AI for Clinical Medicine

Pioneering the integration of artificial intelligence into clinical practice to transform patient care, advance medical knowledge, and ensure safe deployment of AI in healthcare.

Dr. Isaac Kohane, MD, PhD

Isaac "Zak" Kohane, MD, PhD

Professor & Chair, Department of Biomedical Informatics

Editor-in-Chief, NEJM AI

What I'm working on now →

Research Focus

Our work spans three interconnected domains that are reshaping the future of medicine

Clinical AI

Developing AI tools that enhance diagnosis, treatment, and clinical decision-making. From early detection of growth disorders to diagnosis of rare diseases.

  • Predictive modeling for clinical events
  • EHR integration and workflow optimization
  • Rare disease screening and diagnosis
Learn more

Human Values in AI

Ensuring AI systems align with ethical principles, cultural values, and patient preferences. Developing frameworks for value-based clinical decision support.

  • Ethical AI frameworks for healthcare
  • Explicit alignment of AI clinical decisions
  • Patient preference integration
Learn more

Patient Data Control

Empowering patients with granular control over their health data through innovative protocols like PING and Guardian Angel for secure, auditable data sharing.

  • PING: Patient-initiated data granting.
  • Guardian Angel: AI dedicated to patients.
  • People Powered Medicine: Guidelines direct-to-patients.
Learn more

Featured Publications

Key contributions over 30+ years of research

Recent Impact

"The Clinician and Dataset Shift in Artificial Intelligence"

NEJM (2021) - Dataset shift has already impacted healthcare.

Identifies several components of the challenge for AI deployment and regulation in healthcare.

"Compared with What? Measuring AI against the Health Care We Have."

NEJM AI (2024) - Finding the right comparators for clinical AI

Grounding evaluations with respect to the practice we have rather than the clinical care we wish we had.

"The AI Revolution in Medicine: GPT-4 and Beyond"

Pearson (2023) - Conversational lay guide to impact of generative AI on biomedicine.

Early book on LLM transformation of healthcare, research and consumer experiences.

Early Work

"Temporal Reasoning in Medical Expert Systems"

MIT/LCS TR-389 (1987) - Pioneering temporal reasoning across healthcare data types

Applied constraint propagation and context switching techniques.

"Automated Trend Detection with Alternate Temporal Hypotheses"

IJCAI (1993) - Monitoring noisy and sparse clinical data streams.

Collaborative Research Environment

Our lab brings together graduate students, postdocs, and visitors from diverse backgrounds to tackle medicine's most pressing challenges through computational innovation.

30+
Years of Research
500+
Publications
AI in Medicine
PhD Track Director
NEJM AI
Editor-in-Chief

Current Focus Questions

"The skill we most treasure and work to nurture in our lab is the ability to ask interesting, important, and potentially answerable questions."
  • How can AI systems better understand and incorporate patient values?
  • What does patient-controlled data sharing look like in practice?
  • How do we ensure AI recommendations remain robust across populations?