Search

Quantitative volumes from brain magnetic resonance imaging (MRI) acquired across the life course may be useful for investigating long term effects of risk and resilience factors for brain development and healthy ageing, and for understanding early life determinants of adult brain structure. Therefore, there is an increasing need for automated segmentation tools that can be applied to images acquired at different life stages. We developed an automatic segmentation method for human brain MRI, where a sliding window approach and a multi-class random forest classifier were applied to high-dimensional feature vectors for accurate segmentation. The method performed well on brain MRI data acquired from 179 individuals, analysed in three age groups: newborns (38-42 weeks gestational age), children and adolescents (4-17 years) and adults (35-71 years). As the method can learn from partially labelled datasets, it can be used to segment large-scale datasets efficiently. It could also be applied to different populations and imaging modalities across the life course.

Preterm infants are susceptible to inflammation-induced white matter injury but the exposures that lead to this are uncertain. Histologic chorioamnionitis (HCA) reflects intrauterine inflammation, can trigger a fetal inflammatory response, and is closely associated with premature birth. In a cohort of 90 preterm infants with detailed placental histology and neonatal brain magnetic resonance imaging (MRI) data at term equivalent age, we used Tract-based Spatial Statistics (TBSS) to perform voxel-wise statistical comparison of fractional anisotropy (FA) data and computational morphometry analysis to compute the volumes of whole brain, tissue compartments and cerebrospinal fluid, to test the hypothesis that HCA is an independent antenatal risk factor for preterm brain injury. Twenty-six (29%) infants had HCA and this was associated with decreased FA in the genu, cingulum cingulate gyri, centrum semiovale, inferior longitudinal fasciculi, limbs of the internal capsule, external capsule and cerebellum (p in utero for a significant proportion of preterm infants, which focuses attention on the development of methods for detecting fetuses and placentas at risk as a means of reducing preterm brain injury.

The amount of medical knowledge is ever growing. This growth is welcomed and hoped for as new and better information that can improve patient care becomes available. It does, however, have the potential to become overwhelming and unmanageable. This project began as an idea to improve the knowledge development and maintenance process. It was coupled with an idea for online order sets that met the need of physicians within Intermountain Health Care (IHC). After gathering concepts and elements from Knowledge Management (KM) and Configuration Management (CM) and reviewing the successes of similar research, analyses were performed on several processes of groups and individuals who develop and maintain knowledge content at IHC. Requirements were gathered and developed based on the analyses, the concepts from KM and CM, and the lessons learned from previous research. The requirements guided the development of the Knowledge Repository Online (KRO), an application that allows users to find and use knowledge content as well as to perform reviews on knowledge content. The application has been shown to be successful in clinical use as approximately 26% of the patient charts in two pilot sites contain order sets from KRO. There are over 500 documents that have been loaded into KRO for both clinical use and for review. These numbers suggest that the application can be used in the knowledge development and maintenance process and that clinicians can be effective in contributing knowledge content.

Medical informatics technology is advocated by the American Medical Informatics Association and the International Medical Informatics Association as a better way to provide knowledge skills to healthcare providers by implementing better medical data and knowledge management. Although the 'lecture' method is the most traditional form of teaching in higher education, active learning techniques have been shown to be more effective as they require students be actively engaged in the learning process. As part of this dissertation research, computerized ontologies were used as a form of active learning. In medical informatics, ontologies are data models that define medical concepts and the relationships among those concepts. This dissertation analyzed whether ontologies could be effectively used as an active learning activity to teach physical therapy students. The dissertation consisted of three projects: 1 Develop a computerized ontology for physical therapy students, 2.Provide physical therapy students with an opportunity to instantiate the ontology into a computerized form, and 3. Compare three modes of teaching - lecture with an active learning activity, lecture with the computerized ontology as the active learning activity, and using the computerized ontology as the only teaching method. The developed computerized ontology was used to teach musculoskeletal topics to physical therapy students. Key concepts for the ontologies were derived from the curriculum of the Physical Therapy Education Department at Rockhurst University. In the development process, the computerized modules were set up in a manner that allowed future use on other topics. The computerized physical therapy ontology performed very well as an active learning application. Students enjoyed instantiating the ontologies and using them to learn physical therapy concepts. In addition, the students who used the computerized ontologies, performed as well or better, academically, than those who used other learning methods. As a consequence, it can be argued that ontologies should be used more frequently in physical therapy education and perhaps as a teaching method and learning activity for other healthcare professionals. Careful searching of the literature shows that this dissertation work is groundbreaking in uniting the fields of medical informatics and physical therapy, and future development promises to be exciting.