Current prostate cancer diagnosis and treatment methods do not routinely utilise the most advanced soft tissue contrast imaging tool, Magnetic Resonance Imaging (MRI). The incorporation of MRI in diagnosis and treatment planning could improve significantly the clinical management and outcomes for the patients. This thesis addresses the impact MRI could have on these two key areas.
The current routine clinical diagnosis methods, Digital Rectal Exam (DRE), Prostate Specific Antigen (PSA) Blood test and prostate biopsy have a combined 50% success rate for correctly diagnosing cancer. The DRE and PSA tests produce significant false positive results. However, biopsy examination produces significant false negative results. The success of biopsy depends on how many core samples are taken because even with ultrasound guidance the tumours cannot be detected, therefore core samples are taken blindly. MRI sequences can produce contrast dependent on the water diffusion in tissue and this may be used to distinguish tumour from healthy tissue. Diffusion Weighted Imaging (DWI) in conjunction with the quantitative diffusion measurement called the Apparent Diffusion Coefficient (ADC) can be used to create an ADC map. The ADC map has the potential to produce greater contrast than a DWI. The current sequence used to gain DWI and ADC maps, Echo Planar Imaging (EPI), have considerable errors and artefacts associated with the sequence. In this thesis a different sequence with potentially decreased artefacts, Half fourier Acquisition Single shot Turbo Spin Echo (HASTE) was tested for increased potential of detecting prostate cancer over EPI. The first step was to assess the dependency of the calculated ADC value on the acquisition protocols. This allowed selection of optimised protocols for the calculation of ADC using both sequences. The aim was to investigate if MRI can significantly increase the detection of prostate cancer tumours compared to current methods.
If tumour location information could be added, the radiation therapy treatments would benefit from improved prostate location information and greatly benefit. Computer Tomography (CT), currently used for radiation therapy treatment planning has poor soft tissue contrast and relies upon bony structures to define prostate location. MRI has superior soft tissue contrast, compared to CT, clearly showing the prostate boundary which, if implemented would increase the accuracy of the initial planning stage. Prostate motion, caused by primarily bladder and rectal filling increases the error margin placed around the planning target to ensure all cancer tissue is included. This also increases the amount of healthy tissue including the bladder and rectum that may be negatively affected. Using CT the lack of clear prostate boundaries and inaccurate motion definition significantly increase with the treatment margin required. A plan using MRI would be less affected by these issues. If the prostate motion was measured through MRI the error margin could be reduced. The superior soft v tissue contrast of MRI over CT also allows accurate delineation of the prostate. The delineation can produce a computer model of the prostate which can be used for further analysis. In this thesis the prostate computer reconstructions were used to accurately measure the prostate motion based on shape registration. In addition the reconstructions were used to establish the radiation planning coverage on the prostate based on motion between imaging and treatment time.
The work in this thesis will scientifically support the implementation of MRI into prostate cancer detection and treatment which will lead to improved management of patients and treatment success.