Medicine is multimodal, messy, and high-stakes — and I ask one question of every AI system I build: will this work in a real clinical setting?

Meghal Dani

Meghal Dani is a Ph.D. student in Computer Science at the International Max Planck Research School for Intelligent Systems (IMPRS-IS), University of Tübingen, Germany. She is working at the Tübingen AI Center and the Hertie AI institute, advised by Dr. Stefanie Liebe. Following her Bachelor's in Computer Science and Master's in Computational Biology, she worked in industry for two years on medical imaging and 3D vision. Her current research focuses on understanding and adapting foundation models — vision, language, and vision-language — for safe AI in healthcare, with a focus on epilepsy.

Medicine is multimodal, messy, and high-stakes — and I ask one question of every AI system I build: will this work in a real clinical setting?

Dani M, Liebe S.

Parameter-Efficient Self-Supervised Adaptation for EEG-FM under Fixed Computational Budgets

Dani M, Prakash MJ, Rosa F, Akata Z, Liebe S.

Meghal Dani

Parameter-Efficient Self-Supervised Adaptation for EEG-FM under Fixed Computational Budgets

Evaluating large language models for diagnostic reasoning from unstructured clinical narratives in epilepsy