Authors:
Bharath Kumar Nagaraj

Addresses:
1Department of Artificial Intelligence, Digipulse Technologies Inc., Salt Lake City, United States of America. bharathkumarnlp@gmail.com1

Abstract:

This paper explores the theoretical foundations and practical applications of Explainable Artificial Intelligence (XAI) in the context of epilepsy diagnosis. As the adoption of machine learning algorithms in healthcare continues to grow, there is an increasing need for models that provide accurate predictions and offer clinicians transparent and interpretable insights. The theoretical framework of XAI is discussed, emphasizing the importance of model interpretability in the medical domain. The authors review existing literature on AI applications in epilepsy diagnosis and highlight the limitations of traditional black-box models in providing understandable reasoning for their predictions. The paper then introduces various XAI techniques, such as LIME and SHAP, and their application in enhancing the interpretability of epilepsy diagnosis models. Furthermore, the paper presents a case study or empirical results demonstrating the effectiveness of XAI techniques in a real-world epilepsy diagnosis scenario. The study may include the evaluation of model performance, interpretability metrics, and feedback from medical professionals. In conclusion, the paper underscores the significance of integrating XAI into epilepsy diagnosis systems to bridge the gap between AI predictions and clinical decision-making. The insights gained from interpretable models improve the trustworthiness of AI-based diagnostics and empower healthcare professionals with valuable information for personalized patient care. The implications of this research extend beyond epilepsy diagnosis, serving as a foundation for the broader integration of XAI in medical applications.

Keywords: Theoretical Framework; Applications of Explainable; Epilepsy Diagnosis; Explainable Artificial Intelligence (XAI); Medical Applications; Local Interpretable Model-agnostic Explanations (LIME); SHapley Additive exPlanations (SHAP); Convolutional Neural Network.

Received on: 03/03/2023, Revised on: 12/06/2023, Accepted on: 17/08/2023, Published on: 19/12/2023

FMDB Transactions on Sustainable Computing Systems, 2023 Vol. 1 No. 3, Pages: 157-170

• Views : 147
• Downloads : 10

Download PDF