3.1 Research program

The recent advent of an increasing number of new microscopy techniques giving access to high throughput screenings and micro or nano-metric resolutions provides a means for quantitative imaging of biological structures and phenomena. To conduct quantitative biological studies based on these new data, it is necessary to develop non-standard specific tools. This requires using a multi-disciplinary approach. We need biologists to define experiment protocols and interpret the results, but also physicists to model the sensors, computer scientists to develop algorithms and mathematicians to model the resulting information. These different expertises are combined within the Morpheme team. This generates a fecund frame for exchanging expertise, knowledge, leading to an optimal framework for the different tasks (imaging, image analysis, classification, modeling). We thus aim at providing adapted and robust tools required to describe, explain and model fundamental phenomena underlying the morphogenesis of cellular and supra-cellular biological structures. Combining experimental manipulations, in vivo imaging, image processing and computational modeling, we plan to provide methods for the quantitative analysis of the morphological changes that occur during development. This is of key importance as the morphology and topology of mesoscopic structures govern organ and cell function. Alterations in the genetic programs underlying cellular morphogenesis have been linked to a range of pathologies.

Biological questions we will focus on include:

1. what are the parameters and the factors controlling the establishment of ramified structures? (Are they really organize to ensure maximal coverage? How are genetic and physical constraints limiting their morphology?),
2. how are newly generated cells incorporated into reorganizing tissues during development? (is the relative position of cells governed by the lineage they belong to?)

Our goal is to characterize different populations or development conditions based on the shape of cellular and supra-cellular structures, e.g. micro-vascular networks, dendrite/axon networks, tissues from 2D, 2D+t, 3D or 3D+t images (obtained with confocal microscopy, video-microscopy, photon-microscopy or micro-tomography). We plan to extract shapes or quantitative parameters to characterize the morphometric properties of different samples. On the one hand, we will propose numerical and biological models explaining the temporal evolution of the sample, and on the other hand, we will statistically analyze shapes and complex structures to identify relevant markers for classification purposes. This should contribute to a better understanding of the development of normal tissues but also to a characterization at the supra-cellular scale of different pathologies such as Alzheimer, cancer, diabetes, or the Fragile X Syndrome. In this multidisciplinary context, several challenges have to be faced. The expertise of biologists concerning sample generation, as well as optimization of experimental protocols and imaging conditions, is of course crucial. However, the imaging protocols optimized for a qualitative analysis may be sub-optimal for quantitative biology. Second, sample imaging is only a first step, as we need to extract quantitative information. Achieving quantitative imaging remains an open issue in biology, and requires close interactions between biologists, computer scientists and applied mathematicians. On the one hand, experimental and imaging protocols should integrate constraints from the downstream computer-assisted analysis, yielding to a trade-off between qualitative optimized and quantitative optimized protocols. On the other hand, computer analysis should integrate constraints specific to the biological problem, from acquisition to quantitative information extraction. There is therefore a need of specificity for embedding precise biological information for a given task. Besides, a level of generality is also desirable for addressing data from different teams acquired with different protocols and/or sensors. The mathematical modeling of the physics of the acquisition system will yield higher performance reconstruction/restoration algorithms in terms of accuracy. Therefore, physicists and computer scientists have to work together. Quantitative information extraction also has to deal with both the complexity of the structures of interest (e.g., very dense network, small structure detection in a volume, multiscale behavior, $...$) and the unavoidable defects of in vivo imaging (artifacts, missing data, $...$). Incorporating biological expertise in model-based segmentation methods provides the required specificity while robustness gained from a methodological analysis increases the generality. Finally, beyond image processing, we aim at quantifying and then statistically analyzing shapes and complex structures (e.g., neuronal or vascular networks), static or in evolution, taking into account variability. In this context, learning methods will be developed for determining (dis)similarity measures between two samples or for determining directly a classification rule using discriminative models, generative models, or hybrid models. Besides, some metrics for comparing, classifying and characterizing objects under study are necessary. We will construct such metrics for biological structures such as neuronal or vascular networks. Attention will be paid to computational cost and scalability of the developed algorithms: biological experimentations generally yield huge data sets resulting from high throughput screenings. The research of Morpheme will be developed along the following axes:

• Imaging: this includes i) definition of the studied populations (experimental conditions) and preparation of samples, ii) definition of relevant quantitative characteristics and optimized acquisition protocol (staining, imaging, $...$) for the specific biological question, and iii) reconstruction/restoration of native data to improve the image readability and interpretation.
• Feature extraction: this consists in detecting and delineating the biological structures of interest from images. Embedding biological properties in the algorithms and models is a key issue. Two main challenges are the variability, both in shape and scale, of biological structures and the huge size of data sets. Following features along time will allow to address morphogenesis and structure development.
• Classification/Interpretation: considering a database of images containing different populations, we can infer the parameters associated with a given model on each dataset from which the biological structure under study has been extracted. We plan to define classification schemes for characterizing the different populations based either on the model parameters, or on some specific metric between the extracted structures.
• Modeling: two aspects will be considered. This first one consists in modeling biological phenomena such as axon growing or network topology in different contexts. One main advantage of our team is the possibility to use the image information for calibrating and/or validating the biological models. Calibration induces parameter inference as a main challenge. The second aspect consists in using a prior based on biological properties for extracting relevant information from images. Here again, combining biology and computer science expertise is a key point.

5.1 Awards

• V. Stergiopoulou, L. Calatroni, S. Schaub and L. Blanc-Féraud were awarded the best runner-up paper award for the ISBI 2022 proceeding 18.
• V. Stergiopoulou, L. Calatroni and L. Blanc-Féraud were awarded the prix d'excellence UCA 2022 for the best runner-up paper award.
• M. Lazzaretti was awarded with the Verizon Connect Italy grant for young female researchers, Florence, Italy.
• R. Xiao and X. Descombes were awarded the best poster award at REMIA 2022 - MICCAI Workshop on Resource-Efficient Medical Image Analysis.

6.1 New software

7.1 ADMM-Based Residual Whiteness Principle for Automatic Parameter Selection in Single Image Super-Resolution Problems

Participants: Luca Calatroni.

In collaboration with A. Lanza, M. Pragliola, F. Sgallari (University of Bologna).

In 13 we propose an automatic parameter selection strategy for the single image super-resolution problem for images corrupted by blur and additive white Gaussian noise with unknown standard deviation. The proposed approach exploits the structure of both the down-sampling and the blur operators in the frequency domain and computes the optimal regularisation parameter as the one optimising a suitably defined residual whiteness measure. Computationally, the proposed strategy relies on the fast solution of generalised Tikhonov ${\ell }_2$-${\ell }_2$ problems as proposed in 38 and 34. These problems naturally appear as substeps of the Alternating Direction Method of Multipliers used to solve single image super-resolution problems with non-quadratic, non-smooth, sparsity-promoting regularisers both in convex and in non-convex regimes. After detailing the theoretical properties allowing to express the whiteness functional in a compact way, we report an exhaustive list of numerical experiments proving the effectiveness of the proposed approach for different type of problems, in comparison with well-known parameter selection strategies such as, e.g., the discrepancy principle (see figure 1).

Test image (a), observed image (b), reconstruction via bicubic interpolation (c) and reconstructions obtained by applying the residual whiteness principle for automatic parameter selection to Tikhonov-ℓ2\ell _2 reconstruction model (d), and the WTV-ℓ2\ell _2 (e) variational models.

7.2 A Unified Surface Geometric Framework for Feature-Aware Denoising, Hole Filling and Context-Aware Completion

Participants: Luca Calatroni.

In collaboration with G. Reecupero, M. Huska and S. Morigi (University of Bologna).

Technologies for 3D data acquisition and 3D printing have enormously developed in the past few years, and, consequently, the demand for 3D virtual twins of the original scanned objects has increased. In this context, feature-aware denoising, hole filling and context-aware completion are three essential (but far from trivial) tasks. In 7, they are integrated within a geometric framework and realized through a unified variational model aiming at recovering triangulated surfaces from scanned, damaged and possibly incomplete noisy observations. The underlying non-convex optimization problem incorporates two regularization terms: a discrete approximation of the Willmore energy forcing local sphericity and suited for the recovery of rounded features, and an approximation of the ${\ell }_0$ pseudo-norm penalty favoring sparsity in the normal variation. The proposed numerical method solving the model is parameterization-free, avoids expensive implicit volume-based computations and based on the efficient use of the Alternating Direction Method of Multipliers. Experiments show how the proposed framework can provide a robust and elegant solution suited for accurate restorations even in the presence of severe random noise and large damaged areas (see figure 2).

Applications of the proposed surface geometry framework: incomplete and noisy surface input ℳ0\mathcal {M}_0 (left); denoised surface ℳ*\mathcal {M}^* (center, first row), inpainting mask MVM_V and inpainting result ℳ*\mathcal {M}^* (center, second row), context-aware completion of the curly hair detail without inpainting (center, third row); context-aware completion result ℳ*\mathcal {M}^* with preliminary inpainting (right).

The project has been funded by the project Arch-AI-story at UCA and the CNRS PRIME project Imag'In.

7.3 2D & 3D covariance-based ${\ell }_0$ super-Resolution Microscopy with intensity Estimation and automatic parameter estimation

Participants: Luca Calatroni, Laure Blanc-Féraud, Vasiliki Stergiopoulou, Sébastien Schaub.

This work has been carried out in collaboration with J. H de Goulart (INP Toulouse/IRIT).

Super-resolution light microscopy overcomes the physical barriers due to light diffraction. However, state-of-the-art approaches require specific and often demanding acquisition conditions to achieve adequate levels of both spatial and temporal resolution. Analyzing the stochastic fluctuations of the fluorescent molecules provides a solution to the aforementioned limitations. Based on this idea, we propose COL0RME 15, a method for COvariance-based ${\ell }_0$ super-Resolution Microscopy with intensity Estimation, which achieves good spatio-temporal resolution by solving a sparse optimization problem in the covariance domain, discuss automatic parameter selection strategies, estimates background and noise statistics and includes a final estimation step where intensity information is retrieved. Several numerical results both on synthetic and real fluorescent microscopy images and several comparisons with state-of-the art approaches have been done. Our results show that COL0RME outperforms competing methods exploiting analogously temporal fluctuations; in particular, it achieves better localization, reduces background artifacts and avoids fine parameter tuning. In 18 a 3D COL0RME strategy is used to improve spatial resolution in Total Internal Reflectance Fluorescence (TIRF) imaging applications. The approach improves both lateral and axial resolution by combining sparsity-based modelling for precise molecule localisation and intensity estimation in the lateral plane with a 3D reconstruction procedure in the axial one using Multi-Angle TIRF (MA-TIRF). The proposed method has been validated on simulated MA-TIRF blinking-type data and on challenging real MA-TIRF acquisitions, showing significant resolution improvements (see figure 3). The paper was awarded the best runner-up paper award at the ISBI 2022 conference

Vasiliki Stergiopoulou's work is supported by the French Government through the 3IA Côte d’Azur Investments in the Future project managed by the National Research Agency (ANR) with the reference number ANR-19-P3IA-0002.

7.4 Fluorescence image deconvolution microscopy via generative adversarial learning (FluoGAN)

Participants: Vasiliki Stergiopoulou, Laure Blanc-Féraud, Luca Calatroni, Sébastien Schaub.

In 26 we propose FluoGAN, an unsupervised hybrid approach combining the physical modelling of fluorescence microscopy timelapse acquisitions with a Generative Adversarial Network (GAN) for the problem of image deconvolution. Differently from standard approaches combining a least-square data fitting term based on one (longtime exposure) image with sparsity-promoting regularisation terms, FluoGAN relies on a data fitting term defined as a distribution distance between the fluctuating observed timelapse (short-time exposure images) and the generative model. The distance between these two distributions is computed using adversarial training of two competing architectures: a physics-inspired generator simulating the fluctuation behaviour as a Poisson process of the observed images combined with blur and undersampling, and a standard convolutional discriminator. FluoGAN is a fully unsupervised approach requiring only a fluctuating sequence of blurred, undersampled and noisy images of the sample of interest as input and it can be complemented with prior knowledge on the desired solution such as sparsity, non-negativity etc. After having described in depth the main ideas behind FluoGAN, we formulate the corresponding optimisation problem and report several results on a simulated and real phantoms used by microscopy engineers to quantitatively assess spatial resolution. The comparison of FluoGAN with state-of-the-art methodologies shows unprecedented resolution and allows for high-precision reconstruction of very fine structures in challenging real Ostreopsis cf Ovata data (see figure 4).

7.5 Off-the-grid dynamic super-resolution for fluorescence microscopy

Participants: Bastien Laville, Laure Blanc-Féraud, Gilles Aubert.

The last few years have seen the development in super-resolution of variational optimization in measurement spaces. These so-called "gridless" approaches offer both theoretical (uniqueness, reconstruction guarantees) and numerical results, with very convincing results in biomedical imaging. However, gridless variational optimization is formulated for the reconstruction of point sources, which is not always suitable in biomedical imaging applications: one would also need to reconstruct more realistic biological structures such as curves, to represent blood vessels or filaments.

We developed in 2022 a new approach allowing gridless reconstruction of curves, understood as the reconstruction of $m$ measurements carried by curves from a degraded $y$ acquisition. By introducing a new space of $𝒱$ measures with finite divergence, as well as a new CROC functional, we give a result on the extremal points of the unit ball of the $𝒱$ norm; this allows us to reconstruct with a variational gridless approach curves, in an inverse problem setting:

This also opens up promising prospects for the digital implementation, shown on the Figure 5, that we began in 2022 and continue in 2023.

This work is published in 16, 22 and the report 28 appear in SIIMS journal, Janvier 2023.

A code is available at 6.1.5

7.6 Nearest-Neighbor Kernel for the Classification of Plankton Images

Participants: Cédric Dubois, Eric Debreuve.

This work was carried out in collaboration with Jean-Olivier Irisson (LOV, Villefranche-sur-mer, France).

Plankton organisms are a key component of the Earth bio-sphere. Most notably, the phytoplankton captures carbon from the atmosphere, and the zooplankton aggregates it and stores it on the ocean floor. This so-called carbon pump process is studied by ecologists in order to predict its future efficiency in a climate change era. A modern approach consists in studying the relation between the living environment of the organisms and their functional traits, among which the size stands out. Indeed, a high correlation has been observed between zooplankton size and carbon sequestration. Scientists have developed in situ imaging instruments and built large databases. Taxonomic identification from images is required to estimate the individual volumes of the organisms based on their morphology. The development of automated classification methods has been essential to aid operators in their identification task. Specifically, the breakthrough of Artificial Neural Network (ANN) made it possible to come up with efficient classifiers. However, the decisions of such classifiers are hard to interpret or explain. We put forward the idea that following the transform-then-classify-simply approach of ANNs using a simple, explicit transform can result in a classifier whose predictions are both interpretable (thus, trustable) and accurate. The proposed transform is defined as a linear combination of optimal, per-class targets, and the classification is performed, like with ANNs, by a nearest-target decision (see Fig. 6). Furthermore, as a main theoretical result, we establish that the proposed transform defines a kernel associated with the Weigthed-k-Nearest-Neighbor (W-kNN) classifier, and allows interpreting the W-kNN classifier as a member of a larger family of target-based classifiers which satisfies an optimality criterion. We propose a modern W-kNN implementation that allows to deal with large scale databases. We were then able to perform a Leave-One-Out Cross-Validation on a large in situ plankton images dataset from the Underwater Vision Profiler (UVP).

7.7 Dynamic analysis framework to detect cell division and cell death in live-cell imaging, using signal processing and machine learning

Participants: Asma Chalabi, Eric Debreuve.

This work was made in collaboration with Jérémie Roux, Ircan, Nice, France.

The detection of cell division and cell death events in live-cell assays has the potential to produce robust metrics of drug pharmacodynamics and return a more comprehensive understanding of tumor cells responses to cancer therapeutic combinations. As cancer drugs may have complex and mixed effects on the biology of the cell, knowing precisely when cellular events occur in a live-cell experiment allows to study the relative contribution of different drug effects (such as cytotoxic or cytostatic) on a cell population. Yet, classical methods require dyes to measure cell viability as an end-point assay, where the proliferation rates can only be estimated when both viable and dead cells are labeled simultaneously, not to mention that the actual cell division events are often discarded due to analytical limitations. Live-cell imaging is a promising cell-based assay to determine drug efficacies. However, its main limitation remains the accuracy and depth of the analyses, to acquire automatic measures of the cellular response phenotype, making the understanding of drug action on cell populations difficult. We propose a new algorithmic architecture integrating machine learning, image and signal processing methods to perform dynamic image analyses of single cell events in time-lapse microscopy experiments of drug pharmacological profiling. Our event detection method is based on a pattern detection approach on the polarized light entropy, making it free of any labeling step and exhibiting two distinct patterns for cell division and death events. Our analysis framework is an open source and adaptable workflow that automatically predicts cellular events (and their times) from each single cell trajectory, along with other classic cellular features of cell image analyses, as a promising solution in pharmacodynamics. (see Figure 7).

7.8 Developing methodologies to quantitatively analyze RNP granule dynamic properties upon aging

Participants: Assia Chahidi, Eric Debreuve, Xavier Descombes, Florence Besse.

This work was made in collaboration with Fabienne de Graeve, iBV, Nice, France.

Recent technological advances in bioimaging have enabled biologists to investigate dynamic processes in living cells at high spatial and temporal resolutions. Quantitative analysis of these dynamic processes relies on the development of appropriate tracking methods associated with a spatio-temporal analysis. We developed particle tracking and event detection methods to analyze the movements and interactions of Imp-RNP granules in the cell bodies of mushroom neurons in Drosophila brain explants. The individual granule trajectories were considered globally as a graph embedded in a space-time domain. The base granule fusion and splitting events can occur in specific sequences which build particular graph patterns. The proposed analysis consisted in defining the relevant patterns, and then searching them in actual graphs. (see Figure 8).

7.9 Histopathological image classification from nuclei

Participants: Yueqiao Huang, Xavier Descombes.

This work has been carried out in collaboration with Dr Damien Ambrosetti from Nice CHU.

The aim of this project is to segment histopathological images into tumor, non tumor and fibrous areas based on nuclei information. The first step, consisting of nuclei detection, is performed by applying successively color deconvolution, local normalization, Otsu thresholding and spatial smoothing using mathematical morphology. We then extract several parameters describing the nuclei shape and the grey level statistics within the nuclei and in its neighborhood. A statistical analysis of the parameters histograms provides a features selection. The most discriminating parameters are then classified using the unsupervised k-means algorithm. The last step consists in regularizing this first classification. We consider as graph description of the nuclei population provided by the Delaunay tessellation. The nuclei label are finally regularized using a set of rules derived from the majority vote principle.

First results have shown that our approach provide a description of the different tissues based on their nuclei population. On figure 9 can notably distinguish the tumor area on left from a fibrous area. Currently the nuclei classes are defined based on an unsupervised classification. We currently investigate a supervised classification based on a predefined dictionary of cells.

7.10 Simultaneous vascular network segmentation and classification using a multi-tasks semi-supervised neural network

Participants: Rudan Xiao, Xavier Descombes.

This work was carried out in collaboration with Damien Ambrosetti, Nice CHU. It was supported by a PACA PhD grant

Vascular network analysis is crucial to define the tumoral architecture and then diagnose the Renal Cell Carcinoma (RCC) subtypes. However, automatic vascular network segmentation from histopathological images is still a challenge due to the background complexity. Last year, we have proposed a method that reduces reliance on labeled data through semi-supervised learning (SSL). Additionally, considering the correlation between tumor classification and vascular segmentation, we have proposed a multi-task learning (MTL) model, which can simultaneously segment the vascular network using SSL and predict the tumor class in a supervised context. To train the network we have considered weakly labeled data consisting of constant width bands vessels. We used this annotation approach to speed up the process. Besides, we aim to classify RCC subtypes based on the skeleton of the vascular network that reflect the network connectivity. Therefore, the vessel width can be ignored. In this situation the Dice loss function, based on the objects area, is not adapted. We have proposed a new loss function robust with respect to width errors. We dilate (with a disk of radius 3), the segmentation result from S to obtain DS and the ground truth GT to obtain DGT. We computed the ratio of miss-detected vessels as:

and the ratio of falsely detected vessels as:

Finally, we defined the following global performance index:

Figure 10 illustrates the errors between the real mask and the manual detection (so called ground truth) and the errors of the obtained segmentation.

Missing part and false part between the generated segmentations, weakly labels, and real vascular masks (on the overlay images in the last column: cyan represents the real mask, magenta represents the ground truth, yellow represents the segmentation by our proposed MTL-SSL and white represents the overlap part).

7.11 Self Supervised Neural Network for kidney cancer classification from histopathological images

Participants: Mohamad Mohamad, Xavier Descombes.

This work was carried out in collaboration with Francesco Ponzio from Politechnico di Torino (Italy) and Damien Ambrosetti from Nice CHU.

Annotation scarcity is a major obstacle for the effective deployment of deep networks in the medical field. Based on that, our goal was to investigate the uprising Self-Supervision methodology to help us leverage models performance even with few annotations. We investigate a new self supervision pretext task with RCC WSI dataset. The pretext task proceeds as patch allocation/localization in an image by exploiting different level of magnification/resolution benefiting from both object-level (cell, nuclei) and tissue-level features. We consider 2 images from 2 different magnification levels, the image of higher resolution (patch) lies inside the one with lower resolution. The task consists in assigning a label to the patch corresponding to one of the 16 possible locations in the low resolution image. To complexify the task we also consider rotated and flipped patches. This pretext task does not require any annotations, its goal is to learn the features extraction part (or encoder) of the network.

We then consider a classification task with few annotated data by learning only the decoder part. The classification at the patient level is obtained by a majority vote among the different patches. We have compared the proposed approached using a ResNet four times smaller than the network used with transfer learning from ImageNet database (see figure 11)

7.12 Automatic cell identification and analysis on in vivo reflectance confocal microscopy images of the human epidermis

Participants: Imane Lboukili, Xavier Descombes.

This work was carried out in collaboration with Georgios Stamatas (Johnson & Johnson Consumer Health R&D) as part of a collaboration contract with Johnson & Johnson.

The study of Reflectance Confocal Microscopy (RCM) images provides information on epidermal structure, a key to skin health and barrier which changes in each epidermal layer, with age and with certain skin conditions. RCM can also reveal dynamic changes to the epidermis in response to stimuli, as it enables repetitive sampling of the skin without damaging the tissue. Studying RCM images requires manual identification of each cell to derive its geometrical and topological characteristics, which is time-consuming and subject to expert interpretation. More insights could be derived from these data if not for the tediousness of their manual segmentation, highlighting the need for an automated cell identification method.

We have previously proposed a method for automated cell detection in RCM images of the epidermis based on the morphological features of the cells 23. This year we have validated this approach on the study of skin maturation and ageing in people aged 0 to 80 years old on the cheek and volar forearm (paper under review, first large-scale study using RCM images  36).

To overcome manual parametrization and reduce computational time we have proposed a new unsupervised approach based on multi-task learning. Our approach consists of a dual-task Cycle Generative Adversarial Networks model. The architecture is composed on two Cycle GAN to respectively learn RCM images noise and/or texture model from the initial image and the epidermis structure from a Gabor filtering (see figure 12). To our knowledge, this is the first time a Cycle GAN has been embedded in a multi-task learning model.

An international patent has been submitted for this work.

Multi-task Cycle Generative Adversarial Networks architecture. The first task maps domain A (RCM images) to the unpaired domain B (synthetic binary segmentation) while the second task learns the structure RCM images of the epidermis by translating domain C (Gabor filtered RCM images) into domain B.

7.13 Machine learning for organoid phenotyping

Participants: Alexandre Martin, Xavier Descombes.

The aim of this work is to built a phenotypic map of organoid. We model the organoid as a 3D shape containing several layers of cells composed of 2D point patterns on the sphere. Our goal is to classify organoids with respect to these three components.

We first focus on the classification of point patterns on the sphere. We have create a synthetic database of 2D point patterns composed of simulations of different models such as Poisson and Mattern processes (see figure 13 top). We have proposed a new neural network architecture to estimate both the model and its parameters value from a set of points on the sphere, interpreted as a graph using the Delaunay tesselation. This deep learning approach has been compared with the estimation obtained from statistical tests.

Beside, we have proposed a pipeline to analyze 3D confocal images as shown on figure 13 (middle and bottom lines). Nuclei are detected using a template matching approach. We estimate the organoid hull by computing an alpha shape from the set of detected nuclei centers. We then classify the nuclei with respect to their distance from the organoid hull. The obtained classes correspond to the different layers. Finally we approximate the organoid shape by a sphere using a matching to the first layers centers and project nuclei centers on this sphere to obtained the 2D point patterns. This process provides the three components of our organoid model that are a 3D shape, a number of cell layers and 2D point patterns on the sphere corresponding to the different layers.

7.14 Geometrical and statistical analysis of the extra-cellular matrix

Participants: Faisal Jayousi, Esther Jammes, Xavier Descombes, Laure Blanc-Féraud.

This work was carried out in collaboration with Ellen Van Obberghen-Schilling and Morgane Rogissard, iBV, Nice, France.

We studied in this work the protein named fibronectine that is a main component of the extra-cellular matrix. We have imaged a cohort of several hundred tissue samples on a fluorescence multiplex scanner. We then have defined six different classes to characterize the fibronectin patterns as for example aggregates, intracellular, aligned, net shaped. Our goal is to recognize these patterns on images to obtained a segmentation map. We have considered the PanCK channel to segment the tumor, the DAPI channel to detect nuclei. Adding these information to the fibronectine channel we have derived a decision tree to classify the different fibronectine areas. To translate this decision tree into an algorithm, the main challenge was to differentiate aligned and net-shaped structures. To do so, we considered a graph representation of the fibronectine network (see figure 14). The graph has been extracting by first filtering the image with modified Gabor functions, binarizing and skeletonizing the filtered image.

We are currently finalizing the approach and expect to differentiate the classes from the topological information of the graphs (for example from the ratios of the nodes order). Finally, we expect to link these classes to the patient response to immunotherapy.

7.15 Ascidian embryo registration

Participants: Grégoire Malandain, Kevin Yep.

This work has been carried out in collaboration with Killian Biasuz and Patrick Lemaire from CRBM, Montpellier, within the ANR Cell Whisper (section 9.3.6).

It is admitted that ascidian exhibited a stereotyped development in the first stages of their embryogenesis 35. As a consequence, ascidian embryo cells can be unambiguously canonically named after their fate: this provides a common reference space into which population of individuals can be projected and compared. Moreover, this stereotypy also provides a means to name a new sequence from already named ones.

In 5 (see also Astec 6.1.1), we proposed a pipeline to segment and track the embryo cells over several generations from a temporal sequence of light-sheet microscopy 3D images. This pipeline allows to extract a number of quantitative data (lineage, cell volume, cell barycenter, cell contact surfaces, etc.) from a 3D+t image series depicting a developing embryo.

We demonstrated (data not yet published) that the development stereotypy induces cell neighborhood invariance. This allows to entirely name, from a set of atlases (ie already named series) the cells of a developing embryo, according that one time point has been named.

In 37, we have proposed a registration method between segmented embryo 3D images, that offers a first means to transfer names from an atlas to an unnamed embryo, thus to name a time point of an unnamed series. We have investigated whether this approach can be adapted to register two embryos (having the same number of cells) knowing only the quantitative data.

7.16 Segmentation of prostate organoids

Participants: Grégoire Malandain, Apolline Blachet.

This work is made in collaboration with S. Clavel and L. Bastistic (C3M, Nice) and M. Rauzi and B. Delorme (IBV, Nice), within the ANR Morpheus

The ANR Morpheus project aims at using organoids to test the toxicity of endocrine disruptors (EDCs). Within this project, we aimed at studying and quantifying the organoid growth in both normal and perturbed conditions. This requires imaging 3D+t series of developing organoids, and to extract quantitative information at the cell levels.

We have first adapted and test several imaging protocols, with the IBV's light sheet microscope. To assess imaging protocols, resulting 3D+t series were segmented using the Astec pipeline (see section 6.1.1) and segmentation results were visually evaluated (see figure 15). In addition, some Astec parameters were also assessed.

8.1 Bilateral contracts with industry

Participants: Imane Lboukili, Xavier Descombes.

• Johnson & Johnson, Santé Beauté France, Paris: 36-month companion contract (Jan 2021 – Dec 2023) for the PhD thesis of Imane Lboukili.

9.1 International initiatives

9.2 European initiatives

9.3 National initiatives

9.4 Regional initiatives

10.1 Promoting scientific activities

10.2 Teaching - Supervision - Juries

10.3 Popularization

11.1 Major publications

• 1 articleA.Arne Bechensteen, L.Laure Blanc-Féraud and G.Gilles Aubert. A continuous relaxation of the constrained $L_2-L_0$ problem.Journal of Mathematical Imaging and VisionJanuary 2021
  HAL
• 2 articleF.Fabienne De Graeve, E.Eric Debreuve, S.Somia Rahmoun, S.Szilvia Ecsedi, A.Alia Bahri, A.Arnaud Hubstenberger, X.Xavier Descombes and F.Florence Besse. Detecting and quantifying stress granules in tissues of multicellular organisms with the Obj.MPP analysis tool.TrafficJuly 2019
  HALDOI
• 3 articleX.Xavier Descombes. Multiple objects detection in biological images using a marked point process framework.Methods2016
  HALDOI
• 4 articleG.Georgios Efthymiou, A.Agata Radwanska, A.-I.Anca-Ioana Grapa, S.Stéphanie Beghelli-de la Forest Divonne, D.Dominique Grall, S.Sébastien Schaub, M.Maurice Hattab, S.Sabrina Pisano, M.M. Poët, D. F.Didier F Pisani, L.Laurent COUNILLON, X.Xavier Descombes, L.Laure Blanc-Féraud and E.Ellen Van Obberghen-Schilling. Fibronectin Extra Domains tune cellular responses and confer topographically distinct features to fibril networks.Journal of Cell ScienceFebruary 2021
  HAL
• 5 articleL.Léo Guignard, U.-M.Ulla-Maj Fiuza, B.Bruno Leggio, J.Julien Laussu, E.Emmanuel Faure, G.Gaël Michelin, K.Kilian Biasuz, L.Lars Hufnagel, G.Grégoire Malandain, C.Christophe Godin and P.Patrick Lemaire. Contact area-dependent cell communication and the morphological invariance of ascidian embryogenesis.ScienceJuly 2020
  HALDOIback to text
• 6 articleV.Vasiliki Stergiopoulou, L.Luca Calatroni, J. H.José Henrique de M Goulart, S.Sébastien Schaub and L.Laure Blanc-Féraud. COL0RME: Super-resolution microscopy based on sparse blinking/fluctuating fluorophore localization and intensity estimation.Biological Imaging2February 2022, e1
  HALDOI

11.2 Publications of the year

11.3 Other

11.4 Cited publications

• 34 inproceedingsL.Luca Calatroni, M.Monica Pragliola, A.Alessandro Lanza and F.Fiorella Sgallari. Residual Whiteness Principle for Automatic Parameter Selection in $$2 - $$2 Image Super-Resolution Problems.SSVM 2021 - 8th International Conference on Scale Space and Variational Methods in Computer Vision12679Lecture Notes in Computer ScienceVirtual Event, FranceSpringer International PublishingMay 2021, 476-488
  HALDOIback to text
• 35 bookE. G.Edwin Grant Conklin. The organization and cell-lineage of the ascidian egg / by Edwin G. Conklin..Philadelphia :[Academy of Natural Sciences],1905, 182URL: http://www.biodiversitylibrary.org/item/24005
  back to text
• 36 miscI.Imane Lboukili, X.Xavier Descombes and G.Georgios Stamatas. Age-dependent changes in epidermal architecture explored using an automated image analysis algorithm on in vivo reflectance confocal microscopy images .
  back to text
• 37 inproceedingsG.Gaël Michelin, L.Léo Guignard, U.-M.Ulla-Maj Fiuza, P.Patrick Lemaire, C.Christophe Godin and G.Grégoire Malandain. Cell Pairings for Ascidian Embryo Registration.ISBI - International Symposium on Biomedical ImagingThe 2015 IEEE International Symposium on Biomedical Imaging proceedingsInstitute of Electrical and Electronics Engineers (IEEE). USA.New York, United StatesThe Printing House, IncApril 2015
  HALDOIback to text
• 38 articleN.Ningning Zhao, Q.Qi Wei, A.Adrian Basarab, N.Nicolas Dobigeon, D.Denis Kouamé and J.-Y.Jean-Yves Tourneret. Fast Single Image Super-Resolution Using a New Analytical Solution for ${}_2$-${}_2$ Problems.IEEE Transactions on Image Processingvol. 25n°August 2016, pp. 3683-3697
  HALDOIback to text