Guest Column | October 7, 2020

Sharing Ontologies Globally To Speed Science And Healthcare Solutions

By Dave McComb, Semantic Arts

Global-iStock-1142610359

The COVID-19 pandemic is a clear example of how medical practitioners require swift access to enormous amounts of diverse information to efficaciously treat patients. They must synthesize individual data (vital signs, clinical history, demographics, and more) with rapidly evolving knowledge about COVID-19 and make decisions relevant to the conditions from which specific patients suffer.

Medical practitioners rely on point-of-care decision support systems to accelerate patient-care analysis and to scale treatments for intake quantities of global pandemics. They analyze a plethora of inputs to produce tailored treatment recommendations, in near real-time, which significantly enhance the quality of treatment.

Ontologies Create The Foundation For Complex Data Analysis

The underlying utility of these systems is widely based on the vast quantities of healthcare knowledge analyzed. Such knowledge must be uniformly represented (at scale) with rich, contextualized descriptions of the full scope of clinical trials, pharmaceutical information, and research germane to the biomedical field that expands daily with each published paper and new findings. This knowledge should be rapidly accessible, reusable, and a sturdy foundation on which to base present and future research in this field, encompassing everything from long-standing maladies like peanut allergies to emergent ones like COVID-19.

Ontologies—evolving conceptual data models with standardized concepts and uniquely fulfill each of these requirements to fuel healthcare research and point-of-care decision support systems, helping save lives when they need saving most.

International Ontology Sharing Is Becoming A Reality

A consortium of researchers recently formed an organization dedicated to standardizing how scientists define their ontologies, which are essential for retrieving datasets as well as understanding and reproducing research. The group called OntoPortal Alliance is creating a public repository of internationally shared domain-specific ontologies. All the repositories will be managed with a common OntoPortal appliance that has been tested with AllegroGraph Semantic Knowledge Graph software. This enables any OntoPortal adopter to get all the power, features, maintainability, and support benefits that come from using a widely adopted, state-of-the-art semantic knowledge graph database.

The first set of ontology repositories making up the OntoPortal Alliance include BioPortal (biomedical and other ontologies used internationally), SIFR (biomedical ontologies in the French language), BMICC MedPortal (biomedical ontologies focused on Chinese users),  AgroPortal (ontologies focused on agronomy and related sciences), and) EcoPortal (ontologies focused on environmental science. The OntoPortal Alliance will be adding more ontology repositories and is open to working with researchers in other domains who want to offer ontologies publicly.

Exchanging Biomedical Knowledge Through Ontologies

The recent COVID-19 pandemic underscores the need for researchers, scientists, and medical staff to rapidly share information in a standardized way to speed analysis and insights. BioPortal is perhaps the most comprehensive framework for quickly downloading ontologies in the biomedical field. Its advantages are academic and pragmatic, theoretical, and accomplished. Accessing its ontological resources enables researchers to intimately view the data—at the concept and terminology level—for an exhaustive corpus of biomedical research. They can then advance this research by leveraging these data in their studies, elevating the discipline.

As John Graybeal, Director of the OntoPortal Alliance and Technical Program Manager at Stanford Center for Biomedical Informatics Research pointed out, “By sharing biomedical ontologies, you don’t have 10 different groups calling the same thing by different names, and you don’t have people creating overlapping concepts and overlapping terms. The terminology creators can see how their terms relate to each other and they can translate those terms among their different terminologies.”

The knowledge contained in biomedical ontologies is invaluable to decision-support systems. The specificity of the knowledge they contain and their overall collaborative approach to treatment is ideal for cooperative efforts for counteracting the current epidemic’s effects by “being able to supply practitioners with up to the minute information about what are the right things to do, what are the best practices, for managing patients with this virus,”  pointed out  Mark Musen, Director of the Stanford University Center for Biomedical Informatics Research and director of the World Health Organization Collaborating Center for Classification, Terminology, and Standards at Stanford University.

Ontology-Based Decision Support In Action

In clinical settings like New York’s Montefiore Medical Center, the University Hospital for Albert Einstein College of Medicine, BioPortal ontologies are used in decision-support technologies that assist in the treatment of patients who have a range of conditions, from COVID-19 to sepsis and spinal comprehension. In this point-of-care decision-support use case and others, computer programs process enormous amounts of patient data about an individual’s clinical situation, medication, potential genomic factors, and other personal information.

“Montefiore Health System has developed the Patient-centered Analytic Learning Machine (PALM) that combines numerous medical and healthcare ontologies and taxonomies to deliver an enterprise platform for real-time Artificial Intelligence in healthcare,” said Dr. Parsa Mirhaji, Director of Center for Health Data Innovations at the Albert Einstein College of Medicine and Montefiore Health System, NY. “PALM’s ontological and taxonomic core knowledgebase unifies all analytics and data from heterogeneous sources for many applications, including predicting respiratory failure in COVID-19 patients, providing just-in-time recommendations and personalized decision support for clinicians to improve patient outcomes, as well as managing appointments and prescriptions.”)

Simultaneously, these systems access certified knowledge from ontologies about similar cases, previous treatment methods, and other research related to the affliction. The analysis stage is typically two-fold. The Semantic Knowledge Graph technology underpinning these systems gives “their knowledge bases the ability to reason about these concepts most efficiently and robustly possible,” added Musen. Cognitive computing applications also aid the analysis of an individual patient’s situation and knowledge about his condition to determine best practices for treatment.

Scaling Healthcare Knowledge For Real-Time Decision Support

Ontologies are data models for Semantic Knowledge Graphs, which are essential for the scalability necessary to implement the massive datasets for point-of-care clinical support. Decision support systems must couple patient data with this ontological knowledge base to intelligently reason about treatment.

Point-of-care decision support mechanisms enable practitioners to treat patients with the latest information about a quickly evolving situation and other biomedical phenomena. It’s the most effective means of democratizing information about critical care, enabling a community of researchers and practitioners to collaborate for the general public’s greater good.

As Dr. Jans Aasman, CEO of Franz Inc. explains, “When building a Knowledge Graph as your enterprise's single source of truth, it's critical to include ontologies and taxonomies. AI applications and complex reasoning analytics require information from both databases and knowledge bases that contain domain information, taxonomies, and ontologies to solve complex questions. To make this possible, we developed a novel hybrid sharding technology called FedShard, which facilitates the combination of data and knowledge required by applications like Montefiore's PALM. But this approach is not unique or specific to Healthcare, it is applicable in many other industries, which is why we are excited about OntoPortal's plans to bring sharing of domain ontologies to a broad audience.”

Sharing Ontologies Can Fuel Progress In Healthcare And More

The superiority of the ontological approach for point-of-care decision support is manifold. Online ontology repositories provide a central means of sharing knowledge, concepts, and terminology about biomedical, environmental, agricultural, or other domain-specific circumstances. Ontologies offer a launching point for future conceptual applications without having to create conceptual knowledge from scratch and leverage the efforts, energy, and resources of other scientists involved in like-minded pursuits—like eliminating the current epidemic.

About The Author

Dave McComb is the President and cofounder of Semantic Arts. He and his team help organizations uncover the meaning in the data from their information systems. Dave is also the author of "The Data-Centric Revolution," "Software Wasteland" and "Semantics in Business Systems." For 20 years, Semantic Arts has helped firms of all sizes in this endeavor, including Proctor & Gamble, Goldman Sachs, Schneider-Electric, Lexis Nexis, Dun & Bradstreet, and Morgan Stanley. Before Semantic Arts, Dave cofounded Velocity Healthcare, where he developed and patented the first fully model-driven architecture. Before that, he was a part of the problem.