This thesis presents the design of a personal computer (PC) based needle insertion robotic manipulator for biopsy. The robot was designed and built using materials available in the research laboratory. The robot is intended primarily for use inside the confined area of a cylindrical magnetic resonance (MR) scanner. Selection of the robot geometry and novel locations for drive actuators allowed placement of actuators outside the MRI bore. The robot is modeled using Denavit-Hartenberg transformations. Custom developed software control is developed to test the functional aspects of the robot. The robot performs to the tolerance required for the stated clinical application. This thesis addresses only proof of concept chosen for the manipulator design and is not ready for any clinical trials. The work also addresses MRI compatible and safety issues and recommends appropriate materials for future development. Traditionally, neurosurgical navigation has relied on preoperative images and the assumption that anatomical structures of interest remain in the same position with respect to each other and the fiducial markers used for registration. However, during surgery, tissue deformation and shift disrupt the spatial relation between the patient and the preoperative image volumes. This results in localization errors. Developing a manipulator that works inside an imaging machine guided by real time images is expected to minimize the problem of “tissue shift” during the surgery.

Osteoarthritis is characterized by pain and inflammation in joints, typically weight-bearing joints such as the knee. An early warning sign of osteoarthritis is the loss of proteoglycan molecules in the cartilage matrix. A surrogate method for measuring proteoglycan loss is detection of sodium ions, which ionically bond to negatively charged glycosaminoglycan side chains. Sodium MRI has the potential to non-invasively measure proteoglycan content, and hence act as a diagnostic tool for osteoarthritis. However, as sodium MRI suffers from low sodium concentrations in vivo and reduced MR sensitivity compared to standard proton MRI, techniques are required which optimize signal. This thesis examines the hardware, software, and acquisition techniques required in order to achieve high resolution, excellent quality sodium MR images of the human knee in vivo, which has potential applications in early diagnosis as well as pharmacological treatment evaluations of osteoarthritis.

No description available.

Metabolic imaging with hyperpolarized 13C-labeled cell substrates is a promising technique for imaging tissue metabolism in vivo. However, the transient nature of the hyperpolarization - and its depletion following excitation - limits the imaging time and the number of excitation pulses that can be used. A single-shot 3D imaging sequence has been developed and it is shown in this thesis to generate 13C MR images in tumour-bearing mice injected with hyperpolarized [1-13C]pyruvate. The pulse sequence acquires a stack-of-spirals at two spin echoes after a single excitation pulse and encodes the kz-dimension in an interleaved manner to enhance robustness to B0 inhomogeneity. Spectral-spatial pulses are used to acquire dynamic 3D images from selected hyperpolarized 13C-labeled metabolites. A nominal spatial/temporal resolution of 1.25 x 1.25 x 2.5 $mm^3$ x 2 s was achieved in tumour images of hyperpolarized [1-13C]pyruvate and [1-13C]lactate acquired in vivo. An advanced sequence is also described in this thesis in a later study to acquire higher resolution images with isotropic voxels (1.25 x 1.25 x 1.25 $mm^3$) at no cost of temporal resolution. EPI is a sequence widely used in hyperpolarized 13C MRI because images can be acquired rapidly with limited RF exposure. However, EPI suffers from Nyquist ghosting, which is normally corrected for by acquiring a reference scan. In this thesis a workflow for hyperpolarized 13C EPI is proposed that requires no reference scan and, therefore, that does not sacrifice a time point in the dynamic monitoring of tissue metabolism. To date, most of the hyperpolarized MRI on metabolism are based on 13C imaging, while 1H is a better imaging target for its 4 times higher gyromagnetic ratio and hence 16 times signal. In this thesis the world’s first dynamic 1H imaging in vivo of hyperpolarized [1-13C]lactate is presented, via a novel double-dual-spin-echo INEPT sequence that transfers the hyperpolarization from 13C to 1H, achieving a spatial resolution of 1.25 x 1.25 $mm^2$.

Purpose: To assess (1) the potential benefit that MRI may bring to prostate radiotherapy planning and delivery; (2) a method of improving registration of MRI and CT imaging to aid the RT planning workflow; (3) the role of in-bore MRI guided biopsy in informing management; (4) dosimetric outcome and toxicity of an integrated High Dose Rate (HDR-B) or Volumetric Modulated Arc Therapy (VMAT-IB) boost to the area of dominant disease within the prostate; (5) whether a predictive response can be identified measuring changes in Diffusion Weighted Imaging (DWI) and Dynamic Contrast Enhancement (DCE) during prostate RT after neo-adjuvant HT (NA-HT); (6) the necessity of hormone therapy (HT) with dose escalated radiotherapy (DE-RT) for intermediate risk prostate cancer. Methods: (1) Perform a systematic review of literature pertaining to MRI and image guided radiotherapy; (2) compare registration accuracy, based on displacement of fiducial markers or degree of overlap of segmented prostate measured by Dice Similarity Coefficient (DSC), of MRI and CT for 14 patients after conventional operator driven visual matching and then an additional registration step using interstitial points identified on high quality volumetric CT (HQVCT); (3) assess the predictive power of in-bore MRI guided biopsy of areas with suspicious appearance on multi-parametric MRI by comparing biopsy accuracy to histological findings and repeat biopsy results for 42 PIRADS 4-5 lesions in 31 men; (4) analyse patients treated in a prospective study receiving standard radiotherapy to the prostate plus a HDR-B (20 patients) or VMAT-IB (26) to a total dose of 250 Gy BED to assess acute and late toxicity and dosimetric variation between the two methods; (5) prospectively recruit 15 patient who have received NA-HT and perform DWI and DCE before, during and after completion of radiotherapy to look for significant changes in values in normal and malignant tissue which may predict for ultimate outcome; (6) Assess clinical outcome for patients receiving 75.6 – 78 Gy +/- bicalutamide. Results: (1) The review has quantified uncertainties in treatment delivery and the degree that the addition of MRI may mitigate this; (2) point based registration of CT and MRI imaging after visual registration achieved a significant reduction in fiducial marker displacement and a significant increase in DSC; (3) seven lesions targeted by in-bore MR guided biopsy had non-significant or negative results, most with biopsy needle deflected to the target periphery with four confirmed false negative on repeat biopsy; (4) with a median follow up of 12 months acute and late toxicity was similar after either treatment with HDR-B delivering a significantly higher dose to a proportion of the gross tumour volume (GTV) but with significantly lower minimum dose to the planned target volume (PTV); (5) tumour DWI values during RT after NA-HT were not found to significantly alter, DCE was found to vary significantly during treatment and initial changes correlated with changes in DWI; (6) the addition of bicalutamide did not significantly improve biochemical control or overall survival. Conclusions: (1) Routine use of MRI will to improve radiotherapy planning and delivery; (2) repeat point based registration using interstitial points has the potential to improve visual CT and MRI registration; (3) an in-bore MRI guided biopsy has little value in informing a decision to offer focal therapy to an MRI identified PIRADS 4-5 lesion due to its high false negative rate; (4) with limited follow up HDR-B and VMAT-IB appear safe methods of focal dose escalation although with significant dosimetric variations;(5) early changes in DWI and DCE during RT after NA-HT appear to correlate, longer follow up will assess their prognostic value; (6) A benefit of HT combined with DE-RT was not shown in this study.

Diffusion MRI is established as an essential tool for both clinicians as well as biomedical scientists. Its application plays an important role in diagnosis and management of acute stroke, tumors, trauma, and infectious disease, among myriad other applications. Furthermore, diffusion studies are crucial for understanding disease processes caused by developmental and neurodegenerative disorders. The latest developments in quantitative diffusion imaging have broadened the potential application of the technique for both clinical and research applications. However, ongoing research is critical in order to further improve the accuracy and reproducibility of quantitative diffusion MRI techniques. Correlation time diffusion (D-CT) is emerging as an alternative technique for obtaining diffusion qMRI data[1][2][3]. Using the D-CT technique, T1 relaxation data is analyzed, using a modified BPP relaxation theory, in order to calculate the correlation times of protons’ stochastic processes and relate these times to solution viscosity in order to calculate proton diffusion coefficients, ADCs. The purpose of our study was to compare age related changes, during childhood and early adulthood, of global brain diffusion coefficients obtained by correlation time technique to global brain diffusion coefficients obtained by a conventional pulsed field gradient technique. In our study, we used the data of 27 subjects (0.5-24 years old), who were scanned with Mixed-TSE and DW-SS-SE-EPI pulse sequences. Subsequently, we processed the resulting directly acquired images to generate T1, T2, PD, ADC maps as well as volumetric data. We used the student t-test and linear regression analysis to compare and interpret our data. Our results show a strong positive correlation between the volumetric data. Good correlation between ADC values was observed, with the widest discrepancy between DCT, DPFG (about 17%) observed in the youngest subjects, and the smallest discrepancy noted in the older subjects.

No description available.

The claustrum (Cl) is a subcortical gray-matter structure housed between the external capsule medially and the extreme capsule laterally. Due to its extensive reciprocal connections throughout the brain, it has been implicated in consciousness and other higher order functions including linking behavior and emotion. Such linkage may be important in understanding the neurobiology of autism since other cortical and subcortical regions including the spatially and ontologically related basal ganglia, as well as limbic structures, have been implicated in the disorder. Participants were males with autism (n=16) and typically developing (TD; n=14) matched for head circumference and age. The Cl and other structures were identified in 3-Tesla MRI scans using ANALYZE®, then segmented and volume quantified. Four Cl volumes were traced (i.e., right, left, right ventral, left ventral) first in axial plane then in coronal plane for entire Cl visualization. Two-tailed single sample t-tests revealed significant differences in the right claustrum (p=.014), left claustrum (p = .041), right total claustrum (p = .018) and left total claustrum (p=.044). Right Cl volume was found to be significantly larger than left within each of the groups (Autism, p=.021; TD, p=.033). These preliminary results demonstrate that the Cl can be consistently identified in vivo using ROI tracing with apparent right-versus-left asymmetry documented. Smaller claustral volumes in autism support theories of a disconnect in long-range circuitry associated with autism.

With the desired deployment of Artificial Intelligence (AI), concerns over whether AI can “communicate” why it has made its decisions is of particular importance. In this thesis, we utilize predictive entropy (PE) as an surrogate for predictive uncertainty and report it for various test-time conditions that alter the testing distribution. This is done to evaluate the potential for PE to indicate when users should trust or dis- trust model predictions under dataset shift or out-of-distribution (OOD) conditions, two scenarios that are prevalent in real-world settings. Specifically, we trained an ensemble of three 2D-UNet architectures to segment synthetically damaged regions in fractional anisotropy scalar maps, a widely used diffusion metric to indicate mi- crostructural white-matter damage. Baseline ensemble statistics report that the true positive rate, false negative rate, false positive rate, true negative rate, Dice score, and precision are 0.91, 0.091, 0.23, 0.77, 0.85, and 0.80, respectively. Test-time PE was reported before and after the ensemble was exposed to increasing geometric distortions (OOD), adversarial examples (OOD), and decreasing signal-to-noise ratios (dataset shift). We observed that even though PE shows a strong negative correlation with model performance for increasing adversarial severity (ρAE = −1), this correlation is not seen under distortion or SNR conditions (ρD = −0.26, ρSNR = −0.30). However, the PE variability (PE-Std) between individual model predictions was shown to be a better indicator of uncertainty as strong negative correlations between model performance and PE-Std were seen during geometric distortions and adversarial ex- amples (ρD = −0.83, ρAE = −1). Unfortunately, PE fails to report large absolute uncertainties during these conditions, thus restricting the analysis to correlative relationships. Finally, determining an uncertainty threshold between “certain” and “uncertain” model predictions was seen to be heavily dependant on model calibra- tion. For augmentation conditions close to the training distribution, a single threshold could be hypothesized. However, caution must be taken if such a technique is clinically applied, as model miscalibration could nullify such a threshold for samples far from the distribution. To ensure that PE or PE-Std could be used more broadly for uncertainty estimation, further work must be completed. / Thesis / Master of Applied Science (MASc)

No description available.