April 2010
Volume 51, Issue 13
ARVO Annual Meeting Abstract  |   April 2010
An Annotation Software to Help in Early Detection of Amd
Author Affiliations & Notes
  • S. Carvajal-Gonzalez
    Research Statistics, Pfizer Inc, Groton, Connecticut
  • C. S. Tressler
    Global Medical, Pfizer Inc, New York, New York
  • J. G. Cunha-Vaz
    AIBILI, Coimbra, Portugal
  • G. Staurenghi
    Dept of Clinical Science (Luigi Sacco), University of Milan, Milano, Italy
  • U. Chakravarthy
    Ophthalmology, Queens University of Belfast, Belfast, United Kingdom
  • B. Lay
    ADCIS SA, Herouville Saint-Clair, France
  • R. Danno
    ADCIS SA, Herouville Saint-Clair, France
  • J.-C. Klein
    CEntre de Morphologie Mathematique, Ecole des Mines de Paris, Fontainebleau, France
  • Footnotes
    Commercial Relationships  S. Carvajal-Gonzalez, Pfizer Inc, E; C.S. Tressler, Pfizer Inc, E; J.G. Cunha-Vaz, Pfizer Inc., F; G. Staurenghi, Pfizer Inc., F; U. Chakravarthy, Pfizer Inc., F; B. Lay, Pfizer Inc., F; R. Danno, Pfizer Inc., F; J.-C. Klein, Pfizer Inc., F.
  • Footnotes
    Support  None.
Investigative Ophthalmology & Visual Science April 2010, Vol.51, 1794. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      S. Carvajal-Gonzalez, C. S. Tressler, J. G. Cunha-Vaz, G. Staurenghi, U. Chakravarthy, B. Lay, R. Danno, J.-C. Klein; An Annotation Software to Help in Early Detection of Amd. Invest. Ophthalmol. Vis. Sci. 2010;51(13):1794.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Purpose: : ADCIS, in partnership with experts in Ophthalmology and Pfizer Inc., developed a software using imaging tools to track the evolution of pathologies during the Age-Related Macula Degeneration conversion.

Methods: : 100 patients with high risk of conversion were enrolled in a clinical trial at 3 different sites and are examined every 6 months over a period of 2 years. At each visit, color fundus, infrared, red free and autofluorescence images, 3D OCT, fluorescein and indocyanin green angiographies are acquired with the HRA II Spectralis. Then, these images are exported from the Heidelberg Eye Explorer© database and uploaded to a remote database accessible from all sites.Then, the software is used:1.In "Grader" mode to annotate selected images by filling modality-specific forms, outlining pathologies, adding comments, and diagnosing AMD conversion.2.In "Investigator" mode by experts of each site who review the grader annotations in order to agree or disagree, alter or add annotations. Once a visit review is complete, they can view annotations made by other investigators, and correct/update their own if necessary.3.In "Consensus" mode to adjudicate the annotations: When discrepancies appear between experts and/or grader, a "consensus" meeting is called to review and alter any previously made annotation in order to seal the final diagnosis.The annotated database is then used by image processing experts to develop algorithms to automatically detect, classify and quantify the pathologies in order to be used in statistical analyses.

Results: : •Intra- and inter-modality image registration techniques were developed.•On FAF images, algorithms were developed:oFor the automatic detection of anatomical structures (i.e.vascular tree, optic disc and fovea) and the detection of zones of depigmentation and of geographic atrophy,oFor the tracking of pathology evolutions between visits after image registration, standardization and normalization of the images.

Conclusions: : Good results were obtained on FAF images for some pathologies. Future developments aim to detect and track more pathologies on all modalities.

Clinical Trial: : www.clinicaltrials.gov NCT00902785

Keywords: age-related macular degeneration • imaging/image analysis: clinical • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) 

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.