Use this URL to cite or link to this record in EThOS:
Title: Automated shape analysis and visualization of the human back
Author: Twumasi, Bright Osei
ISNI:       0000 0004 2682 2846
Awarding Body: Bournemouth University
Current Institution: Bournemouth University
Date of Award: 2007
Availability of Full Text:
Access from EThOS:
Access from Institution:
Spinal and back deformities can lead to pain and discomfort, disrupting productivity, and may require prolonged treatment. The conventional method of assessing and monitoring tile de-formity using radiographs has known radiation hazards. An alternative approach for monitoring the deformity is to base the assessment on the shape of back surface. Though three-dimensional data acquisition methods exist, techniques to extract relevant information for clinical use have not been widely developed. Thi's thesis presentsthe content and progression of research into automated analysis and visu-alization of three-dimensional laser scans of the human back. Using mathematical shape analysis, methods have been developed to compute stable curvature of the back surface and to detect the anatomic landmarks from the curvature maps. Compared with manual palpation, the landmarks have been detected to within accuracy of 1.15mm and precision of 0.8111m.Based on the detected spinous process landmarks, the back midline which is the closest surface approximation of the spine, has been derived using constrained polynomial fitting and statistical techniques. Three-dimensional geometric measurementsbasedon the midline were then corn-puted to quantify the deformity. Visualization plays a crucial role in back shape analysis since it enables the exploration of back deformities without the need for physical manipulation of the subject. In the third phase,various visualization techniques have been developed, namely, continuous and discrete colour maps, contour maps and three-dimensional views. In the last phase of the research,a software system has been developed for automating the tasks involved in analysing, visualizing and quantifying of the back shape. The novel aspectsof this research lie in the development of effective noise smoothing methods for stable curvature computation; improved shape analysis and landmark detection algorithm; effective techniques for visualizing the shape of the back; derivation of the back midline using constrained polynomials and computation of three dimensional surface measurements. �
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: Medicine and Health