Use this URL to cite or link to this record in EThOS:
Title: Bone ageing and osteoporosis : automated DXA image analysis for population imaging
Author: Farzi, Mohsen
ISNI:       0000 0004 7657 3892
Awarding Body: University of Sheffield
Current Institution: University of Sheffield
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
Osteoporosis is an age-associated bone disease characterised by low bone mass. The consequent fragility fractures with increased follow-up mortality and morbidity underlie the clinical significance of osteoporosis in public health. However, current diagnostic criteria using bone mineral density (BMD) at the femoral neck at most can identify half of the fragility fractures, and thereby the ability to provide new metrics capturing the bone strength beyond neck BMD remains of interest in osteoporosis research. This study aims to, first, quantify pixel BMD at anatomically corresponding locations in the femur; second, model the evolution of spatial BMD patterns with ageing; and third, characterise how trabecular and cortical bone arrangements change at different stages of osteoporosis progression. To construct the atlas, a novel cross-calibration procedure is proposed to integrate data from different DXA manufacturers into an amalgamated largescale dataset (n > 13000). A new technique, termed region free analysis (RFA), is proposed to eliminate morphological variation between scans by warping each image into a reference template. This image warping establishes a correspondence between pixel coordinates that allows modelling pixel BMD evolution with ageing using smooth quantile curves. Given access to largescale datasets, automatic quality control of DXA scans has been identified as an emerging challenge to the community for which an unsupervised, nondistortion-specific, opinion-free framework was proposed. The developed atlas usefully added to our understanding of spatial BMD patterns and their relationship with osteoporosis. The concept of osteoporosis progression is introduced by proposing bone age as the age at which an individual bone map best fits the constructed atlas. Normalising BMD maps for bone age, local fracture-specific patterns were identified. The proposed framework in this thesis constitutes a first step toward modelling osteoporosis progression to identify better bone-based risk factors for prediction of fragility fractures.
Supervisor: Wilkinson, J. Mark ; Frangi, Alejandro F. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available