Presentation is loading. Please wait.

Presentation is loading. Please wait.

International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA SysML and Ontology in Biomedical Modeling Henson Graves Yvonne Bijan 30 January 2011.

Similar presentations


Presentation on theme: "International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA SysML and Ontology in Biomedical Modeling Henson Graves Yvonne Bijan 30 January 2011."— Presentation transcript:

1 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA SysML and Ontology in Biomedical Modeling Henson Graves Yvonne Bijan 30 January 2011

2 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA Outline Our interest in ontology biomedical modeling Our initial focus Modeling objectives by ontology community Ontology modeling achievements How their OWL modeling works Some comparisons between SysML and OWL

3 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA Our Interest In Ontology Modeling Do the modeling principles used by Description Logic (OWL) community offer anything for MBSE? Will these examples and the OWL models help us understand how to integrate formal reasoning with SysML? How do biomedical examples look in SysML? Do the modeling principles used for air vehicles and other systems work in biomedical domain?

4 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA Our Initial Focus Is On Structural Modeling For Anatomy & Chemistry This kind of modeling is being actively investigated using conceptual modeling languages such as OWL

5 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA Ontology Modeling Objectives Construct a model that captures what is common to all (or at least most) human hearts –corresponds to product model, or product line Perform general reasoning about effects of pathology and disease symptom propagation –general properties of operation Use general case to analyze and reason about a specific heart –fault detection

6 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA The Conceptual Modeling Results Large Taxonomies of medical, biological, anatomy terms –SNOMED –reference ontology for healthcare –GALEN – reference ontology for medical terminology Identification of primitive relations for biomedical modeling, e.g., –Instance, subclass, part, Representation as classes and properties in OWL and other Description Logic languages Limited amounts of reasoning

7 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA A General Principle From Conceptual Modeling That Applies To Engineering ModelInterpretation Clear distinction between model and thing being modeled

8 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA hasLayer Heart LeftSideRightSide AorticValve MitralValvePulmonicValveTricuspidValve LeftVentricleSeptumRightVentricle HasComponent hasConnection divisionOF Biomedical Often Starts With Structural Diagram (e.g., PowerPoint Engineering) … and Constructs Axioms (aka Knowledge Base, Ontology)

9 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA Diagrams Are Translated To Axioms Heart subclass (hasComp 1 AorticVentricle) This says that a necessary condition for a heart is that it have one Aortic Ventricle Heart AorticVentricle hasComp

10 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA Axioms Do Not Fully Recognize The Distinction Between Parts and Other Connections For example the arrow labeled hasLayer in the heart structure diagram is translated as LeftVentricle subclass (hasLayer some Septum) This is incorrect outside of the context of the heart. SysML would lead one to represent this as hasPart1.hasLayer1 = hasPart2.hasLayer2 Which says that the left ventricle part of the heat is connected by hasLayer1 to the same thing as the right ventricle part is connected to Heart LeftVentricle RightVentricle haspart1 Septum Haspart1: LeftVentricle Haspart2: RigntVentricle Haslayer1

11 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA Our SysML BDD and IBD for the Heart

12 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA There is a lot to be learned from OWL modeling, - but they don’t always get it right Good modeling principles are critical for any approach –Sort out and define what are the necessary conditions for the model –Sort out what is a part and what is a connection, both parts and connections can be necessary conditions

13 International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA Future Work We are looking at parametric constraints and state diagrams for the heart. Our objective is to have a model that allows us to generate a 3D visualization of the heart with behavior that responds to its neural stimulation


Download ppt "International Workshop 28 Jan – 2 Feb 2011 Phoenix, AZ, USA SysML and Ontology in Biomedical Modeling Henson Graves Yvonne Bijan 30 January 2011."

Similar presentations


Ads by Google