Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Computed Tomography

Computed Tomography

Tomography refers to imaging by sections. Computed tomography (CT) is a non-invasive imaging technique that uses computers to analyze several cross-sectional X-rays to reveal minute details about structures in the body.
The technique was invented in the 1970s and is based on the principle that as X-rays pass through the body, they are absorbed or reflected at different levels. In the technique, a patient lies on a motorized platform while a computerized axial tomography (CAT) scanner rotates...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Association of Pretreatment Serum Albumin and Systemic Inflammatory Markers with Pathologic Response to Neoadjuvant Chemotherapy in Breast Cancer.

Journal of clinical medicine·2026

Same author

YOLO26x-based automated fracture detection on radiographs and its impact on radiologist performance: A multi-reader multi-case study.

European journal of radiology·2026

Same author

Liver Background Uptake of Prostate-specific Membrane Antigen-targeted PET Radiotracers: A Systematic Review and Meta-Analysis.

Radiology. Imaging cancer·2026

Same author

Impact of Deep-Learning Reconstruction on MRI Workflows: A Retrospective Analysis at a Large Academic Tertiary Center.

RoFo : Fortschritte auf dem Gebiete der Rontgenstrahlen und der Nuklearmedizin·2026

Same author

Photon Counting CT of the Head: Retrospective Analysis of Image Quality and Dose in Comparison to Energy Integrating Detector CT.

Academic radiology·2026

Same author

Timing and Pattern of Extraosseous Dissemination Are Key Determinants of Survival in Multiple Myeloma.

American journal of hematology·2026

Same journal

Separating Signal from Noise: When Deep Learning Clarifies Physiology in MR Lymphangiography of Lower Extremity.

Radiology·2026

Same journal

Congratulations to the 2026 Editorial Fellows.

Radiology·2026

Same journal

Ten-year Longitudinal Relationship between Spinal Degenerative Lesions in Axial Spondyloarthritis at MRI and Radiography in the DESIR Cohort.

Radiology·2026

Same journal

Standardized Knee Meniscus MRI Reporting: An Interdisciplinary Delphi Consensus.

Radiology·2026

Same journal

Mind the Gap: Cardiac MRI and the Future of Heart Failure Risk Prediction.

Radiology·2026

Same journal

Beyond Static Prediction: Tracking AI Breast Cancer Risk over Time.

Radiology·2026

See all related articles

Search research articles

Related Experiment Video

Updated: Mar 25, 2026

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

The Rise of Deepfake Medical Imaging: Radiologists' Diagnostic Accuracy in Detecting ChatGPT-generated Radiographs.

Mickael Tordjman^1,2, Murat Yuce^1,2, Amine Ammar³

¹BioMedical Engineering and Imaging Institute, Icahn School of Medicine at Mount Sinai, 1470 Madison Ave, New York, NY 10029.

|March 24, 2026

Summary

This summary is machine-generated.

Synthetic medical images, or deepfakes, generated by large language models (LLMs) are difficult for radiologists and AI to detect. Training is crucial to identify these artificial intelligence-generated radiographs and mitigate risks.

More Related Videos

Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules

Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules

Published on: May 19, 2023

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Published on: February 23, 2024

Related Experiment Videos

Last Updated: Mar 25, 2026

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules

Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules

Published on: May 19, 2023

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Published on: February 23, 2024

Area of Science:

Radiology
Artificial Intelligence
Medical Imaging

Background:

Large language models (LLMs) can create realistic synthetic medical images (deepfakes).
These AI-generated images raise concerns regarding potential misuse in clinical settings.
Assessing the detectability of synthetic radiographs is vital for maintaining diagnostic integrity.

Purpose of the Study:

To evaluate the capability of radiologists and multimodal LLMs in distinguishing synthetic radiographs generated by ChatGPT from authentic clinical images.
To compare the diagnostic accuracy of human readers and various AI models in identifying deepfake medical images.
To identify common characteristics of synthetic radiographs that may aid in their detection.

Main Methods:

A retrospective diagnostic accuracy study involving 17 radiologists and four LLMs (GPT-4o, GPT-5, Gemini 2.5 Pro, Llama 4 Maverick).
Radiologists assessed image quality and provided diagnoses for both synthetic (ChatGPT-generated) and authentic radiographs.
LLMs and radiologists performed classification tasks to differentiate between synthetic and authentic images across multiple datasets.

Main Results:

Purpose-blinded radiologists spontaneously identified artificial intelligence-generated radiographs in 41% of cases.
No significant difference in radiologist accuracy was found between GPT-4o-generated (75%) and RoentGen-generated (70%) synthetic images.
While no LLM detected all synthetic radiographs, GPT-4o (85%) and GPT-5 (83%) showed higher accuracy than Gemini 2.5 Pro (56%) and Llama 4 Maverick (59%) in differentiating GPT-4o-generated images.

Conclusions:

Synthetic radiographs generated by LLMs are challenging for both radiologists and AI models to distinguish from authentic images.
Features such as bilateral symmetry and unnatural textures characterize synthetic radiographs.
Developing effective training protocols for physicians and AI systems to recognize synthetic medical images is essential to address potential risks.