Subject study: Medical Image Analysis for Automated Skin Lesion Segmentation
The project will look to define and create robust and efficient image processing techniques that can be applied to wounds with respect to improving the level of information captures and to automate certain measurements. These include segmentation of a variety of wounds with particular reference to pressure ulcers. There are a wide variety of types of pressure ulcers which therefore provides a unique challenge to processing the images. This research is concerned with efficient quantification of lesions, their healing process and ultimately to predict healing time. It is envisaged that a combination of traditional algorithms and machine learning techniques, in particular deep learning methods, will be used.
Candidates should have a 2.1 or above in their undergraduate degree in Computer Science or a related subject.
Please visit our website for more information on eligibility criteria.