Visual Disease Diagnosis using user-based Judgements

According to the U.S. National Eye Institute the cases of cataract and glaucoma has increased in the past 15 years and is constantly rising. The risk increases constantly, depending on growing age and ethnicity. Moreover, the male population is more affected than females as illustrated below.


While current technology supports surgical interventions on the eye and advanced medicine decreases the likelyhood of ailments earlier diagnosis of diseases can prevent further symptoms and reduce or eliminate their incubation time. And althought cataract can be healed via operational surgery, there is no cure for vision loss, which is precipitated through glaucoma. We are currently studying the effectiveness of diagnosis for eye-diseases like glaucoma (ICD-10, H40-H42) or cataract (ICD-10, H25-H28) with different types of tasks. With our results we want to enable faster and more precise accuracy when it comes to diagnosis. By providing a data set of visual cognitive challenging jobs, with which we can target the following different symptoms:

  • Amblyopia (weak eyesight)
  • Decreased vision angle
  • Weakness in contrast and sharpness detectopm
  • Vision holes
  • Narrow vision (tunnel vision)
  • Difficulty in dark environments
  • Sensitivity to light and glare



The crowdsourcing project is offering a great opportunity to collect judgement data with high-potential-contributors, for providing an overview opinion on tasks, which were created in cooperation with the institutes of psychology and health informatics. With further investigation and attempts to provide a good use-case-scenario for this study, the research can enhance current diagnosis methods for eye-related diseases and symptoms, more specifically Cataract and Glaucoma.



Prof. Dr. Wolf-Tilo Balke, M. Sc. Kinda El Maarry, Duc Hai Le