www.psrg.man.ac.uk Research Courses Clinical Radiology Industrial liaIson Contact Us Join Us Search Home Contact
the Website Administrator
with comments or queries about this website. |
|