This study investigated the influence of semantic complexity treatment in individuals with fluent aphasia on discourse performance. Semantic treatment is an effective way to improve semantically based word retrieval problems in aphasia. Treatment focused on the semantic application of the Complexity Account of Treatment Efficacy (CATE) (Thompson, Shapiro, Kiran, & Sobecks, 2003) promotes training of complex items resulting in generalization to less complex, untrained items. In addition, research has shown that the personal relevance of treatment material can increase treatment efficacy. This study investigated the effect of semantic treatment of atypical personally relevant items among individuals with aphasia on discourse performance. Two treatment phases were applied to examine the influence of personally relevant and non-relevant treatment material on discourse performance. In addition, generalization from trained atypical items to untrained typical items was investigated. Methods and procedures were partially replicated from Kiran, Sandberg, & Sebastian (2011) examining semantic complexity within goal-derived (ad hoc) categories. Three participants with fluent aphasia were trained on three semantic tasks including category sorting, semantic feature generation/selection, and Yes/No feature questions. A generative naming task was used for probe data collection every second session. Stimuli consisted of atypical items only. The hypothesis that semantic complexity training of personally relevant items from ad hoc categories will produce greater generalization to associated, untrained items than training of non-relevant items and consequently increase discourse performance was not supported. The findings revealed a failure to replicate the magnitude and type of improvements previously reported for the typicality effect in generative naming. Clinical significance was found for personally relevant and non-relevant discourse performance. However, no consistent pattern was found within and across participants. In addition, effect size for generalization from trained atypical to untrained typical items was not significant. Limitations of this study lead to future directions to further specify participation selection, such as cognitive abilities, procedural changes, and the inclusion of discourse performance as an outcome measure. Overall, the results of this study provide weak support for replicating semantic treatment of atypical exemplars in ad-hoc categories and hence demonstrate the critical role of replication across labs to identify key issues in the candidacy, procedures, and outcome measurement of any developing treatment.

Standard accounts of mental health are based on a “deficit view” solely focusing on cognitive impairments associated with psychiatric conditions. Based on the principle of neural competition, we suggest an alternative. Rather than focusing on deficits, we should focus on the cognitive potential that selective dysfunctions might bring with them. Our approach is based on two steps: the identification of the potential (i.e., of neural systems that might have benefited from reduced competition) and the development of corresponding training methods, using the testing-the-limits approach. Counterintuitively, we suggest to train not only the impaired function but on the function that might have benefitted or that may benefit from the lesser neural competition of the dysfunctional system.

abstract: Medical images constitute a special class of images that are captured to allow diagnosis of disease, and their "correct" interpretation is vitally important. Because they are not "natural" images, radiologists must be trained to visually interpret them. This training process includes implicit perceptual learning that is gradually acquired over an extended period of exposure to medical images. This dissertation proposes novel computational methods for evaluating and facilitating perceptual training in radiologists. Part 1 of this dissertation proposes an eye-tracking-based metric for measuring the training progress of individual radiologists. Six metrics were identified as potentially useful: time to complete task, fixation count, fixation duration, consciously viewed regions, subconsciously viewed regions, and saccadic length. Part 2 of this dissertation proposes an eye-tracking-based entropy metric for tracking the rise and fall in the interest level of radiologists, as they scan chest radiographs. The results showed that entropy was significantly lower when radiologists were fixating on abnormal regions. Part 3 of this dissertation develops a method that allows extraction of Gabor-based feature vectors from corresponding anatomical regions of "normal" chest radiographs, despite anatomical variations across populations. These feature vectors are then used to develop and compare transductive and inductive computational methods for generating overlay maps that show atypical regions within test radiographs. The results show that the transductive methods produced much better maps than the inductive methods for 20 ground-truthed test radiographs. Part 4 of this dissertation uses an Extended Fuzzy C-Means (EFCM) based instance selection method to reduce the computational cost of transductive methods. The results showed that EFCM substantially reduced the computational cost without a substantial drop in performance. The dissertation then proposes a novel Variance Based Instance Selection (VBIS) method that also reduces the computational cost, but allows for incremental incorporation of new informative radiographs, as they are encountered. Part 5 of this dissertation develops and demonstrates a novel semi-transductive framework that combines the superior performance of transductive methods with the reduced computational cost of inductive methods. The results showed that the semi-transductive approach provided both an effective and efficient framework for detection of atypical regions in chest radiographs. / Dissertation/Thesis / Ph.D. Computer Science 2012