Deep Learning in Medical Forensics and Neurodegeneration: A Survey on Tampered Image Detection and Alzheimer’s Diagnosis
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The widespread use of digital medical imaging modalities such as MRI, CT, and X-ray has increased the risk of malicious manipulation through deepfake techniques, including artificial atrophy insertion, pixel modification, and GAN-generated forgeries.Such tampering can mislead clinicians, compromise diagnostic accuracy, and endanger patient safety, underscoring the need for reliable medical image authentication.This study presents an advanced dualstage framework that integrates deep learning and image forensics to both detect medical image manipulation and provide accurate Alzheimer's disease predictions.In the first stage, tampered and untampered medical scans are analyzed using a hybrid forensic architecture that combines handcrafted feature extraction with deep models such as ResNet and CNNs, enabling precise detection of copymove, splicing, and GAN-based deepfakes across multiple imaging modalities.The second stage employs volumetric learning and transfer-learning approaches, including 3D-CNN and ResNet-18, to classify Alzheimer's disease into its respective stages with high accuracy.Experimental results demonstrate that the proposed system achieves superior performance in both forgery detection and disease prediction, offering a robust, trustworthy, and clinically valuable solution for modern healthcare environments.
Fields of Study & Subjects
Digital Media Forensic Detection
Advanced Steganography and Watermarking Techniques
Benford’s Law and Fraud Detection
Publication details
- Journal / source
- Advances in intelligent systems research/Advances in Intelligent Systems Research
- Pages
- 201–211
- Publication date
- 1 April 2026
- Language
- EN
- Document type
- Book chapter