Search

Electronic medical record (EMR) data is collected on a daily basis at hospitals and other healthcare facilities to track patients’ health situations including conditions, treatments (medications, procedures), diagnostics (labs) and associated healthcare operations. Besides being useful for individual patient care and hospital operations (e.g., billing, triaging), EMRs can also be exploited for secondary data analyses to glean discriminative patterns that hold across patient cohorts for different phenotypes. These patterns in turn can yield high level insights into disease progression with interventional potential. In this dissertation, using a large scale realistic EMR dataset of over one million patients visiting University of Kentucky healthcare facilities, we explore data mining and machine learning methods for association rule (AR) mining and predictive modeling with mood and anxiety disorders as use-cases. Our first work involves analysis of existing quantitative measures of rule interestingness to assess how they align with a practicing psychiatrist’s sense of novelty/surprise corresponding to ARs identified from EMRs. Our second effort involves mining causal ARs with depression and anxiety disorders as target conditions through matching methods accounting for computationally identified confounding attributes. Our final effort involves efficient implementation (via GPUs) and application of contrast pattern mining to predictive modeling for mental conditions using various representational methods and recurrent neural networks. Overall, we demonstrate the effectiveness of rule mining methods in secondary analyses of EMR data for identifying causal associations and building predictive models for diseases.

Memory for the temporal order of events is an essential cognitive function, but has been observed to decline in the process of normal, healthy aging. One avenue that has yet to be explored in this body of research is how temporal order memory relates to the establishment of new, causal associations between ordered events, and how prior causal knowledge influences judgments of event order in older adults. These questions are evaluated in two experiments using a novel stimulus set. The stimulus set introduced here includes paired images depicting both cause-and-effect sequences, and pairs of events that merely share a semantic or thematic association. Continuous sequences of both related and unrelated pairs of these images were presented during an incidental encoding task, and participants' memory for the order of target pairs was later evaluated in a yes / no temporal order recognition judgment task. A subset of participants also completed an item recognition task to address dissociations between memory for item identity and for item order. The results provided new support for age-related declines in memory for order, together with relatively intact performance for recognizing an item's identity. The results also identify age-related changes in the ability to form new ordered associations between unrelated, unassociated events. The basic features of causal order learning are evaluated in light of these results, and point to the importance of temporal proximity between events in establishing ordered cause-and-effect relations. Of particular interest was the influence of prior causal order knowledge on order memory judgments, and mixed support is found for an interaction between prior knowledge of an event sequence's `logical' order, and accurate recognition memory for those events.