Combining and choosing case base maintenance algorithms

Case-Based Reasoning (CBR) uses past experiences to solve new problems. The quality of the past experiences, which are stored as cases in a case base, is a big factor in the performance of a CBR system. The system's competence may be improved by adding problems to the case base after they have been solved and their solutions verified to be correct. However, from time to time, the case base may have to be refined to reduce redundancy and to get rid of any noisy cases that may have been introduced. Many case base maintenance algorithms have been developed to delete noisy and redundant cases. However, different algorithms work well in different situations and it may be difficult for a knowledge engineer to know which one is the best to use for a particular case base. In this thesis, we investigate ways to combine algorithms to produce better deletion decisions than the decisions made by individual algorithms, and ways to choose which algorithm is best for a given case base at a given time. We analyse five of the most commonly-used maintenance algorithms in detail and show how the different algorithms perform better on different datasets. This motivates us to develop a new approach: maintenance by a committee of experts (MACE). MACE allows us to combine maintenance algorithms to produce a composite algorithm which exploits the merits of each of the algorithms that it contains. By combining different algorithms in different ways we can also define algorithms that have different trade-offs between accuracy and deletion. While MACE allows us to define an infinite number of new composite algorithms, we still face the problem of choosing which algorithm to use. To make this choice, we need to be able to identify properties of a case base that are predictive of which maintenance algorithm is best. We examine a number of measures of dataset complexity for this purpose. These provide a numerical way to describe a case base at a given time. We use the numerical description to develop a meta-case-based classification system. This system uses previous experience about which maintenance algorithm was best to use for other case bases to predict which algorithm to use for a new case base. Finally, we give the knowledge engineer more control over the deletion process by creating incremental versions of the maintenance algorithms. These incremental algorithms suggest one case at a time for deletion rather than a group of cases, which allows the knowledge engineer to decide whether or not each case in turn should be deleted or kept. We also develop incremental versions of the complexity measures, allowing us to create an incremental version of our meta-case-based classification system. Since the case base changes after each deletion, the best algorithm to use may also change. The incremental system allows us to choose which algorithm is the best to use at each point in the deletion process.

Open-source virtual bronchoscopy for image guided navigation

This thesis describes the development of an open-source system for virtual bronchoscopy used in combination with electromagnetic instrument tracking. The end application is virtual navigation of the lung for biopsy of early stage cancer nodules. The open-source platform 3D Slicer was used for creating freely available algorithms for virtual bronchscopy. Firstly, the development of an open-source semi-automatic algorithm for prediction of solitary pulmonary nodule malignancy is presented. This approach may help the physician decide whether to proceed with biopsy of the nodule. The user-selected nodule is segmented in order to extract radiological characteristics (i.e., size, location, edge smoothness, calcification presence, cavity wall thickness) which are combined with patient information to calculate likelihood of malignancy. The overall accuracy of the algorithm is shown to be high compared to independent experts' assessment of malignancy. The algorithm is also compared with two different predictors, and our approach is shown to provide the best overall prediction accuracy. The development of an airway segmentation algorithm which extracts the airway tree from surrounding structures on chest Computed Tomography (CT) images is then described. This represents the first fundamental step toward the creation of a virtual bronchoscopy system. Clinical and ex-vivo images are used to evaluate performance of the algorithm. Different CT scan parameters are investigated and parameters for successful airway segmentation are optimized. Slice thickness is the most affecting parameter, while variation of reconstruction kernel and radiation dose is shown to be less critical. Airway segmentation is used to create a 3D rendered model of the airway tree for virtual navigation. Finally, the first open-source virtual bronchoscopy system was combined with electromagnetic tracking of the bronchoscope for the development of a GPS-like system for navigating within the lungs. Tools for pre-procedural planning and for helping with navigation are provided. Registration between the lungs of the patient and the virtually reconstructed airway tree is achieved using a landmark-based approach. In an attempt to reduce difficulties with registration errors, we also implemented a landmark-free registration method based on a balanced airway survey. In-vitro and in-vivo testing showed good accuracy for this registration approach. The centreline of the 3D airway model is extracted and used to compensate for possible registration errors. Tools are provided to select a target for biopsy on the patient CT image, and pathways from the trachea towards the selected targets are automatically created. The pathways guide the physician during navigation, while distance to target information is updated in real-time and presented to the user. During navigation, video from the bronchoscope is streamed and presented to the physician next to the 3D rendered image. The electromagnetic tracking is implemented with 5 DOF sensing that does not provide roll rotation information. An intensity-based image registration approach is implemented to rotate the virtual image according to the bronchoscope's rotations. The virtual bronchoscopy system is shown to be easy to use and accurate in replicating the clinical setting, as demonstrated in the pre-clinical environment of a breathing lung method. Animal studies were performed to evaluate the overall system performance.

Challengers to provisional republicanism: the official republican movement, people’s democracy and the Irish republican socialist party, 1968–98

The Provisional IRA and its political wing Sinn Féin have attracted by far the greatest scholarly interest of all the players in the Northern Irish conflict. This emphasis is perfectly legitimate, given the centrality of the Provos to so many turning-points in the conflict, from the collapse of Stormont in the early 1970s to the hunger strikes of the following decade and the ceasefires which were followed by the Belfast Agreement. My project, however, looks at political groups that at one time or another challenged the Provos for leadership of the militant, anti-state constituency in Northern Ireland (chiefly based in the Catholic working class). Although never as large or influential as the Provisional republicans, groups such as the Official IRA and the Irish Republican Socialist Party sometimes had a discernible impact on the course of events which is overlooked by most studies, and often pioneered ideas and tactics that were later adopted by the Provos themselves. The idea that republicans should embrace political action and work in broad campaigning alliances was promoted by the IRSP and socialist groups such as People’s Democracy before it was taken up by Gerry Adams and his allies, while the Official IRA supported the principle of a settlement based on democratization of the Northern Irish state, which was later accepted by Sinn Féin in the form of the Belfast Agreement. The goal of my research is to provide a novel perspective on the conflict in Northern Ireland, while engaging with theoretical debates about its character.

Multi-party actions as a route to collective redress for mass arm, having particular regard to environmental mass harm

Accepted Version

Relevance of anatomy to medical education and clinical practice: perspectives of medical students, clinicians, and educators

Introduction: Against a backdrop of ever-changing diagnostic and treatment modalities, stakeholder perceptions (medical students, clinicians, anatomy educators) are crucial for the design of an anatomy curriculum which fulfils the criteria required for safe medical practice. This study compared perceptions of students, practising clinicians, and anatomy educators with respect to the relevance of anatomy education to medicine. Methods: A quantitative survey was administered to undergraduate entry (n = 352) and graduate entry students (n = 219) at two Irish medical schools, recently graduated Irish clinicians (n = 146), and anatomy educators based in Irish and British medical schools (n = 30). Areas addressed included the association of anatomy with medical education and clinical practice, mode of instruction, and curriculum duration. Results: Graduate-entry students were less likely to associate anatomy with the development of professionalism, teamwork skills, or improved awareness of ethics in medicine. Clinicians highlighted the challenge of tailoring anatomy education to increase student readiness to function effectively in a clinical role. Anatomy educators indicated dissatisfaction with the time available for anatomy within medical curricula, and were equivocal about whether curriculum content should be responsive to societal feedback. Conclusions: The group differences identified in the current study highlight areas and requirements which medical education curriculum developers should be sensitive to when designing anatomy courses.