Electronic Thesis and Dissertation Repository


Doctor of Philosophy


Electrical and Computer Engineering


Dr. Abbas Samani


Early detection of prostate cancer plays an important role in successful prostate cancer treatment. This requires screening the prostate periodically after the age of 50. If screening tests lead to prostate cancer suspicion, prostate needle biopsy is administered which is still considered as the clinical gold standard for prostate cancer diagnosis. Given that needle biopsy is invasive and is associated with issues including discomfort and infection, it is desirable to develop a prostate cancer diagnosis system that has high sensitivity and specificity for early detection with a potential to improve needle biopsy outcome. Given the complexity and variability of prostate cancer pathologies, many research groups have been pursuing multi-parametric imaging approach as no single modality imaging technique has proven to be adequate. While imaging additional tissue properties increases the chance of reliable prostate cancer detection and diagnosis, selecting an additional property needs to be done carefully by considering clinical acceptability and cost. Clinical acceptability entails ease with respect to both operating by the radiologist and patient comfort. In this work, effective tissue biomechanics based diagnostic techniques are proposed for prostate cancer assessment with the aim of early detection and minimizing the numbers of prostate biopsies. The techniques take advantage of the low cost, widely available and well established TRUS imaging method. The proposed techniques include novel elastography methods which were formulated based on an inverse finite element frame work. Conventional finite element analysis is known to have high computational complexity, hence computation time demanding. This renders the proposed elastography methods not suitable for real-time applications. To address this issue, an accelerated finite element method was proposed which proved to be suitable for prostate elasticity reconstruction. In this method, accurate finite element analysis of a large number of prostates undergoing TRUS probe loadings was performed. Geometry input and displacement and stress fields output obtained from the analysis were used to train a neural network mapping function to be used for elastopgraphy imaging of prostate cancer patients. The last part of the research presented in this thesis tackles an issue with the current 3D TRUS prostate needle biopsy. Current 3D TRUS prostate needle biopsy systems require registering preoperative 3D TRUS to intra-operative 2D TRUS images. Such image registration is time-consuming while its real-time implementation is yet to be developed. To bypass this registration step, concept of a robotic system was proposed which can reliably determine the preoperative TRUS probe position relative to the prostate to place at the same position relative to the prostate intra-operatively. For this purpose, a contact pressure feedback system is proposed to ensure similar prostate deformation during 3D and 2D image acquisition in order to bypass the registration step.

Included in

Biomedical Commons