Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. Vergadering Doctoraat Sven Van Poucke 09/12/2008
  2. 2. Chronic Wound Care Wound bed preparation - a concept aimed at assisting clinicians in wound bed assessment and the development of strategies to maximise healing potential - is now recognised as an important aspect of care. The Red-Yellow-Black-scheme is commonly used for classifying chronic wounds. Simplified direction for selecting appropriate dressings for open wounds by focusing on the phase of healing as evidenced by the predominant condition of the wound base The associated development of the TIME framework (Tissue management; Inflammation and infection control; Moisture balance; Epithelial (edge) advancement) offers clinicians a practical tool for translating wound bed preparation into practice.
  3. 3. Chronic Wound Care – Semantic Link Growing demand for randomized clinical trials health technology assessments an increasing economical pressure on health budgets Key requirement for optimal data sharing is standardization with agreements on types and definitions of structures, processes and formats used to access and share this data plus the implementation of consensus based, standardized terminologies and coding schemes.
  4. 4. Chronic Wound Care – Semantic Link Wound measurement and assessment is important for several reasons including: Tracking patient progress to ensure that healing of a specific wound is progressing with the selected treatment regime  Allowing institution managers to audit patient progress and institution effectiveness Building a reliable and consistent database for outcomes based studies of wound care Improving patient compliance Consistently documenting patient care for reimbursement purposes and for protection of the health care provider against litigation.
  5. 5. <ul><li>Research Questions </li></ul><ul><ul><li>Which colorimetric and geometric parameters are most relevant to analyse tissue repair in a time frame, (macroscopic and non-invasive)? </li></ul></ul><ul><ul><li>Which ultrastructural elements and techniques could add significant value in this analysis? </li></ul></ul><ul><ul><li>Which biomarker(s) might demonstrate any prognostic value in tissue repair? </li></ul></ul>
  6. 6. Chronic Wound Care – Semantic Link 1. Decisions are based on visual perception. 2. Descriptive analysis is poorly standardized and rarely reproducible. 3. The temporal changes using digital images can now be calibrated. The color content of the images becomes independently of any camera settings and illumination features. 4. Consistent, controlled and universally accepted vocabularies: for documenting, describing and comparing wounds. 5. Ontologies (controlled vocabularies) promise to help address many of the difficulties currently experienced in managing large image databases.
  7. 7. COP a semi-open, international, virtual community of practice devoted to advancing the field of research in non-invasive wound assessment by image analysis , ontology and semantic interpretation and knowledge extraction (content–based visual information retrieval) .
  8. 8. Partners
  9. 9. Groups - Wound Image Base Working Group - Wound Image Visual Diagnostic Expert Group - Wound Image Colour, Texture and Shape Group - Wound Image Ontology Working Group
  10. 10. Colour-space
  11. 11. Colour-space
  12. 12. WIBG “ deployment of a searchable repository , up- and downloading features of in-vivo digital wound images with calibration chart, automatic and manual calibration of images”
  13. 13.
  14. 14. WIBG Wound image : as a 3-dimensional (2 dimensions in space, 1 in time) representational artifact, of a wound in the real world. Digital images of human wounds plus a reference chart (the Mac- Beth ColorChecker Chart Mini [MBCCC] or the QP Card 201 are uploaded via a smart client tool. The illumination of the reference chart should be homogeneous over the field of view
  15. 15. WIBG-WIVDEG wound? chronic wound wound bed? wound edge? wound border? granulation tissue? fibrin? necrosis?.....
  16. 16. WIVDEG Roi-annotation-...
  17. 17. WIVDEG Sprague Dawley ratten
  18. 18. WICTSG Multiple algorithms to analyse colors, patterns, textures and shapes of all sorts of objects Disambiguation faced during interpretation of and communication on images. Problems related to the photographical analysis of wounds can be partially solved with the current insights used in geo-spacial sciences and satellite imaging . A standardized protocol for extracting reliable color, texture and shape features from ROIs.
  19. 19. WIOWG
  20. 20. WIOWG CODS Server : A joint BMIR/CIM3 project to develop the infrastructure to host an open Collaborative Ontology Development Service and Ontology Repository for the ontology community at large. The infrastructure is based on the Multiuser Protégé Server hosted on a Tier-1 facility provided by CIM3.NET. 1. CODS Server access details at: 2. directions on how to connect to the server are detailed at:
  21. 21. WIOWG subversion (SVN) repository : Simple: Advanced:
  22. 22. WIOWG
  23. 24. License Creative Commons is a nonprofit organization that works to increase the amount of creativity (cultural, educational, and scientific content) in “the commons” — the body of work that is available to the public for free and legal sharing, use, repurposing, and remixing. Attribution. You let people copy, distribute, display, perform, and remix your copyrighted work, as long as they give you credit the way you request. All CC licenses contain this property. Non-Commercial. You let people copy, distribute, display, perform, and remix your work for non-commercial purposes only. If they want to use your work for commercial purposes, they must contact you for permission. Share Alike. You let people create remixes and derivative works based on your creative work, as long as they only distribute them under the same Creative Commons license that your original work was published under.
  24. 25. Reliability and validity of a computer assisted application to measure wound surface area. Van Poucke, S. et al (2008) Validity is crucial and is an indication that the application is measuring what it is designed to measure. CP, FH, COL Repeatability within each wound measurement method was investigated by calculating a coefficient of variation (CV) for each wound measurement.
  25. 26. THE RED-YELLOW-BLACK SYSTEM: A COLORIMETRIC ANALYSIS OF CONVEX HULLS IN THE SRGB COLOR SPACE =Authors= [[Sven Van Poucke]], [[Raf De Jongh]], [[Roald Nelissen]], [[Yves Vander Haeghen]], [[Philippe Jorens]]. =Aim= The red-yellow-black wound classification system continues to be a cornerstone in clinical practice, although a formal colorimetric description is lacking. The aim of this study is to quantitatively assess the three color distribution of the wound bed in a standard, fixed (device-independent) RGB color space: sRGB. =Methods= 20 Digital images of human wounds plus a reference chart (GretagMac- Beth AG, Regensdorf, Switzerland) are calibrated as previously decribed by Vander Haeghen et al. The illumination of the wound plus its reference chart was homogeneous over the field of view. 2846 pixels from the wound bed area were functionally classified as red-yellow-black or other. All pixels were measured in the sRGB color space and convex polyhedra (convex hull) were calculated for each class and for the complete set of studied pixels. A geometric set is convex if for every two points in the set, the line segment joining them is also in the set. The smallest convex shape, the convex hull, encloses the set of pixels. =Results= Respectively 1125 (39.5%), 949 (33.3%), 331 (11.6%) and 441 (15.5%) of the pixels were functionally classified as red, yellow, black and other. The relative convex hull volume distribution of the classified pixels in the sRGB color space was 11%, 10%, 36% and 64% respectively for the red, yellow, black and other classes. The volume sum of the convex hulls of the classes ended 21% higher than the volume of the convex hull of the complete set. =Conclusion= Although the absolute numbers in this study might depend on the wound types under consideration and the inter- and or intraobserver variability, a significant part (64%) of the pixels are not classified as red, yellow or black but as &quot;other&quot;. Most importantly, however 21% overlap exists between the pixel classes making automatic colorimetric classification almost impossible. Other wound semantics seems necessary to reduce the gap between the functional (clinical) and colorimetric use of colour names.
  26. 27. CONVEX HULL OVERLAP (CIELab) &quot;
  27. 28. THE WOUND IMAGE BASE &quot; Dr. Maria Flour UZ Gasthuisberg, Department of Dermatology Prof. Hilde Beele · UZ Gent, Department of Dermatology Prof. Diane Roseeuw  UZ Brussel, Department of Dermatology Prof. Michel de la Brassinne CHU de Liège, service de dermatologie ......