RTU logo

Running Projects

ISTARE: Intelligent Spatio-Temporal Activity Reasoning Engine    
(July 2010 - July 2012)
Supported by Army Research Lab and DARPA grant W911NF-10-2-0062

Exchange scenarioThe goal of ISTARE is to develop video-analysis software able to capture articulated motion in a threelayer hierarchical dynamical graphical model: (i) development of spatial-temporal atoms for basic features such as points, lines, and regions, (ii) semi-supervisedly examine spatial and temporal coherence inherent in appearance, structure and motion of these lowlevel atoms to induce parts at the mid-level over which joint probabilistic and statistical mechanics models are learned; and (iii) generalizations of these (reusable) mid-level parts into full objects and activities at the high-level by means of a realism-based ontology.

    The ontology will describe types of real-world phenomena relevant for video surveillance including aspects of human motion, with more specific parts focused on the “verbs” that have been recognized to be important for surveillance. These aspects will include two layers: one involving how humans interact with objects and other humans in various scenarios and one how motions of body parts contribute to human body motion (e.g. the specific ways arms and legs move when a person is walking).

ARGOS: Transatlantic Observatory for Meeting Global Health Policy Challenges through ICT-Enabled Solutions    
(Jan 2010 - June 2011)
Supported by the European Union grant SI2.548981

The overall goal of the ARGOS eHealth Pilot Project is to contribute to establishing a “Transatlantic Observatory for Meeting Global Health Policy Challenges through ICT-Enabled Solutions” in order to develop and promote “Common Methods for Responding to Global eHealth Challenges in the EU and the US”.

    Our contribution is to evaluate the role of ontology in semantic interoperability and the modelling and simulation of human physiology and diseases with a focus on the Virtual Physiological Human (VPH) and the use of such solutions to support the diagnosis and treatment of rare diseases.

Related literature:
  • Ceusters W, Smith B. Semantic Interoperability in Healthcare - State of the Art in the US, March 2010. (position paper, slides)

Realism-based versioning for biomedical ontologies    
(Apr 2009 - Apr 2011)
Supported by the National Library of Medicine grant 1R21LM009824-01A1

The vision behind this project is one in which biomedical ontologies do not just reflect the state of the art in biomedical science in terms of what entities exist in reality and of how they are related, but that they also keep track of whether the changes introduced in successive versions of ontologies reflect (1) changes in the underlying reality, (2) in the views of ontology authors - or in associated scientific theories, or (3) are corrections of editorial mistakes. To make this vision come true, especially in the context of ontologies that will work in complex domains such as biomedicine, we must have objective measures for ontology quality.
Our hypothesis is that when this vision is endorsed by the authors of biomedical ontologies and when the ontology authoring environments support the application of this view by means of appropriate software, it would become possible to develop an objective measure for the quality of an ontology as it evolves over time. We test this hypothesis by applying realism-based ontology versioning to the Systematized Nomenclature of Medicine (SNOMED). We do this on the basis of the following specific aims:

  1. analyze SNOMED-CT's existing history mechanism to find out whether the principles of realism-based ontology versioning are able to cope with all requirements put forward by SNOMED-CT. Adjust when needed.
  2. develop a prototype of a realism-based ontology versioning software component that can serve as plug in for ontology authoring systems such as Protégé, ODE or SWOOP.
  3. use the prototype to restructure SNOMED-CT's history information in line with the principles of realism-based ontology versioning.
  4. to compute the quality improvement of SNOMED-CT over time in order to demonstrate the usefulness of the approach and foster its acceptance in other ontologies
Related literature:
  • Ceusters W, Smith B. A Realism-Based Approach to the Evolution of Biomedical Ontologies. Proceedings of AMIA 2006, Washington DC, 2006;:121-125. (draft paper, slides)
  • Ceusters W, Spackman KA, Smith B. Would SNOMED CT benefit from Realism-Based Ontology Evolution? In Teich JM, Suermondt J, Hripcsak C. (eds.), American Medical Informatics Association 2007 Annual Symposium Proceedings, Biomedical and Health Informatics: From Foundations to Applications to Policy, Chicago IL, 2007;:105-109. (abstract, draft)
  • Ceusters W. Applying Evolutionary Terminology Auditing to the Gene Ontology. Journal of Biomedical Informatics 2009;42:518–529. (Official version, accepted draft, reviewers comments and responses)
  • Ceusters W, Smith B. A Unified Framework for Biomedical Terminologies and Ontologies. Proceedings of the 13th World Congress on Medical and Health Informatics (Medinfo 2010), Cape Town, South Africa, 12-15 September 2010 (forthcoming). (Final draft)
  • Ceusters W. Applying Evolutionary Terminology Auditing to SNOMED CT. In American Medical Informatics Association 2010 Annual Symposium (AMIA 2010) Proceedings, Washington DC, November 13-17, 2010. (forthcoming) (paper, response to reviewers)

Referent Tracking enabled web sites    
(Sept 2008 - )

Did you ever think about the fact that when you are viewing a web page, you are not looking at a concrete file which is on some medium on the remote web server, but to an exact copy of that file which has been sent to your computer? In many cases, there is not even a concrete file on the server, but what you see through your browser has been assembled out of several components using, for instance, php. Logically, when you view the page at different times, assuming the browser refreshed its contents, you get each time a different copy, and so do other viewers. Moreover, in between two access events, the page on the remote server might have been changed.
RT-enabled websites keep track of these copies, the distribution of it, and the evolution of the web pages that they contain over time. Never saw one? Well, you are looking at one. On Oct 2, 2008, the RTU RT-enabled web site went life. Each distributed page, at the time of a request, is branded with an IUI which is of course different of the IUI of the remote file. The IUI is printed on the right side of the menu bar.
In addition, a checksum is generated over the IUI and the content of the page which protects authors against claims that certain information was there, when it actually was not. At the same, it is a guarantee for viewers that the authors cannot deny that specific information was there. The checksum is shown on the bottom of the page.
In the background, an RT-database keeps track of all the IUIs, including revisions and new versions of pages, as well as of when and where the copies were sent to.
The RT-enabled website technology is available for licensing.

Exploring eyeGENE from a Bioinformatics Perspective    
(June 2008 - )

The eyeGENE database, launched in 2006, collects genotype and phenotype information for patients with eye diseases as a tool to further eye research. To make the system maximally useful from a bio-informatics perspective, the RTU has been requested to perform the following activities:

  1. Prepare an updated review of existing and emerging medical information standards of note in the U.S. as applicable to clinical research data and all major standards committees, with clarification as to how the various standards and standards committees overlap and which ones are likely to have or gain broad acceptance.
  2. Review the unique features of eyeGENE and identify opportunities for future enhancements of value both for eyeGENE and for the medical informatics community.
  3. Offer recommendations on future steps for synchronizing eyeGENE data with applicable medical information standards.
  4. Present innovative ideas for extending the capabilities of eyeGENE and the richness and availability of anonymized data to the research community.

Ontology for Risks Against Patient Safety (RAPS)    
(April 2008 - November 2009)

We will develop in collaboration with RAMIT vzw a component that contains in a machine understandable way all the domain knowledge that is required to be able to prevent, predict, detect or deal appropriately with RAPS in the context of the disease history of a patient. This component will consist of two parts. One part is an ontology describing the portions of reality salient to the domain of RAPS occurring in hospitals. This ontology will itself consist of two different parts. One part is the RAPS domain ontology which will be developed following the principles of unqualified realism and therefore will be lined up with Basic Formal Ontology. This ontology will thus consist exclusively of

  • representational units that refer to salient universals in the domain covered, examples being person, drug, allergic reaction, and
  • relationships taken from the OBO Relation Ontology that has been developed under the same realist assumptions.
The second part of the ontology is the RAPS application ontology that will use the domain ontology as a reference, but will differ from it in a few aspects:
  • it will contain defined classes to represent characteristics of groups of particulars that do not correspond with universals,
  • it will eliminate detail which is irrelevant for the purposes of the application.
The second part of the component is a taxonomy of terms in various languages that are commonly used as linguistic denotations for the universals, particulars, relationships and the more complex portions of reality of which the former three are constituents. Thus, whereas the ontology part describes (some aspects of) what is the case in reality, the terminology part describes (some aspects of) how humans communicate by means of language about reality.

Downloads:
  1. Ceusters W, Capolupo M, Devlies J. D4.2 – RAPS Domain Ontology (M12 Version). Background materials and methodology used to develop the Domain Ontology for Risks against Patient Safety, January 11, 2009, 55p.
  2. Ceusters W, Capolupo M, Devlies J. D4.3 – RAPS Application ontology (Version 1). Background materials and methodology used to develop Application Ontologies for Risks against Patient Safety, January 11, 2009, 53p.

Browse the ontology: LOGIN or REGISTER (select "Rem Ontology" as resource).

UB Task Force for ontology-based IT support for large scale field studies in Psychiatry    
(Sept 2007 - March 2009) (press release, UB Reporter)

The John R. Oishei Foundation's mission is to enhance the quality of life for Buffalo area residents by supporting education, healthcare, scientific research and the cultural, social, civic and other charitable needs of the community. The Foundation was established in 1940 by John R. Oishei, founder of Trico Products Corporation.
In August 2007, the John R. Oishei Foundation authorized a grant to establish at SUNY at Buffalo (UB) a Taskforce for Ontology-Based IT Support for Large-Scale Studies in Psychiatry. The goal of the taskforce is to create the foundation for making UB the best positioned information technology partner to participate in large, multi-center data collection and analysis efforts that are underway to solve a number of diagnostic problems in the domain of psychiatry. The taskforce, led by Werner Ceusters, MD (PI), will draw on the expertise available in the UB Departments of Psychiatry, Philosophy (Barry Smith, PhD), and Computer Science and Engineering (Stuart C. Shapiro, PhD), including the Center for Cognitive Science. The specific aims addressed by the taskforce are:

  1. to assess the functional and technical requirements to be fulfilled by a data management system able to do justice to both the dimensional and categorical approach in psychiatric diagnosis;
  2. to design an implementation and funding plan for the technical infrastructure to be built in order to support data collection and analyses in large-scale field studies in psychiatry, and;
  3. to initiate the collaborations needed to deliver data collection and analyses services to provide the answers to the questions raised in the DSM-V research agenda.

Making existing EHR systems RT compatible    (Dec 2006 - )

By advocating the use of instance unique identifiers to refer to the entities comprising the subject matter of patient health records, Referent Tracking promises many benefits to those who use health record data to improve patient care. One challenge ahead lies in furthering the adoption of the paradigm by developers of existing EHR applications. To meet this challenge, we have begun the process of integrating RT into commercial EHR applications A part of this process is an analysis of the extent to which the data collected by an EHR application needs to be reformulated to make it compatible with the requirements of RT, namely that the particulars assigned an IUI (for �instance unique identifier�) are instances of the kinds included in Basic Formal Ontology (BFO).

  • An initial exploratory analysis was carried out for Praxis which allowed us to identify the obstacles and the directions that future research in relation to this system should take. (Dec 2006 - July 2007)
  • A more in depth analysis was performed on MedtuityEMR which led to the formulation of a research proposal. (Dec 2006 - )
  • For Sigmund Software LLC, we are exploring how their Target Behavior Tracking system which is currently used within individual organizations can benefit from referent tracking to develop a cross-institutional knowledgebase linked to other behavioral assessment tools. (March 2007 - )



Past Projects

Oral Diagnostic Consultation Tracking    (Oct 2005 - March 2007)

The UB School of Dental Medicine started a project involving tracking patients in their dental school clinics for whom a request has been made for a consultation by an oral medicine specialist from the department of oral diagnostic sciences. The school, who is responsible for all patients in its clinics, wants to ensure that these patients are, in fact, seen by an oral medicine consultant. It wants to know why the consultation was requested, what the findings were, what recommendations were made by the consultant and if and how the recommendations were carried through. A form was constructed to be filled out by the consultants when they are called to see a patient. An ontology has been created, referring to the universals of which the entities that are referred to by the data captured by means of this form are instances.

 

701 Ellicott Street ~ Buffalo, New York 14203 ~ 716.898.8630privacy policy | contact us