Michał Januszewski

The female larval fly CNS connectome: first draft

The paper releases a complete larval Drosophila CNS connectome with automated segmentations and highlights the need for improved machine vision for proofreading.

Fine ultrastructural organization of glomerular neuropil in the developing zebrafish olfactory bulb

The paper reveals the fine ultrastructural organization of protoglomeruli in the developing zebrafish olfactory bulb, showing their subdivision into microglomeruli with distinct mitral cell arrangements.

Connectome analysis of a cerebellum-like circuit for sensory prediction

The paper uses connectomics to map a cerebellum-like circuit in weakly electric fish, revealing structured synaptic connectivity that supports predictive sensory processing and continual learning.

Sexual dimorphism in the complete connectome of the Drosophila male central nervous system

The paper presents the complete connectome of the Drosophila male central nervous system and compares it to the female connectome, identifying sex-specific and dimorphic neuronal types.

Olfactory Loss Enhances Visual Learning in Drosophila through Structural and Functional Reorganisation

This study shows that losing olfaction in Drosophila enhances visual learning through structural and functional changes in neural circuits.

Deep anatomical and ultrastructural classification of neurons in the zebrafish olfactory bulb

The paper classifies neuron types and their connectivity in the zebrafish olfactory bulb using ultrastructural features from electron microscopy.

The songbird basal ganglia connectome

This paper presents the first vertebrate basal ganglia connectome, revealing complex connectivity patterns that challenge the traditional view of discrete pathways.

A connectomic resource for neural cataloguing and circuit dissection of the larval zebrafish brain

The paper presents a connectomic dataset of a larval zebrafish brain, enabling circuit analysis through synaptic and neuronal annotations.

Accelerating Neuron Reconstruction with PATHFINDER

PATHFINDER improves neuron reconstruction accuracy and efficiency in mouse brain connectome mapping by reducing error rates and manual effort.

Light-microscopy based dense connectomic reconstruction of mammalian brain tissue

This paper develops LICONN, a light-microscopy based method for synapse-level connectomic reconstruction of brain tissue using hydrogel expansion and deep learning.

Connectome-driven neural inventory of a complete visual system

A FIB-SEM connectome of the male Drosophila right optic lobe sorts ∼53,000 neurons into 727 types, with neurotransmitter predictions and 577 matched split-GAL4 lines.

An ultrastructural map of a spinal sensorimotor circuit reveals the potential of astroglia modulation

The paper maps the ultrastructural organization of a spinal sensorimotor circuit in zebrafish, showing how astroglia interact with synapses to modulate information flow.

Defining the molecular identity and morphology of glia limitans superficialis astrocytes in mouse and human

The paper identifies Myocilin as a marker for glia limitans superficialis astrocytes in mouse and human, showing their specialized morphology and evolutionary conservation.

Paired and solitary ionocytes in the zebrafish olfactory epithelium

The paper characterizes ionocyte subtypes in zebrafish olfactory epithelium, identifying three types with distinct functions and distributions.

A connectomic study of a petascale fragment of human cerebral cortex

The paper presents a connectomic analysis of a human cerebral cortex fragment, revealing structural and connectivity features of neurons and synapses.

Connectomic Analysis of Mitochondria in the Central Brain of Drosophila

The paper analyzes mitochondria in the Drosophila connectome, showing variations in size, location, and appearance across cell types and brain regions.

Multi-Layered Maps of Neuropil with Segmentation-Guided Contrastive Learning

SegCLR, a self-supervised learning method, enables classification of cellular subcompartments and cell types from 3D electron microscopy data with minimal labeled data.

A Connectome of the Male Drosophila Ventral Nerve Cord

Roughly 23 thousand traced neurons and 10 million TBars of the male Drosophila nerve cord, reconstructed from electron microscopy, typed, and queryable in neuPrint down to motor neurons.

Visual recognition of social signals by a tecto-thalamic neural circuit

The paper identifies a tecto-thalamic neural circuit that processes visual social signals and drives social attraction in zebrafish.

Denoising-based Image Compression for Connectomics

The paper develops a compression method for connectomic images using denoising and standard codecs, achieving 17-fold compression with minimal loss in 3D reconstruction and synaptic detection.

Automated Reconstruction of a Serial-Section EM Drosophila Brain with Flood-Filling Networks and Local Realignment

Flood-filling networks with local section realignment segment a forty-teravoxel whole-brain Drosophila electron microscopy volume, making circuit reconstruction an order of magnitude faster than manual tracing.

A Connectome and Analysis of the Adult Drosophila Central Brain

The paper presents a detailed connectome of the adult Drosophila central brain, including new methods for data analysis and biological insights into neural circuit organization.

A Connectome of the Adult Drosophila Central Brain

A dense electron-microscopy reconstruction of half the fly central brain, with nearly every cell typed and queryable online, turns months of tracing partners into a database lookup.

Accelerated EM Connectome Reconstruction using 3D Visualization and Segmentation Graphs

This paper introduces semi-automated strategies combining 3D visualization and machine guidance to accelerate EM connectome reconstruction, showing improved performance over prior methods.

Neuronal Subcompartment Classification and Merge Error Correction

The paper introduces a 3D convolutional neural network method to classify neuronal subcompartments and correct merge errors in brain reconstructions, achieving high accuracy and error detection rates.

GCIB-SEM: A path to 10 nm isotropic imaging of cubic millimeter volumes

The paper introduces GCIB-SEM, a 3D electron microscopy technique that provides <10 nm isotropic resolution for large volumes, improving z-resolution and compatibility with multibeam SEMs.

Learning cellular morphology with neural networks

The paper introduces CMNs to automate morphological analysis of neuronal reconstructions, identifying glia cells and resolving reconstruction errors.

Segmentation-Enhanced CycleGAN

The paper introduces SECGAN, a method that improves neuron reconstruction from electron microscopy data by integrating segmentation components, achieving high accuracy without requiring new ground truth annotations.

High-Precision Automated Reconstruction of Neurons with Flood-filling Networks

This paper introduces flood-filling networks for automated neuron reconstruction, achieving high precision in tracing neurites from electron microscopy data.