Eye-XAI: an explainable artificial intelligence approach for eye disease detection using symptom analysis

Abstract

The early and accurate detection of eye diseases play a pivotal role in preventing vision loss and improving patients’ quality of life. It is therefore important to search for methods for improving this detection. It turns out that Artificial Intelligence (AI) has shown great promise for the detection task. However, AI models are often opaque and complex and as a result there has been a slow adoption for these models in clinical settings. In this paper Eye-XAI is introduced which provides an effective approach to the detection combining Explainable Artificial Intelligence (XAI) techniques with symptom analysis to enhance the transparency and interpretability of eye disease detection models. Our results demonstrate that Eye-XAI not only achieves high accuracy (99.11%) for eye disease detection but also provides transparent and interpretable insights into the diagnostic process. The adoption of Eye-XAI can therefore significantly enhance the early detection and management of eye diseases while empowering clinicians with a deeper understanding of its AI-based diagnostic recommendations. Furthermore, this approach promotes patient engagement by facilitating communication and trust between patients and their healthcare providers. Eye-XAI represents a major step towards the integration of XAI in ophthalmology, unlocking new possibilities for improved eye disease diagnosis and treatment.

Ähnliche Arbeiten