Documenting the de-identification process of clinical and imaging data for AI for health imaging projects.

Artificial intelligence (AI) is revolutionizing the field of medical imaging, holding the potential to shift medicine from a reactive "sick-care" approach to a proactive focus on healthcare and prevention. The successful development of AI in this domain relies on access to large, comprehensive, and standardized real-world datasets that accurately represent diverse populations and diseases. However, images and data are sensitive, and as such, before using them in any way the data needs to be modified to protect the privacy of the patients. This paper explores the approaches in the domain of five EU projects working on the creation of ethically compliant and GDPR-regulated European medical imaging platforms, focused on cancer-related data. It presents the individual approaches to the de-identification of imaging data, and describes the problems and the solutions adopted in each case. Further, lessons learned are provided, enabling future projects to optimally handle the problem of data de-identification. CRITICAL RELEVANCE STATEMENT: This paper presents key approaches from five flagship EU projects for the de-identification of imaging and clinical data offering valuable insights and guidelines in the domain. KEY POINTS: ΑΙ models for health imaging require access to large amounts of data. Access to large imaging datasets requires an appropriate de-identification process. This paper provides de-identification guidelines from the AI for health imaging (AI4HI) projects.

Socioeconomic Determinants of Remote Patient Monitoring Implementation Among Rural and Urban Hospitals.

Introduction: Remote patient monitoring (RPM) is a form of telehealth that improves quality of care for chronic disease treatment and reduces hospital readmission rates. Geographical proximity to health care is important for individuals of low socioeconomic status (SES) who face additional financial and transportation barriers. The goal of this study was to assess the association between social determinants of health and adoption of RPM. Methods: This cross-sectional study analyzed data from hospitals that responded to the American Hospital Association's Annual Survey (2018) and spatially linked census tract-level environmental and social determinants of health obtained from the Social Vulnerability Index (2018). Results: A total of 4,206 hospitals (1,681 rural and 2,525 urban hospitals) met study criteria. Rural hospitals near households in the lower middle quartile SES were associated with a 33.5% lower likelihood of having adopted RPM for chronic care management compared with rural hospitals near households in the highest quartile SES (adjusted odds ratios [aOR] = 0.665; 95% confidence interval [CI]: 0.453-0.977). Urban hospitals near households in the lowest quartile SES were associated with a 41.9% lower likelihood of having adopted RPM for chronic care management compared with urban hospitals near households in the highest quartile SES (aOR = 0.581; 95% CI: 0.435-0.775). Similar trends in accessibility were found with RPM for postdischarge services among urban hospitals. Conclusion: Our findings highlight the importance of hospital responsibility and state and federal policy approaches toward ensuring equitable access to RPM services for patients characterized by lower SES.

Analysis of Caregiver Burden Expressed in Social Media Discussions.

Almost 40% of US adults provide informal caregiving, yet research gaps remain around what burdens affect informal caregivers. This study uses a novel social media site, Reddit, to mine and better understand what online communities focus on as their caregiving burdens. These forums were accessed using an application programming interface, a machine learning classifier was developed to remove low information posts, and topic modeling was applied to the corpus. An expert panel summarized the forums' themes into ten categories. The largest theme extracted from Reddit's forums discussed the personal emotional toll of being a caregiver. This was followed by logistic issues while caregiving and caring for parents who have cancer. Smaller themes included approaches to end-of-life care, physical equipment needs when caregiving, and the use of wearables or technology to help monitor care recipients. The platform often discusses caregiving for parents which may reflect the age of Reddit's users. This study confirms that Reddit forums are used for caregivers to discuss the burdens associated with their role and the types of stress that can result from informal caregiving.

The informed consent form navigator: a tool for producing readable and compliant consent documents.

Background/Objective: Informed consent forms (ICFs) and practices vary widely across institutions. This project expands on previous work at the University of Arkansas for Medical Sciences (UAMS) Center for Health Literacy to develop a plain language ICF template. Our interdisciplinary team of researchers, comprised of biomedical informaticists, health literacy experts, and stakeholders in the Institutional Review Board (IRB) process, has developed the ICF Navigator, a novel tool to facilitate the creation of plain language ICFs that comply with all relevant regulatory requirements. Methods: Our team first developed requirements for the ICF Navigator tool. The tool was then implemented by a technical team of informaticists and software developers, in consultation with an informed consent legal expert. We developed and formalized a detailed knowledge map modeling regulatory requirements for ICFs, which drives workflows within the tool. Results: The ICF Navigator is a web-based tool that guides researchers through creating an ICF as they answer questions about their project. The navigator uses those responses to produce a clear and compliant ICF, displaying a real-time preview of the final form as content is added. Versioning and edits can be tracked to facilitate collaborative revisions by the research team and communication with the IRB. The navigator helps guide the creation of study-specific language, ensures compliance with regulatory requirements, and ensures that the resulting ICF is easy to read and understand. Conclusion: The ICF Navigator is an innovative, customizable, open-source software tool that helps researchers produce custom readable and compliant ICFs for research studies involving human subjects.

Evaluating the temporal association between the recency of prescribed controlled substance acquisition and fatal and non-fatal opioid overdose.

BACKGROUND: Previous studies have explored psychosocial effects as possible triggers of opioid overdose (OOD). However, little is known about the temporal association between OOD and prescribed controlled substance (CS) acquisition. OBJECTIVE: The objective of this study was to evaluate the temporal relationship between OOD and acquiring prescribed CSs prior to OOD. METHODS: This study is an exploratory descriptive analysis using Arkansas Prescription Drug Monitoring Program (AR-PDMP) data linked to death certificate and statewide inpatient discharge records. All persons with ≥1 AR-PDMP prescription fill(s) between 1 January 2014 and 31 December 2017 were included (n = 1,946,686). For persons that experienced OOD and had ≥1 PDMP record(s), the difference in days between OOD and the most recent AR-PDMP prescription filled prior to an OOD was recorded. To account for censoring, a sensitivity analysis was conducted restricting the study group to "New AR-PDMP Entrants" that had at least a 180-day gap between consecutive AR-PDMP fill dates. RESULTS: 28,998,307 AR-PDMP records were analyzed for 1,946,686 individuals. 7195 persons experienced 9223 OODs and 414 (4.49%) of those were fatal. Of these, 6236 experienced ≥1 OOD and acquired prescribed CSs prior to or on the day of the first OOD. Of those that experienced ≥1 OOD(s), 2201 (30.59%) had an AR-PDMP record in the 0- to 5-day period prior to their overdose and 497 (6.91%) had an AR-PDMP record the day prior to their overdose. Among New AR-PDMP Entrants that experienced ≥1 OOD(s), 408 (27.38%) had an AR-PDMP record in the 0- to 5-day period prior to their overdose. CONCLUSION: Though the vast majority of persons accessing CSs in Arkansas did not experience an OOD, a sizable proportion of persons that experience an OOD(s) obtained prescribed CSs immediately prior.

The PRISM semantic cohort builder: a novel tool to search and access clinical data in TCIA imaging collections.

The cancer imaging archive (TICA) receives and manages an ever-increasing quantity of clinical (non-image) data containing valuable information about subjects in imaging collections. To harmonize and integrate these data, we have first cataloged the types of information occurring across public TCIA collections. We then produced mappings for these diverse instance data using ontology-based representation patterns and transformed the data into a knowledge graph in a semantic database. This repository combined the transformed instance data with relevant background knowledge from domain ontologies. The resulting repository of semantically integrated data is a rich source of information about subjects that can be queried across imaging collections. Building on this work we have implemented and deployed a REST API and a user-facing semantic cohort builder tool. This tool allows allow researchers and other users to search and identify groups of subject-level records based on non-image data that were not queryable prior to this work. The search results produced by this interface link to images, allowing users to quickly identify and view images matching the selection criteria, as well as allowing users to export the harmonized clinical data.

Characteristics and Network Influence of Providers Involved in the Treatment of Patients With Chronic Back, Neck or Joint Pain in Arkansas.

Increasing emphasis on guidelines and prescription drug monitoring programs highlight the role of healthcare providers in pain treatment. Objectives of this study were to identify characteristics of key players and influence of opioid prescribers through construction of a referral network of patients with chronic pain. A retrospective cohort study was performed and patients with commercial or Medicaid coverage with chronic back, neck, or joint pain were identified using the Arkansas All-Payer Claims-Database. A social network comprised of providers connected by patient referrals based on 12-months of healthcare utilization following chronic pain was constructed. Network measures evaluated were indegree and eigen (referrals obtained), betweenness (involvement), and closeness centrality (reach). Outcomes included influence of providers, opioid prescribers, and brokerage status. Exposures included provider demographics, specialties and network characteristics. There were 51,941 chronic pain patients who visited 8,110 healthcare providers. Primary care providers showed higher betweenness and closeness whereas specialists had higher indegree. Opioid providers showed higher centrality compared to non-opioid providers, which decreased with increasing volume of opioid prescribing. Non-pharmacologic providers showed significant brokerage scores. Findings from this study such as primary care providers having better reach, non-central positions of high-volume prescribers and non-pharmacologic providers having higher brokerage can aid interventional physician detailing. PERSPECTIVE: Opioid providers held central positions in the network aiding provider-directed interventions. However, high-volume opioid providers were at the borders making them difficult targets for interventions. Primary care providers had the highest reach, specialists received the most referrals and non-pharmacological providers and specialists acted as brokers between non-opioid and opioid prescribers.

API Driven On-Demand Participant ID Pseudonymization in Heterogeneous Multi-Study Research.

OBJECTIVES: To facilitate clinical and translational research, imaging and non-imaging clinical data from multiple disparate systems must be aggregated for analysis. Study participant records from various sources are linked together and to patient records when possible to address research questions while ensuring patient privacy. This paper presents a novel tool that pseudonymizes participant identifiers (PIDs) using a researcher-driven automated process that takes advantage of application-programming interface (API) and the Perl Open-Source Digital Imaging and Communications in Medicine Archive (POSDA) to further de-identify PIDs. The tool, on-demand cohort and API participant identifier pseudonymization (O-CAPP), employs a pseudonymization method based on the type of incoming research data. METHODS: For images, pseudonymization of PIDs is done using API calls that receive PIDs present in Digital Imaging and Communications in Medicine (DICOM) headers and returns the pseudonymized identifiers. For non-imaging clinical research data, PIDs provided by study principal investigators (PIs) are pseudonymized using a nightly automated process. The pseudonymized PIDs (P-PIDs) along with other protected health information is further de-identified using POSDA. RESULTS: A sample of 250 PIDs pseudonymized by O-CAPP were selected and successfully validated. Of those, 125 PIDs that were pseudonymized by the nightly automated process were validated by multiple clinical trial investigators (CTIs). For the other 125, CTIs validated radiologic image pseudonymization by API request based on the provided PID and P-PID mappings. CONCLUSIONS: We developed a novel approach of an ondemand pseudonymization process that will aide researchers in obtaining a comprehensive and holistic view of study participant data without compromising patient privacy.

Semantic Integration of Multi-Modal Data and Derived Neuroimaging Results Using the Platform for Imaging in Precision Medicine (PRISM) in the Arkansas Imaging Enterprise System (ARIES).

Neuroimaging is among the most active research domains for the creation and management of open-access data repositories. Notably lacking from most data repositories are integrated capabilities for semantic representation. The Arkansas Imaging Enterprise System (ARIES) is a research data management system which features integrated capabilities to support semantic representations of multi-modal data from disparate sources (imaging, behavioral, or cognitive assessments), across common image-processing stages (preprocessing steps, segmentation schemes, analytic pipelines), as well as derived results (publishable findings). These unique capabilities ensure greater reproducibility of scientific findings across large-scale research projects. The current investigation was conducted with three collaborating teams who are using ARIES in a project focusing on neurodegeneration. Datasets included magnetic resonance imaging (MRI) data as well as non-imaging data obtained from a variety of assessments designed to measure neurocognitive functions (performance scores on neuropsychological tests). We integrate and manage these data with semantic representations based on axiomatically rich biomedical ontologies. These instantiate a knowledge graph that combines the data from the study cohorts into a shared semantic representation that explicitly accounts for relations among the entities that the data are about. This knowledge graph is stored in a triple-store database that supports reasoning over and querying these integrated data. Semantic integration of the non-imaging data using background information encoded in biomedical domain ontologies has served as a key feature-engineering step, allowing us to combine disparate data and apply analyses to explore associations, for instance, between hippocampal volumes and measures of cognitive functions derived from various assessment instruments.

DICOM re-encoding of volumetrically annotated Lung Imaging Database Consortium (LIDC) nodules.

PURPOSE: The dataset contains annotations for lung nodules collected by the Lung Imaging Data Consortium and Image Database Resource Initiative (LIDC) stored as standard DICOM objects. The annotations accompany a collection of computed tomography (CT) scans for over 1000 subjects annotated by multiple expert readers, and correspond to "nodules ≥ 3 mm", defined as any lesion considered to be a nodule with greatest in-plane dimension in the range 3-30 mm regardless of presumed histology. The present dataset aims to simplify reuse of the data with the readily available tools, and is targeted towards researchers interested in the analysis of lung CT images. ACQUISITION AND VALIDATION METHODS: Open source tools were utilized to parse the project-specific XML representation of LIDC-IDRI annotations and save the result as standard DICOM objects. Validation procedures focused on establishing compliance of the resulting objects with the standard, consistency of the data between the DICOM and project-specific representation, and evaluating interoperability with the existing tools. DATA FORMAT AND USAGE NOTES: The dataset utilizes DICOM Segmentation objects for storing annotations of the lung nodules, and DICOM Structured Reporting objects for communicating qualitative evaluations (nine attributes) and quantitative measurements (three attributes) associated with the nodules. The total of 875 subjects contain 6859 nodule annotations. Clustering of the neighboring annotations resulted in 2651 distinct nodules. The data are available in TCIA at https://doi.org/10.7937/TCIA.2018.h7umfurq. POTENTIAL APPLICATIONS: The standardized dataset maintains the content of the original contribution of the LIDC-IDRI consortium, and should be helpful in developing automated tools for characterization of lung lesions and image phenotyping. In addition to those properties, the representation of the present dataset makes it more FAIR (Findable, Accessible, Interoperable, Reusable) for the research community, and enables its integration with other standardized data collections.

PRISM: A Platform for Imaging in Precision Medicine.

PURPOSE: Precision medicine requires an understanding of individual variability, which can only be acquired from large data collections such as those supported by the Cancer Imaging Archive (TCIA). We have undertaken a program to extend the types of data TCIA can support. This, in turn, will enable TCIA to play a key role in precision medicine research by collecting and disseminating high-quality, state-of-the-art, quantitative imaging data that meet the evolving needs of the cancer research community. METHODS: A modular technology platform is presented that would allow existing data resources, such as TCIA, to evolve into a comprehensive data resource that meets the needs of users engaged in translational research for imaging-based precision medicine. This Platform for Imaging in Precision Medicine (PRISM) helps streamline the deployment and improve TCIA's efficiency and sustainability. More importantly, its inherent modular architecture facilitates a piecemeal adoption by other data repositories. RESULTS: PRISM includes services for managing radiology and pathology images and features and associated clinical data. A semantic layer is being built to help users explore diverse collections and pool data sets to create specialized cohorts. PRISM includes tools for image curation and de-identification. It includes image visualization and feature exploration tools. The entire platform is distributed as a series of containerized microservices with representational state transfer interfaces. CONCLUSION: PRISM is helping modernize, scale, and sustain the technology stack that powers TCIA. Repositories can take advantage of individual PRISM services such as de-identification and quality control. PRISM is helping scale image informatics for cancer research at a time when the size, complexity, and demands to integrate image data with other precision medicine data-intensive commons are mounting.

Enhancing the drug ontology with semantically-rich representations of National Drug Codes and RxNorm unique concept identifiers.

BACKGROUND: The Drug Ontology (DrOn) is a modular, extensible ontology of drug products, their ingredients, and their biological activity created to enable comparative effectiveness and health services researchers to query National Drug Codes (NDCs) that represent products by ingredient, by molecular disposition, by therapeutic disposition, and by physiological effect (e.g., diuretic). It is based on the RxNorm drug terminology maintained by the U.S. National Library of Medicine, and on the Chemical Entities of Biological Interest ontology. Both national drug codes (NDCs) and RxNorm unique concept identifiers (RXCUIS) can undergo changes over time that can obfuscate their meaning when these identifiers occur in historic data. We present a new approach to modeling these entities within DrOn that will allow users of DrOn working with historic prescription data to more easily and correctly interpret that data. RESULTS: We have implemented a full accounting of national drug codes and RxNorm unique concept identifiers as information content entities, and of the processes involved in managing their creation and changes. This includes an OWL file that implements and defines the classes necessary to model these entities. A separate file contains an instance-level prototype in OWL that demonstrates the feasibility of this approach to representing NDCs and RXCUIs and the processes of managing them by retrieving and representing several individual NDCs, both active and inactive, and the RXCUIs to which they are connected. We also demonstrate how historic information about these identifiers in DrOn can be easily retrieved using a simple SPARQL query. CONCLUSIONS: An accurate model of how these identifiers operate in reality is a valuable addition to DrOn that enhances its usefulness as a knowledge management resource for working with historic data.

Enhancing Clinical Data and Clinical Research Data with Biomedical Ontologies - Insights from the Knowledge Representation Perspective.

OBJECTIVES: There exists a communication gap between the biomedical informatics community on one side and the computer science/artificial intelligence community on the other side regarding the meaning of the terms "semantic integration" and "knowledge representation". This gap leads to approaches that attempt to provide one-to-one mappings between data elements and biomedical ontologies. Our aim is to clarify the representational differences between traditional data management and semantic-web-based data management by providing use cases of clinical data and clinical research data re-representation. We discuss how and why one-to-one mappings limit the advantages of using Semantic Web Technologies (SWTs). METHODS: We employ commonly used SWTs, such as Resource Description Framework (RDF) and Ontology Web Language (OWL). We reuse pre-existing ontologies and ensure shared ontological commitment by selecting ontologies from a framework that fosters community-driven collaborative ontology development for biomedicine following the same set of principles. RESULTS: We demonstrate the results of providing SWT-compliant re-representation of data elements from two independent projects managing clinical data and clinical research data. Our results show how one-to-one mappings would hinder the exploitation of the advantages provided by using SWT. CONCLUSIONS: We conclude that SWT-compliant re-representation is an indispensable step, if using the full potential of SWT is the goal. Rather than providing one-to-one mappings, developers should provide documentation that links data elements to graph structures to specify the re-representation.

The Role of Axiomatically-Rich Ontologies in Transforming Medical Data to Knowledge.

In the biomedical domain, there exist a number of common data models (CDM) that have experienced wide uptake. However, none of these has emerged as the common model. Recently, the demand for integrating and analyzing increasingly large data sets in clinical and translational research has led to numerous efforts to harmonize existing CDMs and integrate data curated based on those models. These efforts raise the question of how to appropriately represent the semantics of data, and, furthermore, they highlight the fact that quite often different groups have greatly different definitions of 'semantics'. The question of how to formally assure that mappings between CDMs are correct is often overlooked. The answer to these challenges lies in using axiomatically-rich ontologies that allow verifying that terms refer to the same set of entities using automatic inference. This verification is only possible by building ontologies that represent the content of the scientific disciplines in accordance with the reality of the domain of the disciplines. Organizing and managing the development of numerous orthogonal domain-specific ontologies would benefit from using an Architecture Reference Model, that helps keeping the relationships consistent within each domain and ensure that appropriate inter-domain relationships are defined. This paper will explore how a strong logical representation of the scientific domain does not only foster harmonization of CDMs, but also informs and facilitates the transition from data over information to knowledge.

Mismatches between major subhierarchies and semantic tags in SNOMED CT.

The fully specified name of a concept in SNOMED CT is formed by a term to which in the typical case is added a semantic tag. The latter is meant to disambiguate homonymous terms and to indicate in which major subhierarchy of SNOMED CT that concept fits. We have developed a method to determine whether a concept's tag correctly identifies its place in the hierarchy, and applied this method to an analysis of all active concepts in every SNOMED CT release from January 2003 to January 2017. Our results show (1) that there are concepts in almost every release whose semantic tag does not match their placement in the hierarchy, (2) that it is primarily disorder concepts that are involved, and (3) that the number of such mismatches increase since the July 2012 version. Our analysis determined that it is primarily the absence of a mechanism in the SNOMED CT authoring environment to suggest stated relationships for very similar concepts that is responsible for the mismatches. We argue that the SNOMED CT authoring environment should treat the semantic tags as part of the formal structure so that methods can be implemented to keep the sub-hierarchies in sync with the semantic tags.

Protein Ontology (PRO): enhancing and scaling up the representation of protein entities.

The Protein Ontology (PRO; http://purl.obolibrary.org/obo/pr) formally defines and describes taxon-specific and taxon-neutral protein-related entities in three major areas: proteins related by evolution; proteins produced from a given gene; and protein-containing complexes. PRO thus serves as a tool for referencing protein entities at any level of specificity. To enhance this ability, and to facilitate the comparison of such entities described in different resources, we developed a standardized representation of proteoforms using UniProtKB as a sequence reference and PSI-MOD as a post-translational modification reference. We illustrate its use in facilitating an alignment between PRO and Reactome protein entities. We also address issues of scalability, describing our first steps into the use of text mining to identify protein-related entities, the large-scale import of proteoform information from expert curated resources, and our ability to dynamically generate PRO terms. Web views for individual terms are now more informative about closely-related terms, including for example an interactive multiple sequence alignment. Finally, we describe recent improvement in semantic utility, with PRO now represented in OWL and as a SPARQL endpoint. These developments will further support the anticipated growth of PRO and facilitate discoverability of and allow aggregation of data relating to protein entities.

Ontological Foundations for Tracking Data Quality through the Internet of Things.

Amongst the positive outcomes expected from the Internet of Things for Health are longitudinal patient records that are more complete and less erroneous by complementing manual data entry with automatic data feeds from sensors. Unfortunately, devices are fallible too. Quality control procedures such as inspection, testing and maintenance can prevent devices from producing errors. The additional approach envisioned here is to establish constant data quality monitoring through analytics procedures on patient data that exploit not only the ontological principles ascribed to patients and their bodily features, but also to observation and measurement processes in which devices and patients participate, including the, perhaps erroneous, representations that are generated. Using existing realism-based ontologies, we propose a set of categories that analytics procedures should be able to reason with and highlight the importance of unique identification of not only patients, caregivers and devices, but of everything involved in those measurements. This approach supports the thesis that the majority of what tends to be viewed as 'metadata' are actually data about first-order entities.

Analyzing SNOMED CT's Historical Data: Pitfalls and Possibilities.

SNOMED CT's Release Format 2 (RF2) has been announced as an improvement over its predecessor, for instance because of its more consistent and almost formal approach towards describing changes in components over different versions, as well as changes in the structure of SNOMED CT itself. We explore two sorts of changes that are only partially formalized in RF2: the relationships between associative relations and reasons for inactivations as expressed in Association Reference Sets and Attribute Value Reference Sets on the one hand, and the various patterns according to which semantic tags appearing in fully specified names change over subsequent versions with or without being related to inactivations. We propose a data conversion methodology that combines assertions about SNOMED CT components into history profiles and use elements of these profiles to build Formal Concept Analysis contexts to discover valid implications that can render implicit assumptions hidden in SNOMED CT's structure explicit.