Usability problems associated with electronic health records can adversely impact clinical workflow, leading to inefficiencies, error, and even clinician burnout. The work presented in this dissertation is concerned with understanding and improving clinical workflow. Towards that end, it is necessary to model physical and cognitive aspects of task performance in clinical settings. Task completion can be significantly impacted by the navigational efficiency of the electronic health record (EHR) interface. Workflow modeling of the EHR-mediated workflow could help identify, diagnose and eliminate problems to reduce navigational complexity. The research goal is to introduce and validate a new biomedical informatics methodological workflow analysis framework that combines expert-based and user-based techniques to guide effective EHR design and reduce navigational complexity. These techniques are combined into a modified walkthrough that aligns user goals and subgoals with estimated task completion time and characterization of cognitive demands. A two-phased validation of the framework is utilized. The first is applied to single EHR-mediated workflow tasks, medication reconciliation (MedRec), and medication administration records (MAR) to refine individual aspects of the framework. The second phase applied the framework to a pre/post EHR implementation comparative analysis of multiple workflows tasks. This validation provides evidence of the framework's applicability and feasibility across several sites, systems, and settings. Analysis of the steps executed within the interfaces involved to complete the medication administration and medication reconciliation and patient order management tasks have provided a basis for characterizing the complexities in EHR navigation. An implication of the work presented here is that small tractable changes in interface design may substantially improve EHR navigation, overall usability, and workflow. The navigational complexity framework enables scrutinizing the impact of different EHR interfaces on task performance and usability barriers across different sites, systems, and settings.

Unstructured texts containing biomedical information from sources such as electronic health records, scientific literature, discussion forums, and social media offer an opportunity to extract information for a wide range of applications in biomedical informatics. Building scalable and efficient pipelines for natural language processing and extraction of biomedical information plays an important role in the implementation and adoption of applications in areas such as public health. Advancements in machine learning and deep learning techniques have enabled rapid development of such pipelines. This dissertation presents entity extraction pipelines for two public health applications: virus phylogeography and pharmacovigilance. For virus phylogeography, geographical locations are extracted from biomedical scientific texts for metadata enrichment in the GenBank database containing 2.9 million virus nucleotide sequences. For pharmacovigilance, tools are developed to extract adverse drug reactions from social media posts to open avenues for post-market drug surveillance from non-traditional sources. Across these pipelines, high variance is observed in extraction performance among the entities of interest while using state-of-the-art neural network architectures. To explain the variation, linguistic measures are proposed to serve as indicators for entity extraction performance and to provide deeper insight into the domain complexity and the challenges associated with entity extraction. For both the phylogeography and pharmacovigilance pipelines presented in this work the annotated datasets and applications are open source and freely available to the public to foster further research in public health.

Clinical Decision Support (CDS) is primarily associated with alerts, reminders, order entry, rule-based invocation, diagnostic aids, and on-demand information retrieval. While valuable, these foci have been in production use for decades, and do not provide a broader, interoperable means of plugging structured clinical knowledge into live electronic health record (EHR) ecosystems for purposes of orchestrating the user experiences of patients and clinicians. To date, the gap between knowledge representation and user-facing EHR integration has been considered an “implementation concern” requiring unscalable manual human efforts and governance coordination. Drafting a questionnaire engineered to meet the specifications of the HL7 CDS Knowledge Artifact specification, for example, carries no reasonable expectation that it may be imported and deployed into a live system without significant burdens. Dramatic reduction of the time and effort gap in the research and application cycle could be revolutionary. Doing so, however, requires both a floor-to-ceiling precoordination of functional boundaries in the knowledge management lifecycle, as well as formalization of the human processes by which this occurs.

This research introduces ARTAKA: Architecture for Real-Time Application of Knowledge Artifacts, as a concrete floor-to-ceiling technological blueprint for both provider heath IT (HIT) and vendor organizations to incrementally introduce value into existing systems dynamically. This is made possible by service-ization of curated knowledge artifacts, then injected into a highly scalable backend infrastructure by automated orchestration through public marketplaces. Supplementary examples of client app integration are also provided. Compilation of knowledge into platform-specific form has been left flexible, in so far as implementations comply with ARTAKA’s Context Event Service (CES) communication and Health Services Platform (HSP) Marketplace service packaging standards.

Towards the goal of interoperable human processes, ARTAKA’s treatment of knowledge artifacts as a specialized form of software allows knowledge engineers to operate as a type of software engineering practice. Thus, nearly a century of software development processes, tools, policies, and lessons offer immediate benefit: in some cases, with remarkable parity. Analyses of experimentation is provided with guidelines in how choice aspects of software development life cycles (SDLCs) apply to knowledge artifact development in an ARTAKA environment.

Portions of this culminating document have been further initiated with Standards Developing Organizations (SDOs) intended to ultimately produce normative standards, as have active relationships with other bodies.

The Morningside Initiative is a public-private activity that has evolved from an August, 2007, meeting at the Morningside Inn, in Frederick, MD, sponsored by the Telemedicine and Advanced Technology Research Center (TATRC) of the US Army Medical Research Materiel Command. Participants were subject matter experts in clinical decision support (CDS) and included representatives from the Department of Defense, Veterans Health Administration, Kaiser Permanente, Partners Healthcare System, Henry Ford Health System, Arizona State University, and the American Medical Informatics Association (AMIA). The Morningside Initiative was convened in response to the AMIA Roadmap for National Action on Clinical Decision Support and on the basis of other considerations and experiences of the participants. Its formation was the unanimous recommendation of participants at the 2007 meeting which called for creating a shared repository of executable knowledge for diverse health care organizations and practices, as well as health care system vendors.

The rationale is based on the recognition that sharing of clinical knowledge needed for CDS across organizations is currently virtually non-existent, and that, given the considerable investment needed for creating, maintaining and updating authoritative knowledge, which only larger organizations have been able to undertake, this is an impediment to widespread adoption and use of CDS. The Morningside Initiative intends to develop and refine (1) an organizational framework, (2) a technical approach, and (3) CDS content acquisition and management processes for sharing CDS knowledge content, tools, and experience that will scale with growing numbers of participants and can be expanded in scope of content and capabilities. Intermountain Healthcare joined the initial set of participants shortly after its formation. The efforts of the Morningside Initiative are intended to serve as the basis for a series of next steps in a national agenda for CDS. It is based on the belief that sharing of knowledge can be highly effective as is the case in other competitive domains such as genomics. Participants in the Morningside Initiative believe that a coordinated effort between the private and public sectors is needed to accomplish this goal and that a small number of highly visible and respected health care organizations in the public and private sector can lead by example. Ultimately, a future collaborative knowledge sharing organization must have a sustainable long-term business model for financial support.

Social media is becoming increasingly popular as a platform for sharing personal health-related information. This information can be utilized for public health monitoring tasks such as pharmacovigilance via the use of Natural Language Processing (NLP) techniques. One of the critical steps in information extraction pipelines is Named Entity Recognition (NER), where the mentions of entities such as diseases are located in text and their entity type are identified. However, the language in social media is highly informal, and user-expressed health-related concepts are often non-technical, descriptive, and challenging to extract. There has been limited progress in addressing these challenges, and advanced machine learning-based NLP techniques have been underutilized. This work explores the effectiveness of different machine learning techniques, and particularly deep learning, to address the challenges associated with extraction of health-related concepts from social media. Deep learning has recently attracted a lot of attention in machine learning research and has shown remarkable success in several applications particularly imaging and speech recognition. However, thus far, deep learning techniques are relatively unexplored for biomedical text mining and, in particular, this is the first attempt in applying deep learning for health information extraction from social media.

This work presents ADRMine that uses a Conditional Random Field (CRF) sequence tagger for extraction of complex health-related concepts. It utilizes a large volume of unlabeled user posts for automatic learning of embedding cluster features, a novel application of deep learning in modeling the similarity between the tokens. ADRMine significantly improved the medical NER performance compared to the baseline systems.

This work also presents DeepHealthMiner, a deep learning pipeline for health-related concept extraction. Most of the machine learning methods require sophisticated task-specific manual feature design which is a challenging step in processing the informal and noisy content of social media. DeepHealthMiner automatically learns classification features using neural networks and utilizing a large volume of unlabeled user posts. Using a relatively small labeled training set, DeepHealthMiner could accurately identify most of the concepts, including the consumer expressions that were not observed in the training data or in the standard medical lexicons outperforming the state-of-the-art baseline techniques.

Text mining of biomedical literature and clinical notes is a very active field of research in biomedical science. Semantic analysis is one of the core modules for different Natural Language Processing (NLP) solutions. Methods for calculating semantic relatedness of two concepts can be very useful in solutions solving different problems such as relationship extraction, ontology creation and question / answering [1–6]. Several techniques exist in calculating semantic relatedness of two concepts. These techniques utilize different knowledge sources and corpora. So far, researchers attempted to find the best hybrid method for each domain by combining semantic relatedness techniques and data sources manually. In this work, attempts were made to eliminate the needs for manually combining semantic relatedness methods targeting any new contexts or resources through proposing an automated method, which attempted to find the best combination of semantic relatedness techniques and resources to achieve the best semantic relatedness score in every context. This may help the research community find the best hybrid method for each context considering the available algorithms and resources.

Colorectal cancer is the second-highest cause of cancer-related deaths in the United States with approximately 50,000 estimated deaths in 2015. The advanced stages of colorectal cancer has a poor five-year survival rate of 10%, whereas the diagnosis in early stages of development has showed a more favorable five-year survival rate of 90%. Early diagnosis of colorectal cancer is achievable if colorectal polyps, a possible precursor to cancer, are detected and removed before developing into malignancy.

The preferred method for polyp detection and removal is optical colonoscopy. A colonoscopic procedure consists of two phases: (1) insertion phase during which a flexible endoscope (a flexible tube with a tiny video camera at the tip) is advanced via the anus and then gradually to the end of the colon--called the cecum, and (2) withdrawal phase during which the endoscope is gradually withdrawn while colonoscopists examine the colon wall to find and remove polyps. Colonoscopy is an effective procedure and has led to a significant decline in the incidence and mortality of colon cancer. However, despite many screening and therapeutic advantages, 1 out of every 4 polyps and 1 out of 13 colon cancers are missed during colonoscopy.

There are many factors that contribute to missed polyps and cancers including poor colon preparation, inadequate navigational skills, and fatigue. Poor colon preparation results in a substantial portion of colon covered with fecal content, hindering a careful examination of the colon. Inadequate navigational skills can prevent a colonoscopist from examining hard-to-reach regions of the colon that may contain a polyp. Fatigue can manifest itself in the performance of a colonoscopist by decreasing diligence and vigilance during procedures. Lack of vigilance may prevent a colonoscopist from detecting the polyps that briefly appear in the colonoscopy videos. Lack of diligence may result in hasty examination of the colon that is likely to miss polyps and lesions.

To reduce polyp and cancer miss rates, this research presents a quality assurance system with 3 components. The first component is an automatic polyp detection system that highlights the regions with suspected polyps in colonoscopy videos. The goal is to encourage more vigilance during procedures. The suggested polyp detection system consists of several novel modules: (1) a new patch descriptor that characterizes image appearance around boundaries more accurately and more efficiently than widely-used patch descriptors such HoG, LBP, and Daisy; (2) A 2-stage classification framework that is able to enhance low level image features prior to classification. Unlike the traditional way of image classification where a single patch undergoes the processing pipeline, our system fuses the information extracted from a pair of patches for more accurate edge classification; (3) a new vote accumulation scheme that robustly localizes objects with curvy boundaries in fragmented edge maps. Our voting scheme produces a probabilistic output for each polyp candidate but unlike the existing methods (e.g., Hough transform) does not require any predefined parametric model of the object of interest; (4) and a unique three-way image representation coupled with convolutional neural networks (CNNs) for classifying the polyp candidates. Our image representation efficiently captures a variety of features such as color, texture, shape, and temporal information and significantly improves the performance of the subsequent CNNs for candidate classification. This contrasts with the exiting methods that mainly rely on a subset of the above image features for polyp detection. Furthermore, this research is the first to investigate the use of CNNs for polyp detection in colonoscopy videos.

The second component of our quality assurance system is an automatic image quality assessment for colonoscopy. The goal is to encourage more diligence during procedures by warning against hasty and low quality colon examination. We detect a low quality colon examination by identifying a number of consecutive non-informative frames in videos. We base our methodology for detecting non-informative frames on two key observations: (1) non-informative frames

most often show an unrecognizable scene with few details and blurry edges and thus their information can be locally compressed in a few Discrete Cosine Transform (DCT) coefficients; however, informative images include much more details and their information content cannot be summarized by a small subset of DCT coefficients; (2) information content is spread all over the image in the case of informative frames, whereas in non-informative frames, depending on image artifacts and degradation factors, details may appear in only a few regions. We use the former observation in designing our global features and the latter in designing our local image features. We demonstrated that the suggested new features are superior to the existing features based on wavelet and Fourier transforms.

The third component of our quality assurance system is a 3D visualization system. The goal is to provide colonoscopists with feedback about the regions of the colon that have remained unexamined during colonoscopy, thereby helping them improve their navigational skills. The suggested system is based on a new 3D reconstruction algorithm that combines depth and position information for 3D reconstruction. We propose to use a depth camera and a tracking sensor to obtain depth and position information. Our system contrasts with the existing works where the depth and position information are unreliably estimated from the colonoscopy frames. We conducted a use case experiment, demonstrating that the suggested 3D visualization system can determine the unseen regions of the navigated environment. However, due to technology limitations, we were not able to evaluate our 3D visualization system using a phantom model of the colon.

A core principle in multiple national quality improvement strategies is the engagement of chronically ill patients in the creation and execution of their treatment plans. Numerous initiatives are underway to use health information technology (HIT) to support patient engagement however the use of HIT and other factors such as health literacy may be significant barriers to engagement for older adults. This qualitative descriptive study sought to explore the ways that older adults with multi-morbidities engaged with their plan of care. Forty participants were recruited through multiple case sampling from two ambulatory cardiology practices. Participants were English-speaking, without a dementia-related diagnosis, and between the ages of 65 and 86. The older adults in this study performed many behaviors to engage in the plan of care, including acting in ways to support health, managing health-related information, attending routine visits with their doctors, and participating in treatment planning. A subset of patients engaged in active decision-making because of the point they were at in their chronic disease. At that cross roads, they expressed uncertainly over which road to travel. Two factors influenced the engagement of older adults: a relationship with the provider that met the patient's needs, and the distribution of a Meaningful Use clinical summary at the conclusion of the provider visit. Participants described the ways in which the clinical summary helped and hindered their understanding of the care plan.

Insights gained as a result of this study include an understanding of the discrepancies between what the healthcare system expects of patients and their actual behavior when it comes to the creation of a care plan and the ways in which they take care of their health. Further research should examine the ability of various factors to enhance patient engagement. For example, it may be useful to focus on ways to improve the clinical summary to enhance engagement with the care plan and meet standards for a health literate document. Recommendations for the improvement of the clinical summary are provided. Finally, this study explored potential reasons for the infrequent use of online health information by older adults including the trusting relationship they enjoyed with their cardiologist.

Medical students acquire and enhance their clinical skills using various available techniques and resources. As the health care profession has move towards team-based practice, students and trainees need to practice team-based procedures that involve timely management of clinical tasks and adequate communication with other members of the team. Such team-based procedures include surgical and clinical procedures, some of which are protocol-driven. Cost and time required for individual team-based training sessions, along with other factors, contribute to making the training complex and challenging. A great deal of research has been done on medically-focused collaborative virtual reality (VR)-based training for protocol-driven procedures as a cost-effective as well as time-efficient solution. Most VR-based simulators focus on training of individual personnel. The ones which focus on providing team training provide an interactive simulation for only a few scenarios in a collaborative virtual environment (CVE). These simulators are suited for didactic training for cognitive skills development. The training sessions in the simulators require the physical presence of mentors. The problem with this kind of system is that the mentor must be present at the training location (either physically or virtually) to evaluate the performance of the team (or an individual). Another issue is that there is no efficient methodology that exists to provide feedback to the trainees during the training session itself (formative feedback). Furthermore, they lack the ability to provide training in acquisition or improvement of psychomotor skills for the tasks that require force or touch feedback such as cardiopulmonary resuscitation (CPR). To find a potential solution to overcome some of these concerns, a novel training system was designed and developed that utilizes the integration of sensors into a CVE for time-critical medical procedures. The system allows the participants to simultaneously access the CVE and receive training from geographically diverse locations. The system is also able to provide real-time feedback and is also able to store important data during each training/testing session. Finally, this study also presents a generalizable collaborative team-training system that can be used across various team-based procedures in medical as well as non-medical domains.

Surgery as a profession requires significant training to improve both clinical decision making and psychomotor proficiency. In the medical knowledge domain, tools have been developed, validated, and accepted for evaluation of surgeons' competencies. However, assessment of the psychomotor skills still relies on the Halstedian model of apprenticeship, wherein surgeons are observed during residency for judgment of their skills. Although the value of this method of skills assessment cannot be ignored, novel methodologies of objective skills assessment need to be designed, developed, and evaluated that augment the traditional approach. Several sensor-based systems have been developed to measure a user's skill quantitatively, but use of sensors could interfere with skill execution and thus limit the potential for evaluating real-life surgery. However, having a method to judge skills automatically in real-life conditions should be the ultimate goal, since only with such features that a system would be widely adopted. This research proposes a novel video-based approach for observing surgeons' hand and surgical tool movements in minimally invasive surgical training exercises as well as during laparoscopic surgery. Because our system does not require surgeons to wear special sensors, it has the distinct advantage over alternatives of offering skills assessment in both learning and real-life environments. The system automatically detects major skill-measuring features from surgical task videos using a computing system composed of a series of computer vision algorithms and provides on-screen real-time performance feedback for more efficient skill learning. Finally, the machine-learning approach is used to develop an observer-independent composite scoring model through objective and quantitative measurement of surgical skills. To increase effectiveness and usability of the developed system, it is integrated with a cloud-based tool, which automatically assesses surgical videos upload to the cloud.