|
|
BRIEF REPORT |
|
Year : 2006 | Volume
: 54
| Issue : 2 | Page : 126-129 |
|
Evaluation of web-based annotation of ophthalmic images for multicentric clinical trials
KV Chalam, P Jain, VA Shah, Gaurav Y Shah
University of Florida College of Medicine, Jacksonville, Florida, Indent, Inc, Jacksonville, Florida, USA
Correspondence Address: Gaurav Y Shah 580 West 8th Street, Jacksonville, Florida - 32209 USA
Source of Support: None, Conflict of Interest: None | Check |
DOI: 10.4103/0301-4738.25838
An Internet browser-based annotation system can be used to identify and describe features in digitalized retinal images, in multicentric clinical trials, in real time. In this web-based annotation system, the user employs a mouse to draw and create annotations on a transparent layer, that encapsulates the observations and interpretations of a specific image. Multiple annotation layers may be overlaid on a single image. These layers may correspond to annotations by different users on the same image or annotations of a temporal sequence of images of a disease process, over a period of time. In addition, geometrical properties of annotated figures may be computed and measured. The annotations are stored in a central repository database on a server, which can be retrieved by multiple users in real time. This system facilitates objective evaluation of digital images and comparison of double-blind readings of digital photographs, with an identifiable audit trail. Annotation of ophthalmic images allowed clinically feasible and useful interpretation to track properties of an area of fundus pathology. This provided an objective method to monitor properties of pathologies over time, an essential component of multicentric clinical trials. The annotation system also allowed users to view stereoscopic images that are stereo pairs. This web-based annotation system is useful and valuable in monitoring patient care, in multicentric clinical trials, telemedicine, teaching and routine clinical settings. Keywords: Web, annotation, images.
How to cite this article: Chalam K V, Jain P, Shah V A, Shah GY. Evaluation of web-based annotation of ophthalmic images for multicentric clinical trials. Indian J Ophthalmol 2006;54:126-9 |
Multicentric clinical trials[1],[2],[3],[4],[5] have utilized film based photography to monitor the progress of disease during the course of the study. Typically standardized fundus photographs acquired from study subjects at participating centers, are transported to photography reading centers for grading and evaluation. This process is labor intensive, expensive, inefficient and does not take advantage of current advances in web based digital technology.
Commercially available digital fundus photography (Zeiss Visupac, Topcon Imagenet) and ophthalmic patient management systems (Medinote Corporation, Ophthalmic Image System) are not web-based and have limited annotation capability. These software systems for viewing and annotating fundus images are on proprietary client-server architecture (not accessible through the web), with limited capabilities for saving, sharing and comparing annotations. We describe a prototype Internet browser-based annotation system for identifying and describing features in digitalized retinal images.
System description
In the web-based annotation system, the bottom layer contains the image to be annotated [Figure - 1]. A variety of images (color fundus photograph, red-free photograph, fundus fluorescein angiograph and indocyanine green angiograph), can be included in this group. A transparent annotation layer is placed on top of the image [Figure - 2]. An annotation layer may contain multiple annotations that encapsulate the observations and interpretations of a user for a specific image.[6] Multiple annotation layers may be overlaid on a single image [Figure - 3]. These layers may correspond to annotations by different users on the same image or annotations of a temporal sequence of images of disease process over a period of time (such as diabetic macular edema).
Users employ a mouse or pen-based device to choose an annotation tool from a customizable palette. These tools [Figure - 2] and [Figure - 3] are used to mark or draw on a transparent layer, thereby identifying pathologies in the image. The annotated pathology is subsequently described by entering data in a customizable form, as illustrated in [Figure - 2] and [Figure - 3].
Geometrical properties of annotated figures may be computed and measured. Two markers are provided, one to identify the macula and the other to identify the optic disk [Figure - 4]. The markers are concentric circles with radial lines. The distance between the center of the macula and optic disc margin is assumed to be 4 mm.[7] After the user places the two markers, the distance is calibrated by dividing 4 mm by distance in the number of pixels. This yields a uniform scaling factor in mm per pixel, which is used to transform pixel (image) measurements into actual distance and area. Recalibration is required when the camera specifications are modified.
Multiple annotated layers with different sets of data (e.g. from multiple visits) can be overlaid and compared. The registration process involves placing of four markers at blood vessel bifurcation points, to align the images.
The annotations are stored in a central repository database. They can be retrieved, based on a variety of criteria including patient identifier, properties of annotations and changes to properties over time [Figure - 5].
The annotation system also allows users to view stereoscopic images that are stereo pairs. The image annotation system is built using Scalable Vector Graphics. It is an extended markup language (XML) based standard, that allows annotations to be stored as vectors with associated data.
Discussion | | |
The currently available digital ophthalmic systems have the following drawbacks: The annotation files created by the systems have a large file size. Geometrical properties like area and greatest linear dimension of the lesion can be computed for a single annotation, but not for custom computations. Annotations cannot be overlaid and sequential changes in a disease process over a period of time that is documented through digital photography, cannot be compared. Multicentric collaboration is not easily supported, because specialized software is required to access these systems from other computers, inside or outside the network.
Our system allows overlay of multiple annotation layers, created for a sequence of images from multiple patient visits and facilitates study of temporal properties of pathologies in the images. This provides an objective method to monitor properties of pathologies over time, an essential component of multicentric clinical trials. This system allows multiple users in remote locations to use a web browser, to annotate the same image in a "double-blind" manner and compare multiple images of a patient over a time period.
This is a prototype software and not currently available for commercial use.
There are other systems that attempt to identify features in images through automated image processing algorithms, based on pattern matching.[8],[9],[10] Our annotation system instead, provides rich annotation tools to users to manually draw on a computer screen and create digital annotations.
Annotation of ophthalmic images is clinically feasible and useful to interpret, like track properties of an area of fundus pathology (such as area of retinal edema, capillary nonperfusion, drusen, size of neovascular membrane and capillary leakage) between multiple sites over multiple patient visits. Various properties of the lesion that can be tracked, include geometry of annotation, area, greatest linear distance and grading of pathology. This system is suitable in multi-centric clinical trials for double blind grading of pathology. Web application reduces the time to collect and read images, reduces cost of managing images and associated data, enhances objectivity of reading images and increases efficiency of collection and analysis of data[6] [Figure - 6], [Figure - 7].
In summary, our web-based annotation system is useful and valuable in monitoring patient care in multicentric clinical trials, telemedicine and routine clinical settings.
References | | |
1. | Design and methods of a clinical trial for a rare conditions: The Collaborative Ocular Melanoma Study. COMS report No. 3. Control Clin Trials 1993;14:362-91. |
2. | Complications of Age- Related Macular Degeneration Prevention Trial Study Group. The Complications of Age- Related Macular Degeneration Prevention Trial (CAPT): rationale, design and methodology. Clin Trials 2004;1:91-107. |
3. | Cleary PA, Beck RW anderson MM Jr, Kenny DJ, Backlund JY, Gilbert PR. Design, methods and conduct of the Optic Neuritis Treatment Trial. Control Clin Trials 1993;14:123-42. |
4. | Age- Related Eye Disease Study Research Group. Risk factors associated with age- related macular degeneration. A case - control study in the age-related eye disease study: Age- Related Eye Disease Study Report Number 3. Ophthalmology 2000;107:2224-32. |
5. | Bressler NM, Bressler SB, Hawkins BS, Marsh MJ, Sternberg P Jr, Thomas MA. Submacular Surgery Trials Pilot Study Investigatiors. Submacular surgery trials randomized pilot trial of laser photocoagulation versus surgery for recurrent choroidal neovascularization secondary to age-related macular degeneration: I. Ophthalmic outcomes submacular surgery trials pilot study report number 1. Am J Ophthalmol 2000;130:387-407. |
6. | Jain P, Komatineni S. Data Driven Web Graphics with SVG- A Declarative Approach. XML Jr 2002;3:32-41. |
7. | Grand MG, Bressier NM, Brown GC, Flynn HW Jr. Marmor MF. Basic Anatomy. Basic and Clinical Science Course. Section 12. Am Acad Ophthalmol: San Francisco; 1996. p. 11. |
8. | Goldbaum MH, Katz NP, Nelson MR, Haff LR. The discrimination of similarly colored objects in computer images of the ocular fundus. Invest Ophthalmol Vis Sci 1990;31:617-23. [ PUBMED] |
9. | Shin DS, Javornik NB, Berger JW. Computer-assisted, interactive fundus image processing for macular drusen quantitation. Ophthalmology 1999;106:1119-25. [ PUBMED] [ FULLTEXT] |
10. | Berger JW, Shin DS. Computer-vision-enabled augmented reality fundus biomicroscopy. Ophthalmology 1999;106:1935-41. [ PUBMED] [ FULLTEXT] |
[Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4], [Figure - 5], [Figure - 6], [Figure - 7]
This article has been cited by | 1 |
A system for the acquisition, interactive exploration and annotation of stereoscopic images |
|
| Benzeroual, K., Haouach, M., Guinot, C., Venturini, G. | | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2009; 5651(156): 160 | | [Pubmed] | | 2 |
A 3D hypermedia with biomedical stereoscopic images: From creation to exploration in virtual reality |
|
| Haouach, M., Venturini, G., Guinot, C. | | Proceedings of the 20th ACM Conference on Hypertext and Hypermedia, HTę09. 2009; (337): 338 | | [Pubmed] | | 3 |
Acquisition, annotation and interactive exploration of stereo images with virtual reality |
|
| Haouach, M., Benzeroual, K., Guinot, C., Venturini, G. | | HEALTHINF 2009 - Proceedings of the 2nd International Conference on Health Informatics. 2008; : 369-374 | | [Pubmed] | |
|
|
|
|