Hanspeter Pfister

The paper presents a challenge for 3D registration of expansion microscopy volumes, evaluates top methods, and highlights future directions for ExM registration.

This study develops a quantitative framework to distinguish between clock and domino mechanisms in embryogenesis, finding domino governs cleavage timings while a clock mechanism governs morula and blastocyst timings.

The paper introduces niiv, a self-supervised method for isotropic reconstruction of 3D microscopy data, achieving faster and higher-quality results than existing methods.

SmartEM accelerates electron microscopy image acquisition by using machine learning to intelligently allocate imaging time, achieving up to 7-fold speedup in connectomic samples.

MoMo is an interactive visualization framework that analyzes neuron morphology-aware motifs in connectomes, enabling efficient subgraph isomorphism searches and intuitive exploration through a sketch-based interface.

SynAnno is an interactive tool that streamlines synaptic annotation proofreading in connectomics by integrating guided workflows and machine learning assistance, improving efficiency and reducing errors.

SEAL integrates spatial imaging context with dimensionality reduction embeddings to improve interpretation of complex datasets through interactive visualization and feature importance analysis.

This paper develops a deep learning method to correct non-uniform milling artifacts in FIB-SEM images using unsupervised cross-plane image-to-image translation.

This study investigates synaptic connectivity development in the mouse cerebellum, revealing sequential input-dependent wiring and individualized parallel fiber inputs to Purkinje cells.

This paper introduces a method that combines autoencoders with segmentation to improve single-cell analysis of multiplexed tissue images by capturing morphology and reducing signal spillover.

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

CAVE enables reproducible connectome analysis in large-scale datasets during ongoing annotation and proofreading.

EM-Compressor uses Variational Autoencoders to compress EM images for connectomics, reducing data size to 1/128th while preserving features needed for segmentation.

This paper introduces a point affinity transformer framework for global neuron shape reasoning, improving segmentation error correction and neuron type classification in connectomics.

psudo is an interactive system that optimizes pseudocolor palettes for multichannel biomedical images to enhance perceptual distinction and reduce color confusion in overlapping channels.

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

The paper develops an AI tool for quantifying ultrastructural traits from 3D microscopic datasets and improves fixation protocols for plants.

Vimo enables interactive visualization of neuronal connectivity motifs in large brain networks, allowing experts to sketch and analyze motif instances and chains in 3D.

The paper presents a vCLEM protocol to map Aβ and pTau in AD mouse hippocampus connectomics data, revealing new molecular and structural alterations.

This paper introduces a method for saturated reconstruction of living brain tissue using advanced imaging and deep learning to achieve high-resolution, volumetric analysis of neural structures over time.

The paper introduces a vCLEM method using fluorescent scFvs to combine molecular labeling with ssEM, revealing new cell types and subcellular features in mouse cerebellar cortex.

The paper introduces two methods to improve tissue segmentation in highly multiplexed images using real augmentations and nuclear envelope imaging, resulting in more accurate models.

PO_SCPLOWEAXC_SCPLOW enables interactive visual pattern search in sequential data using unsupervised deep representation learning, outperforming existing techniques in retrieving similar patterns.

Facetto combines unsupervised and supervised learning to analyze multi-channel microscopy images, enabling discovery of cell types and phenotypes in cancer tissues.

The paper shows that during cerebellar development, climbing fibers add synapses to a subset of Purkinje cells through positive feedback, without pruning other connections, and that this process is economical due to matching axon and Purkinje cell numbers.

This paper introduces Scalable Insets, a technique for exploring annotated patterns in multiscale visualizations, improving search speed and comparison accuracy.

HiGlass, a web-based, containerized viewer, enables continuous multiscale navigation of Hi-C contact maps alongside genomic tracks using composable views linked by location and zoom.

HiPiler is an interactive visualization tool for exploring large genome interaction matrices, enabling users to define and analyze regions of interest through thumbnails and linked exploration.