Medical AI Research & Publications

Born at the MIT Center for Collective Intelligence, Centaur.ai continues to partner with researchers at leading institutions all over the world to advance the field of AI.

featured research

Nonexpert crowds outperform expert individuals in diagnostic accuracy

In collaboration with researchers at Memorial Sloan Kettering

The Collective Intelligence Methodology centaur.ai

Want to learn more about our
research-backed methodology?

Download our publications →

Research

MedPAIR: Measuring Physicians and AI Relevance Alignment in Medical Question Answering

With MIT, Cornell University, Oxford University, University of Potsdam, and the Hasso Plattner Institute

Research

A Dataset of Medical Questions Paired with Automatically Generated Answers and Evidence-supported References

In partnership with the NIH and the National Library of Medicine

Research

Evaluating crowdsourcing for ICU EEG annotation

In Collaboration with the NIH and the National Science Foundation

Research

The use of LLMs to enhance clinical trial educational materials

In collaboration with Dana Farber Cancer Institute and Loyola University

Research

Can crowdsourced annotations improve AI-based congestion scoring?

In collaboration with the MICCAI Society in agreement with Springer

Research

A randomized controlled educational pilot trial of interictal epileptiform discharge identification for neurology residents

Presented at the American Academy of Neurology annual meeting

Research

Real-time near infrared AI using scalable non-expert crowdsourcing

In collaboration with Activ Surgical and The OSU Wexner Medical Center

Research

Leveraging AI to determine fetal sex

Published at the Society of Maternal Fetal Medicine Annual Conference

Research

Expert-level annotation quality via gamified crowdsourcing

In collaboration with researchers at Brigham and Women’s Hospital

Research

Boosting the wisdom of the crowd for image annotation

In collaboration with researchers at Indiana University

Research

On-demand EEG education through competition

In collaboration with researchers at Massachusetts General Hospital

Research

Gamified crowdsourcing for dataset labeling

In collaboration with researchers at Brigham and Women’s Hospital

Research

Deep learning for detection and localization

Originally published in the IEEE Journal of Biomedical and Health Informatics

Research

Comparison of pulmonary congestion severity using AI-assisted scoring versus clinical experts

An analysis from the NHLBI-sponsored BLUSHED-AHF study

Research

Crowd-sourcing delineations of disease affected skin in patient photographs

Originally published in JMIR Dermatology

Research

Artificial intelligence for fetal sex determination

Originally published in the American Journal of Obstetrics & Gynecology

Research

Agreement between experts and an untrained crowd fusing a gamified app

Originally published in JMIR Medical Informatics

Research

Towards transparency in image datasets with annotations by experts, crowds, and an algorithm

In collaboration with researchers at MIT and Stanford

Research

The viability of an AI/ML learning prediction model to determine surgical candidates

Originally published in the The Journal of Arthroplasty

Research

How one block of trials influences the next

In collaboration with researchers at Brigham and Women’s Hospital

Research

The American Journal of Psychotherapy, Both Old and New

Originally published in the American Journal of Psychotherapy