Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Radiological Investigation II: MRI and Ventilation Perfusion Scan01:30

Radiological Investigation II: MRI and Ventilation Perfusion Scan

Description
Magnetic Resonance Imaging (MRI) and Ventilation Perfusion Scans are two radiological investigations that offer detailed diagnostic images of the body, particularly lung structures.
MRI
MRI uses magnetic fields and radiofrequency signals to distinguish between normal and abnormal tissues. This technology provides a more detailed diagnostic image than CT scans, enabling it to characterize pulmonary nodules, stage bronchogenic carcinoma, and evaluate inflammatory activity in...

You might also read

Related Articles

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

Sort by
Same author

Mast cell score associates with wide-spread mast cell symptoms and comorbidities in patients with hEDS and HSD.

medRxiv : the preprint server for health sciences·2026
Same author

Age-related symptom clustering in pediatric hypermobility spectrum disorders: a scoping review.

Orphanet journal of rare diseases·2026
Same author

Limited Feasibility Study of Holographic Display Technology for Interprofessional Team Training.

Healthcare (Basel, Switzerland)·2026
Same author

Prevalence of Intracranial and Cervical Artery Abnormalities in Patients with Hypermobile Ehlers-Danlos Syndrome and Hypermobility Spectrum Disorders Presenting to an Academic Headache Clinic.

Neurology international·2026
Same author

Comparative assessment of left common iliac vein compression in patients with hypermobile Ehlers-Danlos syndrome, hypermobility spectrum disorder and healthy controls - A retrospective single-centre study.

Phlebology·2025
Same author

3D Printing Today, AI Tomorrow: Rethinking Apert Syndrome Surgery in Low-Resource Settings.

Healthcare (Basel, Switzerland)·2025

Related Experiment Video

Updated: Jun 23, 2026

A Pipeline for 3D Multimodality Image Integration and Computer-assisted Planning in Epilepsy Surgery
09:41

A Pipeline for 3D Multimodality Image Integration and Computer-assisted Planning in Epilepsy Surgery

Published on: May 20, 2016

11.3K

The Performance of DeepSeek R1 and Gemini 3 in Complex Medical Scenarios: Comparative Study.

Maria Bajwa1, Robert Hoyt2, Dacre Knight3

  • 1MGH Institute of Health Professions, Boston, MA, United States.

Jmirx Med
|April 27, 2026
PubMed
Summary

This study evaluated two reasoning large language models (LLMs), DeepSeek R1 and Gemini 3 Pro, for diagnostic accuracy in medical scenarios. Gemini 3 Pro generally outperformed DeepSeek R1, especially in open-ended questions, highlighting the importance of model choice for clinical applications.

Keywords:
DeepSeek R1Gemini 3LLMLRMaccuracylarge language modellarge reasoning modelmedical scenario

More Related Videos

Whole-body PET/MRI of Pediatric Patients: The Details That Matter
10:02

Whole-body PET/MRI of Pediatric Patients: The Details That Matter

Published on: December 19, 2017

14.8K
Author Spotlight: Advancing Hepatobiliary and Pancreatic Tumor Treatment with Minimally Invasive Surgical Techniques
03:33

Author Spotlight: Advancing Hepatobiliary and Pancreatic Tumor Treatment with Minimally Invasive Surgical Techniques

Published on: September 27, 2024

1.5K

Related Experiment Videos

Last Updated: Jun 23, 2026

A Pipeline for 3D Multimodality Image Integration and Computer-assisted Planning in Epilepsy Surgery
09:41

A Pipeline for 3D Multimodality Image Integration and Computer-assisted Planning in Epilepsy Surgery

Published on: May 20, 2016

11.3K
Whole-body PET/MRI of Pediatric Patients: The Details That Matter
10:02

Whole-body PET/MRI of Pediatric Patients: The Details That Matter

Published on: December 19, 2017

14.8K
Author Spotlight: Advancing Hepatobiliary and Pancreatic Tumor Treatment with Minimally Invasive Surgical Techniques
03:33

Author Spotlight: Advancing Hepatobiliary and Pancreatic Tumor Treatment with Minimally Invasive Surgical Techniques

Published on: September 27, 2024

1.5K

Area of Science:

  • Artificial Intelligence in Healthcare
  • Medical Diagnostics
  • Clinical Education Technology

Background:

  • Reasoning large language models (LLMs) are increasingly used in healthcare for decision support and education.
  • DeepSeek R1 offers explicit decision-making through chain-of-thought explanations.
  • The Massive Multitask Language Understanding Pro (MMLU-Pro) professional medicine subset provides complex, realistic clinical scenarios for LLM evaluation.

Purpose of the Study:

  • To assess the diagnostic accuracy, reasoning quality, transparency, and usability of DeepSeek R1 and Gemini 3 Pro.
  • To compare model performance on closed- and open-ended clinical scenarios.
  • To guide the application of these LLMs in clinical education and training.

Main Methods:

  • A dual-model evaluation of DeepSeek R1 and Gemini 3 Pro using 162 clinical vignettes from the MMLU-Pro health subset.
  • Construction of closed-ended (multiple-choice) and open-ended prompts for each scenario.
  • Coding of model outputs for accuracy, reasoning steps, and citation behavior, with statistical comparison.

Main Results:

  • Gemini 3 Pro achieved higher accuracy (90.7% closed-ended, 88.9% open-ended) than DeepSeek R1 (86.4% closed-ended, 80.9% open-ended).
  • Both models showed decreased performance on open-ended questions compared to closed-ended ones.
  • Error analysis suggested overthinking in longer reasoning chains; Gemini 3 Pro provided more relevant citations and fewer reasoning steps than DeepSeek R1.

Conclusions:

  • DeepSeek R1 and Gemini 3 Pro show moderate-to-excellent performance in medical scenario evaluation, with Gemini 3 Pro demonstrating higher accuracy.
  • Open-ended evaluations provide deeper insights into reasoning fidelity than closed-ended formats.
  • Further validation is crucial for establishing trustworthiness in clinical and educational settings.