Crohn's disease is an inflammatory bowel disease associated with substantial morbidity. Complications arising from this disease affect many organ systems. In particular, hepatitis, pancreatitis, blood dyscrasias, and renal disease are believed to occur more frequently in patients with Crohn's disease. The incidence of these conditions is also believed to be increased by some of the medications used to treat Crohn's disease. Sulfasalazine as well as mesalamine have been associated with hepatitis, pancreatitis, renal disease, and blood dyscrasias. In addition to characterizing the demographics and severity of Crohn's disease in Saskatchewan, the purpose of this study was also to determine if, in patients with Crohn's disease, there is an increased risk of developing these adverse conditions associated with the medications used to treat this condition. In this study 1999 patients with Crohn's disease who met inclusion criteria, were identified in the Saskatchewan Healthcare datafiles. (Abstract shortened by UMI.)

A lesion interface consisting of a glia limitans lined by a laminin-rich basal lamina forms within 3 weeks after leptomeningeal cells infiltrate into penetrating wounds in the adult mammalian CNS. This impedes axon regrowth. We have examined the effects of steroids on the formation of the lesion interface in the adult rat cerebral cortex. Topical glucocorticoids were applied on the surface of cortex encompassing stab wounds. Three weeks later, cryostat sections through the lesioned area were labeled with anti-laminin, anti-GFAP, anti-ED-1 and Nuclear Yellow. Steroid treatment attenuated all components of the lesion, including leptomeningeal cell infiltration. In vitro, steroid treatment did not alter laminin secretion but reduced cell proliferation in leptomeningeal cultures. These results suggest that steroids modulate lesion interface formation in the CNS in part by decreasing leptomeningeal cell proliferation.

Pre-operative image based image guided neurosurgery (IGNS) systems involve the mapping of intraoperative real world coordinates and trajectories to one or more preoperative image data spaces. This is accomplished using a rigid body transformation from patient to image coordinate systems, which is usually obtained by identifying external landmarks on the patient and on the images, and employing a least squares minimization technique. The validity of the rigid body transformation relies on the assumption that the patient and localizing device form a completely rigid body for the duration of the procedure. It has been observed that the brain tissue moves significantly within the skull, particularly after large open craniotomies, violating this assumption of rigidity. / A surgical guidance system has been developed that combines pre-operative image information (e.g. MRI or CT) and intraoperative ultrasound imaging to detect brain tissue deformation during IGNS. The system includes hardware and software to track the ultrasound transducer during image acquisition, and visualization software to view the live ultrasound and co-planar MRI images. The software includes tools to delineate structures in either modality and overlay these images on one another, and warp the preoperative image based on the delineated structures. / The thesis includes a discussion of techniques, instruments and results, using a novel calibration tool and a multi-modality deformable phantom. Finally, the clinical experience from the use of this system in the operating room are presented.

Experimental investigations were carried out concerning the use of bovine bone (Unilab Surgibone) grafted in the tibiae of rats. The first experiment evaluated tissue response of bovine bone as an inlay graft and the second experiment as an onlay graft. Histological and morphometric analyses were performed in order to obtain baseline data on tissue response for future experiments using titanium implants with bovine bone grafts in this model. Light microscopy demonstrated rapid incorporation of the inlay graft by new bone, whereas, the onlay graft was mainly encapsulated by fibrous tissue. However, a residual increase in the thickness of the outer cortex of the tibiae was observed with onlay graft. From the results of this study it appeared that Unilab Surgibone was biocompatible and did not induce a foreign body reaction. Future investigations using titanium implants in combination with the bovine bone grafts appears to be possible especially if an inlay/onlay design is attempted.

To investigate differences in neuronal responses to axotomy in the mammalian central nervous system, I compared the patterns of retinal ganglion cell (RGC) survival after intracranial optic nerve (ON) cut or crush in adult rats. After ON cut, approximately 40% of the RGCs were lost in the second week, followed by a more protracted cell loss, while ON crush gave rise to a slow RGC loss beginning at one week. Retrograde axonal degeneration and macrophage invasion into the ON stump was more marked after ON cut than crush, but progressive retrograde degeneration of axons into the retina did not precede the onset of RGC loss. In parallel experiments, I found evidence that brain derived neurotrophic factor mRNA is expressed in non-neuronal cells of the uninjured ON. These observations suggest that events other than neuron-target separation play a role in the timing and severity of neuronal loss after axotomy.

This study was designed to characterize postcesarean pain and examine its relationship with surgical anesthesia. Pain intensity was measured using a 0 to 10 numerical rating scale. The Short-Form McGill Pain Questionnaire was used to measure quality. The most common pain types reported were movement-associated and constant incisional pain (100% of subjects), gas pain (88.1%), and uterine contraction pain (83.3%). Each pain type differed in its intensity, duration and quality. Movement-associated incisional pain was the most intense and long-lasting pain type reported. On postoperative days 2 to 4, mothers who received complete epidural anesthesia during surgery reported less intense movement-associated incisional pain than those who received general or incomplete epidural anesthesia. However, statistical significance was not often obtained. These findings demonstrate the uniqueness of each pain type composing the postcesarean pain experience. The clinical data also support the hypothesis that epidural anesthesia diminishes the sustained hyperexcitability of the central nervous system caused by surgery.

In this thesis, measurement theory and the statistical evaluations of diagnostic test performance are applied to the measurement of injury severity. Relevant issues in traumatology, the rationale for measurement of injury severity and the importance of likelihood ratios (LRs) and Receiver operator characteristic (ROC) curve analyses are discussed. An assessment of the definition, scaling mechanism, reliability and validity of 22 severity instruments is organized into a reference guide. Data sources for this thesis include the literature, the Vancouver General Hospital trauma registry and the Pennsylvania State University trauma registry. / The LR and the area under ROC curves (A$ sb{ rm uc}$) are calculated from the best published evaluations of four triage instruments. The CRAMS, PHI, and RTI, which include anatomic information, are superior to the RTS, which contains only physiologic information. / A paired ROC curve analysis of the RTS and the RTI demonstrates that the performance of the RTI is equivalent to the RTS when mortality is the outcome; however the performance gain of the RTI over the RTS is 35% when major trauma is the outcome evaluated. / Explanations for differences in predictive validity are sought by evaluating scaling mechanisms, reliability and content validity. (Abstract shortened by UMI.)

No description available.

The family of proprotein convertases has been recently implicated in tumorigenesis and metastasis in animal models. However, these studies have not yet been completely corroborated in human tumors. Here, we show that mRNA, protein expression, and protein cleavage profiles of proprotein convertases 1 and 2 are altered in liver colorectal metastasis, compared to unaffected and normal liver. Active PC1 is overexpressed in tumor, correlating with its mRNA profile. Moreover, the enhanced PC2 processing pattern in tumor correlates with the overexpression of its specific chaperone 7B2, which in turn may represent a target for early diagnosis and treatment. The increased PC2 maturation, as well as the overexpression and altered processing of PC1 may be either a cause or a consequence for the observed metastasic phenotype. Nevertheless, they may result in the alteration of the secretory pathway, which could therefore, modify the cellular microenvironment and thus favor tumor growth and/or metastasis.

This dissertation presents a series of studies aimed at applying machine learning methods to information derived from magnetic resonance (MR) based examinations in order to aid in the pre-surgical evaluation of patients with epilepsy. Two forms of epilepsy were studied: non-lesional temporal lobe epilepsy (TLE) and extra-temporal lobe epilepsy (ETLE) due to malformations of cortical development (MCD). / Regarding patients with TLE, our aim was to predict outcome, in terms of reduction of seizure frequency, following surgical resection. To this end, we trained a Bayesian classifier on results from volumetric magnetic resonance imaging (MRI) and magnetic resonance spectroscopic imaging (MRSI), which allow rapid, non-invasive measurement of structural and metabolic data, respectively. We demonstrated that the pattern of MR markers can predict whether or not a patient with TLE will benefit from surgery. / In our studies of patients with ETLE, we focused on patients whose epilepsy was due to focal cortical dysplasia (FCD), a common form of MCD. In these patients, the identification of FCD lesions is critical in helping to direct the site of surgical resection. This is commonly performed by visual analysis of conventional MRI. The MRI characteristics of FCD are well known; however, in many patients, lesions of FCD are characterized by minor structural abnormalities that go unrecognized or are too subtle to be detected by standard radiological analysis. Thus, the objective of this part of the dissertation was to use mathematical models of the MRI characteristics of FCD as a basis for automated detection of FCD lesions. The mathematical models included first-order statistical and morphological operators which can help measure visually discernable MRI characteristics of FCD lesions, and second-order texture analysis, which can quantify information regarding tissue structure or organization not readily accessible through visual analysis. A Bayesian classifier trained on these models demonstrated a significantly increased sensitivity in lesion detection compared to standard analysis of conventional MRI. / Both components of this thesis present clinically useful techniques for applying machine learning methods to MR data to assist in the pre-surgical evaluation of epilepsy patients. These methods are intended to be used in conjunction with conventional approaches.