Cell Image Analysis Techniques

The paper introduces a multi-parametric particle-based method to analyze organelle landscapes, enabling visualization of heterogeneous organelle structures and transitional states using fluorescence microscopy and UMAP.

This paper uses video analysis to predict bacterial genotypes from spatiotemporal growth patterns, showing that low-resolution microscopy videos can classify genetic modifications in Vibrio cholerae.

MultiMatch analyzes chain-like particle arrangements in multi-color microscopy using optimal transport, outperforming existing methods in recovering chain structures.

Convpaint provides an interactive pixel classification framework using pretrained CNNs or ViTs, integrated into the napari ecosystem for versatile image processing.

VISION is an open-source software for automated multi-dimensional image analysis of cellular biophysics, enabling high throughput and correlative analysis of biophysical properties.

DIRT/μ automatically extracts root hair traits from 2D images using combinatorial optimization, enabling precise and unbiased measurements of root hair length and density.

MorphoNavigator-3D enables automated, annotation-free single-cell analysis of 3D tumor-immune co-cultures, revealing drug-induced phenotypes and their biological relevance.

EmbSAM provides a framework for 3D cell membrane segmentation in C. elegans embryos using deep learning and the Segment Anything Model, enabling analysis of cell shape dynamics during development.

Flynotyper 2.0 improves upon its predecessor by enhancing computer vision and streamlining analysis for high-throughput Drosophila eye phenotype quantification.

CellSeg3D introduces self-supervised 3D cell segmentation models for fluorescence microscopy, achieving performance comparable to supervised methods without requiring ground truth labels.

The paper introduces ChannelNet, a deep learning architecture for segmenting cells and nuclei in multiplexed images, and demonstrates its effectiveness when combined with InstanSeg.

ColocZStats is a 3D Slicer extension that enables colocalization analysis on z-stacks with user-defined ROIs and thresholding.

ExoJ is an ImageJ2/Fiji plugin that automates the detection and analysis of exocytosis events, providing quantitative features and robust performance across varying signal-to-noise ratios.

The paper introduces GAT, a tool for automated analysis of enteric neurons in gut tissue images, improving accuracy and throughput over manual methods.

SynBot is an open-source software for automated synapse quantification from light microscopy images, improving speed, reproducibility, and ease of use.

UNSEG provides an unsupervised method for segmenting cells and their nuclei in complex tissue samples without training data, outperforming deep learning methods in generalization and consistency.

This paper evaluates a deep learning approach to estimate cell viability without stains by classifying live/dead cells and counting total cells using ResNet CNNs.

This paper introduces a novel image clustering approach for CAR-T/tumor cocultures using U-net and clustering algorithms, demonstrating superior performance in classification and data interpretation.

This study applies FastSAM for automated algal cell segmentation and clustering in water samples, showing high accuracy for single cells and varying clustering performance.

The paper introduces da_Tracker, an automated workflow for high-throughput single-cell and single-phagosome tracking in infected cells, enabling quantitative analysis of pathogen-host interactions.

The paper introduces an image analysis pipeline to segment, quantify, and visualize contraction waves in 2D+t and 3D+t microscopy data of Tribolium castaneum embryos, revealing novel dynamics in extra-embryonic membranes.

This paper introduces MouseMapper, a deep learning tool combined with 3D imaging to analyze whole-body structural changes in obesity, revealing nerve and immune system alterations.

This paper introduces motion blur microscopy to image cell adhesion in whole blood flow without dilution, enabling analysis of adhesion dynamics in physiological conditions.

This paper introduces a semi-automatic segmentation method for bacterial colonies to improve growth characterization efficiency and accuracy.

This study uses subcellular morphology and AI to classify cell types and predict molecular differences without relying on markers or destructive sampling.

This paper introduces embGAN, a deep learning pipeline for automated cell detection and tracking in label-free 3D time-lapse imaging without manual annotation.

ORACLE, a deep-learning model, accurately identifies ovarian cancer cells from healthy cells in 3D co-cultures using nuclear shape, outperforming existing methods.

Human-in-the-loop labeling of nuclear live-cell movies produced DynamicNuclearNet (over 2,621 divisions); Caliban, a tracker trained on it, beat four published trackers at division detection.

PyPlaque is an open-source Python package for analyzing virus plaque assays, offering flexibility and modularity to improve measurement consistency and adaptability to different experimental setups.

This paper describes a postdoctoral training program in bioimage analysis, highlighting its components, outcomes, and implications for training experts in professional skills.