Informatics Theses and Dissertations

Permanent URI for this community

https://hdl.handle.net/1805/4070

Browse

Now showing 1 - 10 of 165

Academic laboratory information management system: a tool for science and computer science students
(2011-07-08) Lerch, Spencer; Merchant, Mahesh; Wild, David; Doman, Thompson N.
Proof of Concept - An Academic LIMS application: The aim of this project is the creation of an open-source, freeware LIMS application that can be used in an academic setting as a teaching tool for both chemistry and computer science students. The LIMS package will combine an application, developed using VB.NET, to manage the data with other open-source or freeware programs such as MySQL and WEKA. The numerous commercial chemical informatics applications available are useful tools to learn how to manage data from a user's standpoint. However, they are not readily available to the average student, nor do they offer a great understanding into how they were developed from a programmer's frame of mind. There is a great void here that, if filled can greatly help the academic community.
Acceptance of use of personal health record: factors affecting physicians' perspective
(2011-10-19) Agrawal, Ekta; Jones, Josette F.; Weiner, Michael; Simmermaker, Jennifer
Acceptance of PHR by physicians is fundamental as they play important role towards the promotion of PHR adoption by providing the access to the data to be maintained in PHR and also, using the information within the PHR for decision making. Therefore it is important to measure physicians' perspective on usefulness of PHR, and also the value and trust they have in PHR usage. Review of previous researches identifies the lack of availability of a valid survey instrument that can be used to measure physicians' perception on all different aspects of PHR use and acceptance. Using the integrated literature review methodology and Unified Theory of Acceptance and Use of Technology (UTAUT) as a guiding framework, this study was aimed to identify the factors that can be used in the development of comprehensive evaluation instrument to understand physicians' acceptance of PHR. Total 15 articles were selected for literature review and using the content analysis method, 189 undifferentiated data units were extracted from those articles. These data units were then categorized into the four core constructs of UTAUT. ―Other categorization system was also created for the data units that could not be classified into one of the UTAUT core constructs. Among four core UTAUT constructs, Performance Expectancy is found to be the most influential factor in physicians' acceptance of PHR, followed by ―Other factors, Facilitating Condition and Social Influence. Effort expectancy was found to be the least influential. The identified specific factors within each domain can be used to develop a valid survey instrument to measure physicians' perception on PHR.
ACTIVE READING ON TABLET TEXTBOOKS
(2015-04-17) Palilonis, Jennifer Ann; Defazio, Joseph; Bolchini, Davide; Butler, Darrell; Voida, Amy
To study a text, learners often engage in active reading. Through active reading, learners build an analysis by annotating, outlining, summarizing, reorganizing and synthesizing information. These strategies serve a fundamental meta-cognitive function that allows content to leave strong memory traces and helps learners reflect, understand, and recall information. Textbooks, however, are becoming more complex as new technologies change how they are designed and delivered. Interactive, touch-screen tablets offer multi-touch interaction, annotation features, and multimedia content as a browse-able book. Yet, such tablet textbooks-in spite of their increasing availability in educational settings-have received little empirical scrutiny regarding how they support and engender active reading. To address this issue, this dissertation reports on a series of studies designed to further our understanding of active reading with tablet textbooks. An exploratory study first examined strategies learners enact when reading and annotating in the tablet environment. Findings indicate learners are often distracted by touch screen mechanics, struggle to effectively annotate information delivered in audiovisuals, and labor to cognitively make connections between annotations and the content/media source from which they originated. These results inspired SMART Note, a suite of novel multimedia annotation tools for tablet textbooks designed to support active reading by: minimizing interaction mechanics during active reading, providing robust annotation for multimedia, and improving built-in study tools. The system was iteratively developed through several rounds of usability and user experience evaluation. A comparative experiment found that SMART Note outperformed tablet annotation features on the market in terms of supporting learning experience, process, and outcomes. Together these studies served to extend the active reading framework for tablet textbooks to: (a) recognize the tension between active reading and mechanical interaction; (b) provide designs that facilitate cognitive connections between annotations and media formats; and (c) offer opportunities for personalization and meaningful reorganization of learning material.
Advanced natural language processing and temporal mining for clinical discovery
(2015-08-17) Mehrabi, Saeed; Jones, Josette F.; Palakal, Mathew J.; Chien, Stanley Yung-Ping; Liu, Xiaowen; Schmidt, C. Max
There has been vast and growing amount of healthcare data especially with the rapid adoption of electronic health records (EHRs) as a result of the HITECH act of 2009. It is estimated that around 80% of the clinical information resides in the unstructured narrative of an EHR. Recently, natural language processing (NLP) techniques have offered opportunities to extract information from unstructured clinical texts needed for various clinical applications. A popular method for enabling secondary uses of EHRs is information or concept extraction, a subtask of NLP that seeks to locate and classify elements within text based on the context. Extraction of clinical concepts without considering the context has many complications, including inaccurate diagnosis of patients and contamination of study cohorts. Identifying the negation status and whether a clinical concept belongs to patients or his family members are two of the challenges faced in context detection. A negation algorithm called Dependency Parser Negation (DEEPEN) has been developed in this research study by taking into account the dependency relationship between negation words and concepts within a sentence using the Stanford Dependency Parser. The study results demonstrate that DEEPEN, can reduce the number of incorrect negation assignment for patients with positive findings, and therefore improve the identification of patients with the target clinical findings in EHRs. Additionally, an NLP system consisting of section segmentation and relation discovery was developed to identify patients' family history. To assess the generalizability of the negation and family history algorithm, data from a different clinical institution was used in both algorithm evaluations.
Advancing Toxicology-Based Cancer Risk Assessment with Informatics
(2010-05-03T19:38:33Z) Bercu, Joel P.; Mahoui, Malika; Romero, Pedro R.; Stevens, James L.; Jones, Josette F.; Palakal, Mathew J.
Since exposure to carcinogens can occur in the environment from various point sources, cancer risk assessment attempts to define and limit potential exposure such that the risk of developing cancer is negligible. While cancer risk assessment is widely used with certain methodologies well accepted in the scientific literature and regulatory guidances, there are still gaps which increase uncertainties when assessing risk including: (1) mixtures of genotoxins, (2) genotoxic metabolites, and (3) nongenotoxic carcinogens. An in silico model was developed to predict the cancer risk of a genotoxin which improved methodology for a single compound and mixtures. Monte Carlo simulations performed with a carcinogenicity potency database to estimate the overall carcinogenic risk of a mixture of genotoxic compounds showed that structural similarity would not likely increase the overall cancer risk. A cancer risk model was developed for genotoxic metabolites using excretion material in both animals and humans to determine the probability not exceeding a 1 in 100,000 excess cancer risk. Two model nongenotoxic compounds (fenofibrate and methapyraline) were tested in short-term microarray studies to develop a framework for cancer risk assessment. It was determined that a threshold for potential key events could be derived using benchmark dose analysis in combination with well developed ontologies (Kegg/GO), which were at or below measured tumorigenic and precursor events. In conclusion, informatics was effective in advancing toxicology-based cancer risk assessment using databases and predictive techniques which fill critical gaps in its methodology.
Ambulatory Computerized Provider Order Entry and PDA-Based Clinical Decision Support Systems: An Investigation of their Patient Safety Effectiveness via an Integrative and Systematic Review
Taffel, Jared Ross; Jones, Josette F.
Substantial research has been done on inpatient provider order entry systems with varying degrees of clinical decision support. Such studies have examined how these technologies impact patient safety as well as the quality and cost of care. However, given that most medical care and prescriptions are administered in an ambulatory setting, the dearth of research on ACPOE systems is quite astonishing. This knowledge gap demonstrates the need for an integrative and systematic literature review that attempts to assess the research done on computerized patient safety interventions in ambulatory care. This review’s findings provided adequate evidence that ACPOE systems are effective interventions for reducing medication errors. Other evidence further indicated that, in terms of functional capabilities, commercial ACPOE and e-prescribing systems may be catching up with their homegrown counterparts. PDA-based CDSSs were depicted as useful tools for raising adherence to guidelines and inducing safer prescribing. These findings suggest that ACPOE And PDA-based CDS systems show promise for improving safety and healthcare quality in ambulatory settings. ACPOE specifically, tended to have more advanced CDS attributes but, nonetheless, showed more negative results compared to the e-prescribing systems. Close scrutiny should therefore be given to the elements of decision support that ambulatory physicians find most useful.
An Application for Downloading and Integrating Molecular Biology Data
Fontaine, Burr R.; Foroud, Tatiana
Integrating large volumes of data from diverse sources is a formidable challenge for many investigators in the field of molecular biology. Developing efficient methods for accessing and integrating this data is a major focus of investigation in the field of bioinformatics. In early 2003, the Hereditary Genomics division of the department of Medical and Molecular Genetics at IUPUI recognized the need for a software application that would automate many of the manual processes that were being used to obtain data for their research. The two primary objectives for this project were: 1) an application that would provide large-scale, integrated output tables to help answer questions that frequently arose in the course of their research, and 2) a graphic user interface (GUI) that would minimize or eliminate the need for technical expertise in computer programming or database operations on the part of the end-users. In early 2003, Indiana University (IU), IBM, and the Indiana Genomics Initiative (INGEN) introduced a new resource called Centralized Life Sciences Data Services (CLSD). CLSD is a centralized data repository that provides programmatic access to biological data that is collected and integrated from multiple public, online databases. METHODS 1. an in-depth analysis was conducted to assess the department's data requirements and map these requirements to the data available at CLSD 2. CLSD incorporated new data as necessary 3. SQL was written to generate tables that would replace the targeted manual processes 4. a DB2 client was installed in Medical and Molecular Genetics to establish remote access to CLSD 5. a graphic user interface (GUI) was designed and implemented in HTML/CGI 6. a PERL program was written to accept parameters from the web input form, submit queries to CLSD, and generate HTML-based output tables 7. validation, updates, and maintenance procedures were conducted after early prototype implementation RESULTS AND CONCLUSIONS This application resulted in a substantial increase in efficiency over the manual methods that were previously used for data collection. The application also allows research teams to update their data much more frequently. A high level of accuracy in the output tables was confirmed by a thorough validation process.
Application of Data Pipelining Technology in Cheminformatics and Bioinformatics
(2002-12) Mao, Linyong; Perry, Douglas G.
Data pipelining is the processing, analysis, and mining of large volumes of data through a branching network of computational steps. A data pipelining system consists of a collection of modular computational components and a network for streaming data between them. By defining a logical path for data through a network of computational components and configuring each component accordingly, a user can create a protocol to perform virtually any desired function with data and extract knowledge from them. A set of data pipelines were constructed to explore the relationship between the biodegradability and structural properties of halogenated aliphatic compounds in a data set in which each compound has one degradation rate and nine structure-derived properties. After training, the data pipeline was able to calculate the degradation rates of new compounds with a relatively accurate rate. A second set of data pipelines was generated to cluster new DNA sequences. The data pipelining technology was applied to identify a core sequence to represent a DNA cluster and construct the 95% confidence distance interval for the cluster. The result shows that 74% of the DNA sequences were correctly clustered and there was no false clustering.
Assessing the Potential Utility of a Virtual and Mixed/Augmented Reality System to Assist in Stroke Rehabilitation
Leventhal, Jeremy; McDaniel, Anna M.
Stroke is the number one cause of disability in the United States. This thesis summarizes current techniques and technologies for stroke rehabilitation and in addition, describes a revolutionary new concept and rehabilitation system, Visually Directed Intention (VDI), created by Dr. Jill Bolte Taylor (Indiana University). The purpose of this research is to determine the feasibility and potential of her system through comparative research and expert opinion. Dr. Taylor‟s rehabilitation system harnesses several technologies such as mixed reality, biofeedback, and game-like environments. Key concepts such as visualization, intention, motivation and repetition are also pivotal to her ideology. Specifically the system uses biofeedback, viewed through a mixed reality headset to motivate a user to utilize nerves and muscles he/she may have lost through experiencing a stroke. In order to properly identify and analyze current methods used in stroke rehabilitation, several subject matter experts (SME) at the University of Chicago‟s Rehabilitation Institute of Chicago (RIC) were interviewed. The SME provided useful critique on current stroke rehabilitation techniques, technologies and Dr. Taylor‟s innovative concept. Through a general qualitative interview, examining the SMEs research and actually experimenting with some of their technologies, meaningful insight into expert opinions on stroke rehabilitation technologies was obtained. After several detailed interviews at the RIC, the experts agreed that VDI is noble concept and has great potential. Although they had some specific comments about how to properly utilize the technologies involved, overall they believe the system encompasses Assessing the Potential Utility for a Mixed Reality System (vii) several exciting and motivating features that will significantly improve the rehabilitation process.
The Audio Implicit Association Test: Human Preferences and Implicit Associations Concerning Machine Voices
Mitchell, Wade Joseph; MacDorman, Karl F.
Auditory human-machine interfaces are becoming ubiquitous. Interactive voice response systems, navigation systems, socially assistive robots, and smart houses are just a few examples of technologies that support auditory interactions. This study uses the implicit association test (IAT) to measure participants’ associative strength between human and machine voices and pleasant or unpleasant attributes. To accomplish this, the IAT needed to be validated using audio stimuli and the associative strength of secondary features of stimuli, that is, features other than their semantic content. Six IAT experiments were conducted to test the ability of the IAT to measure association strengths of the target concepts of audio stimuli and an attribute dimension in addition to target concepts of secondary features and an attribute dimension. Results support the effectiveness of an audio IAT, an IAT for secondary features, and an IAT that combines audio with secondary features. Results also show that participants had a stronger association between human voices and pleasant attributes than machine voices and pleasant attributes.

Browse

Browsing Informatics Theses and Dissertations by Title

Results Per Page

Sort Options