Samik Banerjee

The paper introduces a method combining deep learning and Discrete Morse techniques to skeletonize neuronal processes from tracer injection data, improving the accuracy of axon length density estimation.

The paper presents a multimodal, multiscale atlas of the mouse brain and head with high-resolution histology and co-registered MRI data, enabling better integration of molecular and connectivity data.

This study creates a 3D histological cell atlas of the developing human brain, providing detailed cell counts, proliferative dynamics, and cell type distribution at 13-14 and 20-21 gestational weeks.

The paper introduces a hybrid deep learning architecture combining topological priors with encoder-decoder networks to improve semantic segmentation of neuroanatomical data with high precision and recall.

This paper introduces a generative framework to register multimodal, incomplete neuroimaging data into a reference brain atlas, enabling geometric quantification of biological and technical variation.

This paper introduces topological methods to detect and skeletonize neurons and tracer injections, showing improved performance over existing techniques.

This paper presents an unsupervised algorithm for detecting labeled nuclei in whole mouse brain images, achieving performance close to human accuracy.