Biology subjects

Lavoie-Cardinal, F.

Publications and source records attributed to Lavoie-Cardinal, F..

11 recordsLinked to original sources

bioRxiv2026

PhenoMe: An interactive AI framework for interpretable, microscopy-driven phenotyping

Extracting meaningful biological patterns from microscopy image datasets requires tools that quantitatively connect visual similarities and experimental context. Deep learning models, and in particular novel foundation models, have the capacity to translate images into high-dimensional embeddings that can better represent complex visual patterns. Using these complex representations analytically and linking them to experimental information, it is possible to build advanced exploratory and phenotyping capabilities. Here, we present PhenoMe, an open-source, model-agnostic framework to handle pretrained foundation model-derived image representations alongside quantitative image properties and experimental metadata. PhenoMe integrates multiple scriptable and interactive workflows for embedding space visualisation, dimensionality reduction and clustering, similarity-based comparison and retrieval of representative images for each phenotype, among others. It also supports correlation analyses that link the embedding structure to interpretable image properties. We demonstrate PhenoMe's capabilities on seven public microscopy datasets, ranging from cell painting assays, nanoscale synaptic images, drug-response analyses and dynamic cell-phenotyping. These case studies show how readily available image representations can support data contextualisation, curation and analysis. Through guided notebooks, reproducible configurations, extensible model wrappers and a napari plugin, PhenoMe makes representation-based microscopy exploration accessible to researchers without dedicated technical expertise. PhenoMe lays the groundwork for applying foundation models and representation learning to microscopy data analysis.

bioinformatics↗

bioRxiv2026

Calcium transient detection and segmentation with the astronomically motivated algorithm for background estimation and transient segmentation (Astro-BEATS)

SignificanceFluorescence-based Ca2+-imaging is a powerful tool for studying localized neuronal activity, including miniature Synaptic Calcium Transients, providing real-time insights into synaptic activity. These transients induce only subtle changes in the fluorescence signal, often barely above baseline, which poses a significant challenge for automated synaptic transient detection and segmentation. AimDetecting astronomical transients similarly requires efficient algorithms that will remain robust over a large field of view with varying noise properties. We leverage techniques used in astronomical transient detection for miniature Synaptic Calcium Transient detection in fluorescence microscopy. ApproachWe present Astro-BEATS, an automatic miniature Synaptic Calcium Transient segmentation algorithm that incorporates image estimation and source-finding techniques used in astronomy and designed for Ca2+-imaging videos. Astro-BEATS uses the Rolling Hough Transform filament detector to construct an estimate of the expected (transient-free) fluorescence signal of both the dendritic foreground and the background. Subtracting this baseline signal yields difference images displaying transient signals. We use Density-Based Spatial Clustering of Applications with Noise to find sources clustered in spatial and temporal space. ResultsAstro-BEATS outperforms current threshold-based approaches for synaptic Ca2+ transient detection and segmentation. The produced segmentation masks can be used to train a supervised deep learning algorithm for improved synaptic Ca2+ transient detection in Ca2+-imaging data. The speed of Astro-BEATS and its applicability to previously unseen datasets without re-optimization makes it particularly useful for generating training datasets for deep learning-based approaches. ConclusionAstro-BEATS greatly reduces the time needed for the annotation of synaptic Ca2+ transient and removes the significant overhead of human expert annotation, enabling consistent analysis of new Ca2+-imaging datasets.

bioinformatics↗

bioRxiv2025

Enhancing STED Microscopy via Fluorescence Lifetime Unmixing and Filtering in Two-Species SPLIT-STED

Simultaneous super-resolution imaging of multiple fluorophores remains a major challenge in STimulated Emission Depletion (STED) microscopy due to spectral overlap of STED-compatible fluorophores. The combination of STED microscopy and Fluorescence Lifetime Imaging Microscopy (FLIM) offers a powerful alternative for super-resolved, multiplexed imaging of biological samples but is hindered by lifetime convergence at high depletion powers. Here, we present an analysis method, two-species Separation of Photons by LIfetime Tuning (SPLIT)-STED, that uses a linear system of equations in phasor-based STED-FLIM to enhance both fluorophore unmixing and spatial resolution. It defines the fluorescence signal as a mixture of three lifetime components: the two target fluorophores and a short-lifetime contribution from undepleted peripheral fluorescence photons. Two-species SPLIT-STED disentangles overlapping lifetimes and selectively filters low-resolution signal. The method enables accurate unmixing of spectrally overlapping fluorophores and, by enhancing resolution through lifetime-based filtering, allows the use of lower depletion powers, thereby improving fluorescence lifetime separation.

bioinformatics↗

bioRxiv2025

A Self-Supervised Foundation Model for Robust and Generalizable Representation Learning in STED Microscopy

Foundation Models (FMs) have dramatically increased the potential and power of deep learning algorithms through general capacities over a variety of tasks. The performance increase they offer is obtained without elaborated specific trainings for domains such as natural language processing and computer vision. However, their application in specialized fields like biomedical imaging and fluorescence microscopy remains difficult due to distribution shifts and the scarcity of high-quality annotated datasets. The high cost of data acquisition and the requirement for in-domain expertise further exacerbate this challenge in microscopy. To address this we introduce STED-FM, a foundation model specifically designed for super-resolution STimulated Emission Depletion (STED) microscopy. STED-FM leverages a Vision Transformer architecture trained at scale with Masked Autoencoding on a new dataset of nearly one million STED images. STED-FM learns expressive latent representations without requiring extensive annotations, yielding robust performance across diverse downstream microscopy image analysis tasks. Unsupervised experiments demonstrate the discriminative structure of its learned latent space. These representations can be leveraged for multiple downstream applications, including fully supervised classification and segmentation with reduced annotation requirements. Moreover, STED-FM representations enhance the performance of deep learning-based image denoising and improve the quality of images generated by diffusion models, enabling latent attribute manipulation for the data-driven discovery of subtle nanostructures and phenotypes, as well as algorithmic super-resolution. Moreover, its powerful structure retrieval capabilities are integrated into automated STED microscopy acquisition pipelines, paving the way for smart microscopy. In sum, we demonstrate that STED-FM lays a robust foundation for state-of-the-art algorithms across a wide array of tasks, establishing it as a highly valuable and scalable resource for researchers in super-resolution microscopy.

bioinformatics↗

bioRxiv2024

Quantitative Analysis of Miniature Synaptic Calcium Transients Using Positive Unlabeled Deep Learning

Ca2+ imaging methods are widely used for studying cellular activity in the brain, allowing detailed analysis of dynamic processes across various scales. Enhanced by high-contrast optical microscopy and fluorescent Ca2+ sensors, this technique can be used to reveal localized Ca2+ fluctuations within neurons, including in sub-cellular compartments, such as the dendritic shaft or spines. Despite advances in Ca2+ sensors, the analysis of miniature Synaptic Calcium Transients (mSCTs), characterized by variability in morphology and low signal-to-noise ratios, remains challenging. Traditional threshold-based methods struggle with the detection and segmentation of these small, dynamic events. Deep learning (DL) approaches offer promising solutions but are limited by the need for large annotated datasets. Positive Unlabeled (PU) learning addresses this limitation by leveraging unlabeled instances to increase dataset size and enhance performance. This approach is particularly useful in the case of mSCTs that are scarce and small, associated with a very small proportion of the foreground pixels. PU learning significantly increases the effective size of the training dataset, improving model performance. Here, we present a PU learning-based strategy for detecting and segmenting mSCTs. We evaluate the performance of two 3D deep learning models, StarDist-3D and 3D U-Net, which are well established for the segmentation of small volumetric structures in microscopy datasets. By integrating PU learning, we enhance the 3D U-Nets performance, demonstrating significant gains over traditional methods. This work pioneers the application of PU learning in Ca2+ imaging analysis, offering a robust framework for mSCT detection and segmentation. We also demonstrate how this quantitative analysis pipeline can be used for subsequent mSCTs feature analysis. We characterize morphological and kinetic changes of mSCTs associated with the application of chemical long-term potentiation (cLTP) stimulation in cultured rat hippocampal neurons. Our data-driven approach shows that a cLTP-inducing stimulus leads to the emergence of new active dendritic regions and differently affects mSCTs subtypes.

bioinformatics↗

bioRxiv2024

Development of AI-assisted microscopy frameworks through realistic simulation in pySTED

The integration of artificial intelligence (AI) into microscopy systems significantly enhances performance, optimizing both the image acquisition and analysis phases. Development of AI-assisted super-resolution microscopy is often limited by the access to large biological datasets, as well as by the difficulties to benchmark and compare approaches on heterogeneous samples. We demonstrate the benefits of a realistic STED simulation platform, pySTED, for the development and deployment of AI-strategies for super-resolution microscopy. The simulation environment provided by pySTED allows the augmentation of data for the training of deep neural networks, the development of online optimization strategies, and the training of reinforcement learning models, that can be deployed successfully on a real microscope.

bioinformatics↗

bioRxiv2022

Chronic stress exposure alters the gut barrier: sex-specific effects on microbiota and jejunum tight junctions

Major depressive disorder (MDD) is the leading cause of disability worldwide. However, 30-50% of patients are unresponsive to commonly prescribed antidepressants, highlighting untapped causal biological mechanisms. Dysfunction in the microbiota-gut-brain axis, the bidirectional communications between the central nervous system and gastrointestinal tract that are modulated by gut microorganisms, has been implicated in MDD pathogenesis. Exposure to chronic stress disrupts blood-brain barrier integrity, still, little is known about intestinal barrier function in these conditions particularly for the small intestine where most food and drug absorption takes place. Thus, here we investigate how chronic social or variable stress, two mouse models of depression, impact the jejunum (JEJ) intestinal barrier in males and females. Mice were subjected to stress paradigms followed by analysis of gene expression profiles of intestinal barrier-related targets, fecal microbial composition, and blood-based markers. Altered microbial populations as well as changes in gene expression of JEJ tight junctions were observed depending on the type and duration of stress, with sex-specific effects. We took advantage of machine learning to characterize in detail morphological tight junction properties identifying a cluster of ruffled junctions in stressed animals. Junctional ruffling is associated with inflammation, so we evaluated if LPS injection recapitulates stress-induced changes in the JEJ and observed profound sex differences. Finally, LPS-binding protein (LBP), a marker of gut barrier leakiness, was associated with stress vulnerability in mice and translational value was confirmed on blood samples from women with MDD. Our results provide evidence that chronic stress disrupts intestinal barrier homeostasis in conjunction with the manifestation of depressive-like behaviors in a sex-specific manner in mice and possibly, human depression.

neuroscience↗

bioRxiv2021

Task-Assisted GAN for Resolution Enhancement and Modality Translation in Fluorescence Microscopy

AbstractWe introduce a deep learning model that predicts super-resolved versions of diffraction-limited microscopy images. Our model, named Task- Assisted Generative Adversarial Network (TA-GAN), incorporates an auxiliary task (e.g. segmentation, localization) closely related to the observed biological nanostructures characterization. We evaluate how TA-GAN improves generative accuracy over unassisted methods using images acquired with different modalities such as confocal, brightfield (diffraction-limited), super-resolved stimulated emission depletion, and structured illumination microscopy. The generated synthetic resolution enhanced images show an accurate distribution of the F-actin nanostructures, replicate the nanoscale synaptic cluster morphology, allow to identify dividing S. aureus bacterial cell boundaries, and localize nanodomains in simulated images of dendritic spines. We expand the applicability of the TA-GAN to different modalities, auxiliary tasks, and online imaging assistance. Incorporated directly into the acquisition pipeline of the microscope, the TA-GAN informs the user on the nanometric content of the field of view without requiring the acquisition of a super-resolved image. This information is used to optimize the acquisition sequence, and reduce light exposure. The TA-GAN also enables the creation of domain-adapted labeled datasets requiring minimal manual annotation, and assists microscopy users by taking online decisions regarding the choice of imaging modality and regions of interest.

bioinformatics↗

bioRxiv2021

MICRA-Net: MICRoscopy Analysis Neural Network to solve detection, classification, and segmentation from a single simple auxiliary task

High throughput quantitative analysis of microscopy images presents a challenge due to the complexity of the image content and the difficulty to retrieve precisely annotated datasets. In this paper we introduce a weakly-supervised MICRoscopy Analysis neural network (MICRA-Net) that can be trained on a simple main classification task using image-level annotations to solve multiple the more complex auxiliary semantic segmentation task and other associated tasks such as detection or enumeration. MICRA-Net relies on the latent information embedded within a trained model to achieve performances similar to state-of-the-art architectures when no precisely annotated dataset is available. This learnt information is extracted from the network using gradient class activation maps, which are combined to generate detailed feature maps of the biological structures of interest. We demonstrate how MICRA-Net significantly alleviates the Expert annotation process on various microscopy datasets and can be used for high-throughput quantitative analysis of microscopy images.

bioinformatics↗

bioRxiv2020

Neuronal activity remodels the F-actin based submembrane lattice in dendrites but not axons of hippocampal neurons

The nanoscale organization of the F-actin cytoskeleton in neurons comprises membrane-associated periodical rings, bundles, and longitudinal fibers. The F-actin rings have been observed predominantly in axons but only sporadically in dendrites, where fluorescence nanoscopy reveals various patterns of F-actin arranged in mixed patches. These complex dendritic F-actin patterns pose a challenge for investigating quantitatively their regulatory mechanisms. We developed here a weakly supervised deep learning segmentation approach of fluorescence nanoscopy images of F-actin in cultured hippocampal neurons. This approach enabled the quantitative assessment of F-actin remodeling, revealing the disappearance of the rings during neuronal activity in dendrites, but not in axons. The dendritic F-actin cytoskeleton of activated neurons remodeled into longitudinal fibers. We show that this activity-dependent remodeling involves Ca2+ and NMDA-dependent mechanisms. This highly dynamic restructuring of dendritic F-actin based submembrane lattice into longitudinal fibers may serve to support activity-dependent membrane remodeling, protein trafficking and neuronal plasticity.

neuroscience↗

bioRxiv2020

Nano-positioning and tubuline conformation determine transport of mitochondria along microtubules

Correct spatiotemporal distribution of organelles and vesicles is crucial for healthy cell functioning and is regulated by intracellular transport mechanisms. Controlled transport of bulky mitochondria is especially important in polarized cells such as neurons that rely on these organelles to locally produce energy and buffer calcium. Mitochondrial transport requires and depends on microtubules which fill much of the available axonal space. How mitochondrial transport is affected by their position within the microtubule bundles is not known. Here, we found that anterograde transport, driven by kinesin motors, is susceptible to the molecular conformation of tubulin both in vitro and in vivo. Anterograde velocities negatively correlate with the density of elongated tubulin dimers, similar to GTP-tubulin, that are more straight and rigid. The impact of the tubulin conformation depends primarily on where a mitochondrion is positioned, either within or at the rim of microtubule bundle. Increasing elongated tubulin levels lowers the number of motile anterograde mitochondria within the microtubule bundle and increases anterograde transport speed at the microtubule bundle rim. We demonstrate that the increased kinesin step processivity on microtubules consisting of elongated dimers underlies increased mitochondrial dynamics. Our work indicates that the molecular conformation of tubulin controls mitochondrial motility and as such locally regulates the distribution of mitochondria along axons.

neuroscience↗