Matthew Tyler Kaufman

The paper infers motor cortex inputs during movement tasks and finds similar encoding for initial and corrective movements despite nonidentifiability challenges.

The paper reanalyzes neural recordings from rhesus monkeys during grasping tasks and finds that a flexible dynamical model applies to grasp control, showing conserved rotational frequencies and reoriented state-space planes.

The paper uses two-photon calcium imaging to analyze neural activity in mouse sensorimotor cortex during a reach-to-grasp task, revealing overlapping subpopulations across area borders.

This study uses machine vision to analyze larval Drosophila crawling and finds regional differences in muscle coordination during locomotion.

This study investigates how motor commands for reaching and grasping are represented in the mouse sensorimotor cortex, finding shared and distinct roles in encoding joint angles and target-specific signals.

The paper investigates how task-relevant visual information is routed from the PPC to M1 in mice during a visuomotor task, finding that PPC-M1 projections prioritize movement over visual signals.

This study identifies location-dependent rotational dynamics in motor cortex during reaching, improving decoding of movement kinematics from neural activity.

This study investigates how microstimulation of somatosensory cortex affects motor cortex activity in paralyzed individuals, finding task-dependent spatial patterns and showing that biomimetic stimulation improves control of a virtual hand.

This paper introduces RADICaL, a deep learning framework that infers single-trial neural population dynamics from calcium imaging with sub-frame temporal resolution, showing improved accuracy in synthetic and real mouse data.

The paper shows that neuronal population dynamics in motor cortex differ between reaching and grasping movements, suggesting different circuit mechanisms for each task.

BRICseq maps brain-wide connectivity, activity, and gene expression in single animals, showing gene expression patterns predict connectivity and activity correlations.

The paper shows that neural activity during expert task performance is dominated by varied, uninstructed movements, revealing that movement-related activity often outweighs task-specific computations.

The paper shows that neural dynamics can be accurately estimated without spike sorting by using multi-unit threshold crossings, enabling new analyses and electrode array design.

This paper introduces LFADS, a deep learning method to infer single-trial neural population dynamics from spiking data, demonstrating improved accuracy in predicting behavior and extracting dynamics across datasets.

OnACID processes streaming calcium imaging data in real time by combining online dictionary learning with motion correction, source extraction, and activity deconvolution, tracking hundreds of neurons for closed-loop experiments.