Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.498868
Title: Active pixel sensors for breast biopsy analysis using X-ray diffraction
Author: Bohndiek, Sarah Elizabeth
Awarding Body: University of London
Current Institution: University College London (University of London)
Date of Award: 2008
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Breast cancer diagnosis currently requires biopsy samples to be analysed by a histopathologist a time consuming, highly specialised process. X-ray diffraction is a quantitative technique that can distinguish between healthy and diseased breast biopsy samples using the change in proportions of fat and fibrous tissue that occurs when cancer invades. A semi-automated breast biopsy analysis system based on X-ray diffraction could yield a faster patient diagnosis. Recording X-ray diffraction patterns is a challenging task needing low noise, large area, and wide dynamic range detectors. Scientific complementary metal oxide semiconductor (CMOS) Active Pixel Sensors will soon be able to meet all of these demands in a single device. Characterization of two novel Active Pixel Sensors that advance towards an ideal X-ray diffraction detector is presented. 'Vanilla' exhibits a low read noise of 55e r.m.s. and high quantum efficiency of up to 70% so was selected for the design and implementation of the first 'Active Pixel X-ray Diffraction' (APXRD) system. Following on from Vanilla, the 'Large Area Sensor' (LAS) covered an area of over 29cm2 and had a wide dynamic range of over 95dB. The first linear systems model of an Active Pixel Flat Panel Imager (scintillator coupled APS) was formulated in the design of the APXRD system, to select filters to narrow the spectral width of the X-ray beam and predict the recorded scatter intensity. Following system implementation, scatter signatures were recorded for numerous breast tissue equivalent samples. A multivariate analysis model calibrated with these was able to predict the percentage fat content of an 'unknown' sample to within 3% a very promising result. The width of the filtered polychromatic X-ray spectrum had only a minor influence on the APXRD scatter signatures indicating that the system preserves all relevant structural information.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.498868  DOI: Not available
Share: