Aayushi Ranjan
Barnard College, Computer Science. Summer research grant recipient. Investigating the neural correlates of music on working memory using EEG and the n-back task.
Principal Investigator
Associate Research Scientist, Neurology
Columbia University Irving Medical Center
Roman Family Faculty, Computer Science
Barnard College, Columbia University
Murad Megjhani, PhD
I work on the problem of knowing what is happening inside a critically ill patient’s brain — without a probe. That means building machine learning systems that recover clinically meaningful signal from the noisy, heterogeneous streams already flowing out of the ICU: transcranial Doppler, arterial blood pressure, EEG. The complementary thread of the Megjhani Lab asks what sound and aesthetic experience can reveal about that same brain — in health, in pain, and in altered states of consciousness.
The work spans neurocritical care AI, foundation models for physiological waveforms, federated learning for multi-center clinical deployment, and the neuroscience of music and aesthetics. The connective tissue is a conviction that better inference — not better hardware — is what will change outcomes.
AHA Second Century Faculty Independence Award — PI (2024–2027)
NIH R01 Co-Investigator (2023–2028)
Current students
Working on hard problems together.
Pari Bhandari
Barnard College, Computer Science. Summer research grant recipient. Developing multimodal data analysis pipelines for sepsis prediction using the MIMIC-IV clinical database.
Justin Zeng
Working on physiological signal analysis and machine learning for clinical monitoring applications in collaboration with the neurocritical care AI program.
Collaborators & co-advising
Park Lab — PHICI, Columbia Neurology
Students in the Megjhani Lab are co-advised through the Park Lab at the Program for Hospital Innovation and Clinical Intelligence (PHICI) in the Department of Neurology at Columbia University Irving Medical Center. The Park Lab focuses on translational machine learning for critical care, and provides a direct clinical context for the Megjhani Lab’s signal-processing and AI work.