Michigan State University

MSU Libraries

Digital Repository

Search results

(1 - 4 of 4)
Title
Contributions to machine learning in biomedical informatics
Creator
Baytas, Inci Meliha
Date
2019
Collection
Electronic Theses & Dissertations
Description
"With innovations in digital data acquisition devices and increased memory capacity, virtually all commercial and scientific domains have been witnessing an exponential growth in the amount of data they can collect. For instance, healthcare is experiencing a tremendous growth in digital patient information due to the high adaptation rate of electronic health record systems in hospitals. The abundance of data offers many opportunities to develop robust and versatile systems, as long as the...
Show more"With innovations in digital data acquisition devices and increased memory capacity, virtually all commercial and scientific domains have been witnessing an exponential growth in the amount of data they can collect. For instance, healthcare is experiencing a tremendous growth in digital patient information due to the high adaptation rate of electronic health record systems in hospitals. The abundance of data offers many opportunities to develop robust and versatile systems, as long as the underlying salient information in data can be captured. On the other hand, today's data, often named big data, is challenging to analyze due to its large scale and high complexity. For this reason, efficient data-driven techniques are necessary to extract and utilize the valuable information in the data. The field of machine learning essentially develops such techniques to learn effective models directly from the data. Machine learning models have been successfully employed to solve complicated real world problems. However, the big data concept has numerous properties that pose additional challenges in algorithm development. Namely, high dimensionality, class membership imbalance, non-linearity, distributed data, heterogeneity, and temporal nature are some of the big data characteristics that machine learning must address. Biomedical informatics is an interdisciplinary domain where machine learning techniques are used to analyze electronic health records (EHRs). EHR comprises digital patient data with various modalities and depicts an instance of big data. For this reason, analysis of digital patient data is quite challenging although it provides a rich source for clinical research. While the scale of EHR data used in clinical research might not be huge compared to the other domains, such as social media, it is still not feasible for physicians to analyze and interpret longitudinal and heterogeneous data of thousands of patients. Therefore, computational approaches and graphical tools to assist physicians in summarizing the underlying clinical patterns of the EHRs are necessary. The field of biomedical informatics employs machine learning and data mining approaches to provide the essential computational techniques to analyze and interpret complex healthcare data to assist physicians in patient diagnosis and treatment. In this thesis, we propose and develop machine learning algorithms, motivated by prevalent biomedical informatics tasks, to analyze the EHRs. Specifically, we make the following contributions: (i) A convex sparse principal component analysis approach along with variance reduced stochastic proximal gradient descent is proposed for the patient phenotyping task, which is defined as finding clinical representations for patient groups sharing the same set of diseases. (ii) An asynchronous distributed multi-task learning method is introduced to learn predictive models for distributed EHRs. (iii) A modified long-short term memory (LSTM) architecture is designed for the patient subtyping task, where the goal is to cluster patients based on similar progression pathways. The proposed LSTM architecture, T-LSTM, performs a subspace decomposition on the cell memory such that the short term effect in the previous memory is discounted based on the length of the time gap. (iv) An alternative approach to T-LSTM model is proposed with a decoupled memory to capture the short and long term changes. The proposed model, decoupled memory gated recurrent network (DM-GRN), is designed to learn two types of memories focusing on different components of the time series data. In this study, in addition to the healthcare applications, behavior of the proposed model is investigated for traffic speed prediction problem to illustrate its generalization ability. In summary, the aforementioned machine learning approaches have been developed to address complex characteristics of electronic health records in routine biomedical informatics tasks such as computational patient phenotyping and patient subtyping. Proposed models are also applicable to different domains with similar data characteristics as EHRs."--Pages ii-iii.
Show less
Title
An examination of the adoption of electronic medical records by rural hospital nurses through the unified theory of acceptance and use of technology lens
Creator
Holtz, Bree E.
Date
2010
Collection
Electronic Theses & Dissertations
Title
The use of health information exchange organizations in clinical research : current status, challenges and opportunities
Creator
Parker, Carol J. (Carol Jean)
Date
2016
Collection
Electronic Theses & Dissertations
Description
"Significant federal investment has led to the increased use of electronic health record (EHR) technology and electronic exchange of health information across health care providers. Health information exchange organizations (HIOs) are organizations that provide technology and infrastructure to enable electronic health information to be exchanged across disparate EHR technologies and between different health care provider organizations. While designed to support patient care delivery, this...
Show more"Significant federal investment has led to the increased use of electronic health record (EHR) technology and electronic exchange of health information across health care providers. Health information exchange organizations (HIOs) are organizations that provide technology and infrastructure to enable electronic health information to be exchanged across disparate EHR technologies and between different health care provider organizations. While designed to support patient care delivery, this technology has the potential to support clinical research and improve efficiencies for data collection including patient identification and data monitoring. This study sought to determine whether HIOs have the necessary infrastructure, technological capacity and agreements among participating providers to support research using exchanged clinical data; if HIOs facilitate the development of multi-institutional datasets that can be used for research; and whether the application of HIO data (Indiana Network for Patient Care, INPC) resulted in an accurate, representative, and comprehensive foundation for a specific research question (transitions of care in intracerebral hemorrhage, ICH, patients). Our scoping review to identify published studies that used HIOs as data sources for clinical research found that, outside of the evaluation of HIOs themselves, HIO data were being used to a limited extent in clinical research studies, with only a limited number of specific HIOs involved in generating the majority of the published research. We then used data from a national survey of HIOs to determine the extent HIOs report supporting research by allowing exchanged patient data to be aggregated and used for clinical, health services or epidemiologic research. We found that most HIOs reported supporting, or planning to support research, and that support for research is closely aligned with advanced technological infrastructure and functionality. This study culminated in the use of data from one HIO, the INPC, to study transitions of care for ICH patients. We found that the HIO's ability to provide sufficient data to study transitions of care was hindered by two problems: 1) missing clinical data among providers that share data with the INPC and the lack of participation in the INPC for several important post-acute care settings. Most notably, the INPC data did not include encounter information from hospice providers, free-standing acute rehabilitation facilities, skilled nursing facilities (nursing home), home health, or long term care hospitals. For some of these settings (e.g. skilled nursing facilities and home health), this is in part due to the slow implementation of electronic health record and exchange technologies. In addition, we found that encounters are collapsed into broad categories (inpatient, outpatient and emergency) that do not reflect the variety of clinical interactions in a way most useful to researchers and other analysts of healthcare delivery. As the rapid expansion in EHR use and health information exchange are relatively recent, HIO support for research is still developing. While we found limited utility of HIO data to study transitions of care for ICH patients, we only used data from one specific HIO. Additional research is required to determine whether HIOs are viable partners for research outside of the evaluation of HIOs themselves."--Pages ii-iii.
Show less
Title
Health media in the golden years : how American seniors are using media to communicate with healthcare product and service providers
Creator
Hamrick, Christopher D.
Date
2011
Collection
Electronic Theses & Dissertations
Description
ABSTRACTHEALTH MEDIA IN THE GOLDEN YEARS: HOW AMERICAN SENIORS ARE USING MEDIA TO COMMUNICATE WITH HEALTHCARE PRODUCT AND SERVICE PROVIDERSByChristopher D. HamrickThe American population is aging, and as they do their healthcare needs are growing. Americans are among the largest populations of mass media consumers in the world per capita. The Internet revolution has added an entirely new dimension to media production and consumption while altering or supplanting more traditional media....
Show moreABSTRACTHEALTH MEDIA IN THE GOLDEN YEARS: HOW AMERICAN SENIORS ARE USING MEDIA TO COMMUNICATE WITH HEALTHCARE PRODUCT AND SERVICE PROVIDERSByChristopher D. HamrickThe American population is aging, and as they do their healthcare needs are growing. Americans are among the largest populations of mass media consumers in the world per capita. The Internet revolution has added an entirely new dimension to media production and consumption while altering or supplanting more traditional media. Healthcare providers as well as product and service vendors are spending billions of dollars in advertising and informational campaigns as they vie for consumers while using new information and communication technologies to modernize healthcare delivery.Given the previous statements, is the money and effort invested by the U.S. healthcare industry into modern media technologies actually reaching their largest target audience? Are American seniors receptive to these innovations? This study analyzes the current trends in healthcare communications, the related information available about senior citizens and media consumption, and presents a survey of uses of and attitudes towards healthcare informatics, technology uses, and technology adoption among middle class independent seniors living in a single community. The analysis of this data is to inform the choices of those seeking to reach seniors to influence their healthcare media consumption.
Show less