The most frequent cause of blindness in Europe is to be searched in both non-detected diabetic retinopathy and glaucoma. Several telemedical initiatives have been launched in different member states to reduce this risk via early detection screening and shared care support. TOSCA's goal is to co-ordinate the existing services into one standardised platform. If patients who are threatened with blindness receive timely and proper treatment, their eye sight can be spared whereas the costs for the community might also be kept at a reasonable level. Scientific and clinical institutions, and industries from 8 European countries are involved in TOSCA, including Germany, Denmark, The United Kingdom, Greece, Italy, Ireland, The Netherlands, and Russia. Six of them contributed to OPHTEL and have built up experience in synchronous as well as asynchronous tele-consultation.
The TOSCA project consists of four distinct but merging tasks. The team first will establish a common telemedical communication infrastructure, based on existing or still evolving standards for data exchange and integration such as XML and DICOM for imaging, and CORBAmed. TrustHealth will be installed to protect the patient's privacy and security. To achieve data integration on a semantic level, the partners will create a special database, using harmonised ophthalmologic terminology. The standard terminology server has to provide descriptive definitions, hierarchical relations among concepts, and language specific terms. The user can access the various tele-ophthalmologic services through a virtual centre which serves as a brokerage portal for telescreening, patient home-monitoring, image processing and the consultation of reference images in knowledge-based information systems.
Once the platform established, it will be applied in first instance for the tele-screening of diabetic retinopathy. The TOSCA team will strive to implement European standards for telescreening procedures and settings by developing advanced methods based on stereo images; to train screeners; and to assure overall quality. As a result, the partners hope to contribute to the aim of the Saint Vincent Declaration, issued by the World Health Organisation (WHO), to reduce diabetic retinopathy-related blindness by one third. The detection of glaucoma will be the second telescreening task for the TOSCA platform. A home-monitoring and telecommunication system will evaluate the course of the patient's glaucoma as well as the communication with the physician. The partners will also increase the awareness and compliance of the patient via the development of a patient-centred glaucoma information system.
Apart from the infrastructure for the platform, the TOSCA project will design special software to automatically assess the quality of fundus images as well as to classify lesions. The image processing activities from the past OPHTEL project and from the institutions in Aarhus, Aalborg, London, and Munich, which are currently partnering in TOSCA, will be integrated. The objective is to deliver software that can be applied to all colour fundus images, no matter what the technical source is. In addition, a comprehensive reference image database will be set up to serve as a source of comparison for real life patient data, to facilitate diagnosis and for research and educational purposes.
Project co-ordinator Dr. Zahlmann expressed the firm belief that the visual data has to be evaluated at the local level before transmitting it at the remote site to prevent patient images lacking sufficient quality from being submitted to telescreening. Also, it should be checked whether or not the image shows a normal fundus because the OPTHEL project has proven that 60 percent of the patient data assessed in telescreening are images with a normal fundus. In those cases, it is quite unnecessary to bother the telescreening service in the first place. Therefore, an algorithm will be developed to sort out patients with no retinopathy.
At the end of the TOSCA project in December 2002, the different imaging components will be implemented into ophthalmologic image capturing and archiving systems. For more details about the use of tele-ophthalmological services for detecting diabetic retinopathy and glaucoma, you can visit the Web sites of the OPTHEL and TOSCA projects.