Explainable AI for Early Diagnosis of Alzheimer’s Disease: A Transparent Deep Learning Approach

Abstract

Alzheimer’s disease is a degenerative cognitive disease that primarily affects the memory and behavior of an individual. Early and accurate prediction of Alzheimer’s disease is essential for timely prevention, which can potentially slow down the progression of the disease. Earlier diagnostic strategies relied on clinical evaluations, but these methods are very time-consuming. In recent years, deep learning has emerged as a powerful method for evaluating medical images. It is capable of detecting early indicators of Alzheimer’s disease. The ongoing advances in deep learning have shown hope in improving the early prediction of Alzheimer’s disease. This research used a neural network for deep learning for the early prediction of Alzheimer’s disease with an accuracy of 84.97±0.84%. Additionally, techniques like Explainable AI such as SHAP (SHapley Additive Explanations), including summary charts and local and global explanations, provide valuable insights, which enable the researchers to identify the critical factors of the disease. The principal objective of our research on forecasting Alzheimer’s disease with Explainable AI is to develop an interpretable model for the early diagnosis of the condition via SHAP, hence enhancing trust in AI-driven prognostics.

Ähnliche Arbeiten