particular, remains elusive. Recent advancements in deep learning and imaging technologies have shown promise in aiding the early and accurate diagnosis of AD, which is crucial for effective treatment and management. This article discusses the use of multi-modal deep learning models in automated medical image analysis for AD diagnosis. The models leverage both 2D and 3D MRI images and amyloid PET scans to achieve state-of-the-art performance on the OASIS-3 cohort. The integration of multiple imaging modalities has shown to significantly enhance model performance, highlighting the potential of these models in aiding our understanding of the disease’s causes.
