Medical Deepfakes in Healthcare — Detection, Generation, Human Perception, and Clinical Implications: Scoping Review with ☸️SAIMSARA

Name: SAIMSARA Evidence Object digital::MEDICAL_DEEPFAKES_SS
Creator: SAIMSARA
License: https://saimsara.com/license/

SAIMSARA

doi:10.62487/saimsara570eea90

SAIMSARA Journal

Machine-Readable Science • ISSN 3054-3991

Medical Deepfakes in Healthcare — Detection, Generation, Human Perception, and Clinical Implications: Scoping Review with ☸️SAIMSARA

Digital Health & Biotech

Issue 3, Volume 1, 2026

Chat with this issue

DOI: 10.62487/saimsara570eea90

Editorial note

• Last update: 2026-06-17 22:08:27

What is this paper about

Medical deepfakes can already fool radiologists and advanced AI systems, while detection models often report near-perfect accuracy only in controlled datasets. This review reveals where the real risks lie, how synthetic medical data may also benefit healthcare, and what must change before clinical systems can be trusted.

Additional notes

Reference [55] was identified through Semantic Scholar; its publisher landing page was unavailable during verification on 17 June 2026.

Human-verified editorial review Verified by World ID proof-of-human. This editorial layer was submitted from a SAIMSARA account verified as a unique human.

Generated with SAIMSARA v5.1

Abstract: This scoping review aims to systematically map the existing experimental and cross-sectional evidence on the generation and detection of medical deepfakes, synthesizing the performance characteristics of detection algorithms, the realism of synthetic medical data, and the emerging clinical and infrastructural implications of this dual-use technology. The review uses 56 references and builds its evidence map from 57 original studies with 171646 total participants/sample observations (topic-deduplicated ΣN). The mapped evidence highlights a clear asymmetry in the medical deepfake landscape: automated detectors achieve near-perfect performance in controlled settings, yet synthetic images remain realistic enough to deceive clinicians and even advanced language models. This dual-use tension was recurrent, with lightweight detectors exceeding 99% accuracy on benchmark CT data while radiologists distinguished synthetic from authentic radiographs at only 75% accuracy. Generative approaches were simultaneously associated with legitimate uses in data augmentation and privacy-preserving de-identification, suggesting both opportunity and risk. Because most evidence derives from in-silico experiments rather than validated clinical workflows, the practical vulnerability at the human-AI interface remains unresolved. Future work should prioritize prospectively validated human-AI teaming protocols and adversarially robust, cross-modal benchmarks to determine whether detection performance translates into protection of diagnostic integrity in real clinical practice.
.

Keywords: medical deepfake detection; AI-generated medical images; deep learning forensics; CT scan manipulation; MRI deepfake; adversarial robustness; synthetic medical data; generative adversarial networks; healthcare cybersecurity; medical image integrity

Review Stats

Final search date and database lock: 2026-06-04 14:25:52 CEST
Plan: Pro (expanded craft tokens; source: Semantic Scholar)
Source: Semantic Scholar
Total Abstracts/Papers: 4497697
Downloaded Abstracts/Papers: 1000
Included original and non-original Abstracts/Papers (all): 92
Included original Abstracts/Papers (Vote counting by direction of effect): 57
Reference Index (links used in paper): 56
Total participants/sample observations (topic deduplicated ΣN): 171646

Get access to the full paper

Unlock the full evidence map

Full paper access includes the complete human-readable review, figures, reference index, PDF export, and machine-readable Evidence JSON download.
Evidence JSON can also be purchased separately if you only need the LLM-ready object for agent, AI, or RAG workflows.
Institutional or library access? Sign in with your institution email to open all available SAIMSARA papers under your institution access arrangement.
Need a SAIMSARA review on your own topic? ☸️Request.

Reference Index (56)

[1] Toward Medical Deepfake Detection: A Comprehensive Dataset and Novel Method — https://doi.org/10.1007/978-3-032-05185-1_60
[2] A Hybrid Approach for Medical Deepfake Detection Using Depth-Wise Convolutions in Vision Transformer and Frequency Domain Analysis — https://doi.org/10.1007/978-3-031-97632-2_16
[3] Exploring Ensemble Learning Strategies for Robust Medical Deepfake Detection — https://doi.org/10.1109/ciscon66933.2025.11337869
[4] MedFakeDet: multi-organ medical deepfake detection and inpainting localization system via effective deep learning models and density maps — https://doi.org/10.1016/j.softx.2026.102668
[5] Explainable and proactive fragile watermarking for medical Deepfake detection — https://doi.org/10.1016/j.iswa.2026.200684
[6] Benchmarking Deep Learning Architectures for Deepfake Detection in Medical Images — https://doi.org/10.48198/njpas/24.b38
[7] A Lightweight Deep Learning Framework for Real-time Detection of Deepfake Medical Images — https://doi.org/10.1109/amathe65477.2025.11081304
[8] Detection of Deepfake Medical Images Based on Spatial and Frequency Domain Analysis — https://doi.org/10.1109/cicn63059.2024.10847427

Unlock the full paper to view the complete Reference Index.