Research

Led by Hossein Estiri the Clinical Augmented Intelligence Group’s research is at the intersection of medicine, data science, and phenomenology, developing enriched computational models of human phenotypes.

Optimization Instability in Autonomous Agents

When AI Systems That Improve Themselves Make Themselves Worse.

Building a Digital Clinical Team

Five AI Agents That Critique Each Other — And Outperform Human Prompt Engineering.

Diagnostic Signal Decay and the Signal Fidelity Index

Why AI Models Fail When They Leave Home—And How to Fix It.

The Average Patient Fallacy

Why Medical AI Fails the Patients Who Need It Most—And How to Fix It.

The AI Black Box

A Looming Threat to Medical Transparency.

Post-COVID AI Symposium

Boston, MA - September 20th, 2023.

May 23 MGH COVID AI Symposium

Agenda.

Temporal SARS-CoV-2 Severity Estimates

estimating and predicting the true severity of SARS-CoV-2.

CLAI

a lab for CLinical Augmented Intelligence.

transitive sequential pattern mining (tSPM)

Mining meaningful temporal representations from clinical data.

An ICD story

story of how we find true moments of individual records from the EHR with tSPM.

Research Impact

Research with national and international impact.

MLOps for life sciences

Instrumenting Machine Learning Operations for Clinical Research.

Evaluating Bias in Medical AI

An objective framework for evaluating unrecognized bias in medical AI models.

Breaking New Ground; Generative Models in Informatics and Computational Medicine

Introducing Generative Transfer Learning.

Generative models for computational phenotyping

A generative modeling approach for ML-based phenotyping with observational data.

Thinkin' AI with MLHO

A data-centric approach AI/ML system that learns over time.