Investigation of undesired errors relating to the planar array system of electrical impedance mammography for breast cancer detection

摘要

Breast cancer in women continues to be one of the leading causes of death in the world. Since the exact causes are not completely known, the most important approach is to reduce this mortality by early detection and treatment. Although the current detection techniques for breast cancer such as X-ray mammography provide useful information for diagnosis; development of a new imaging technique using non-ionising radiation is highly desirable in order to detect breast cancer at an early stage and overcome current limitations, such as age-dependent sensitivity. Electrical Impedance Mammography (EIM) provides a new solution to break through the current limitation for early cancer detection. The focus of this thesis is to investigate the current fourth generation Sussex EIM system. This system implements the EIM technique by examination of the tissue response to a multi-frequency injected current. The Sussex Mk4 system is discussed in detail followed by system hardware modelling. The hardware modelling includes both analogue and digital components. The analogue part includes modelling of the voltage to current converter (V-I) and analogue multiplexer while the digital section consists of modelling the signal generation, measurement and demodulating components. In the analogue section, bandwidth limitation due to the current source and the analogue multiplexer's configuration is also the prime focus of investigation along with the proposal to overcome it. Possible factors affecting the system performance and signal quality are also part of the research. In this section, possible factors are characterized and discussed in detail on the basis of external and internal sources of possible errors along with predictable and unpredictable noise sources. External sources of error artefacts introduced by the patients and their movements while scanning are most likely to affect the image reconstruction. Predictable and unpredictable causes may introduce frequency dependent noise whereas internal sources, which can be also be classified as systematic errors, degrade system performance due to electronic circuit design, configuration, stray capacitance and cable connections. Further, comprehensive investigation is performed on the in-vivo undesired voltage threshold levels which come hand-in-hand with the methods to mitigate the possible factors responsible for them. A comprehensive study and analysis is also carried out to determine what ratio of electrode blockage can affect the acquired raw data and how this may compromise reconstruction. Techniques for fast detection of any such occurrences are also discussed.