Nilanjan Panda

The paper introduces a dynamical manifold theory of cognitive difficulty using excitable neural networks, showing difficulty emerges as a multidimensional structure rather than a single scalar value.

This paper develops a gradient-boosted model using molecular fingerprints to predict blood-brain barrier permeability, achieving high accuracy and interpretability.

This paper introduces a hybrid SEIR-FitzHugh-Nagumo model to simulate information flow in neural networks, showing improved directional information flow and reduced latency compared to conventional models.

The paper introduces a quantum-mechanochemical model to explain neurotransmitter release latency by combining proton-coupled electron transfer and mechanical barrier softening.

The paper studies adaptive control of neural synchrony in modified FHN networks with variance-driven coupling, showing topology-dependent effects on synchrony and introducing a feedback controller for desynchronization.