Massimiliano Passaretti (left) and Yu-Wei Chang (right) at NEME 2025. (Photo courtesy of Clarion Hotel Draken.)Graph theory and deep learning pipelines
Yu-Wei Chang, Massimiliano Passaretti NEMES 2025, 24-26 September, 2025 Date: 25 September 2025 Time: 12:45 – 14:00 Place: Clarion Hotel Draken
This workshop begins with a practical introduction to graph theory, then guides participants through BRAPH 2 to build connectomes, compute graph measures, and run group comparisons, followed by a hands-on deep-learning pipeline. It demonstrates a unified GUI/command-line workflow, a unique architecture of BRAPH 2, helping participants move smoothly from the GUI to scripts. This workshop also guides participants to reproduce multiplex and deep-learning results on their computers from the BRAPH 2 bioRxiv preprint.
From images to graphs, this plenary shows how parcellations and tractography become connectomes and how network analysis reveals brain-network signatures. (Image by Y.-W. Chang.)Network analysis of neuroimaging data, and deep learning pipelines
Yu-Wei Chang NEMES 2025, 24-26 September, 2025 Date: 25 September 2025 Time: 09:00 – 09:45 Place: Clarion Hotel Draken
This plenary presents a practical framework for analysing neuroimaging data with network science and deep learning. It moves from modality-specific preprocessing to graph construction (single-layer and multiplex), then covers core graph measures, group inference, and brain-surface visualization, highlighting recent work from Associate Professor Joana B. Pereira’s group (Department of Clinical Neuroscience, Karolinska Institutet). It also introduces deep-learning pipelines for neuroimaging data: reservoir-computing memory capacity analysis, GapNet for handling missing data, and a robust feature-attribution method combined with SNP (single nucleotide polymorphism) information. The plenary concludes with the BRAPH 2 framework, which supports these pipelines and extends to other ongoing projects (e.g., light-sheet microscopy, Raman spectroscopy).
Alex Lech at the BNMI poster session. (Photo by M. Granfors)DeepTrack2: Microscopy Simulations for Deep Learning
Alex Lech, Mirja Granfors, Benjamin Midtvedt, Jesús Pineda, Harshith Bachimanchi, Carlo Manzo, Giovanni Volpe BNMI 2025, 19-22 August 2025, Gothenburg, Sweden Date: 20 August 2025 Time: 15:15-19:00 Place: Wallenberg Conference Centre
DeepTrack2 is a flexible and scalable Python library designed for simulating microscopy data to generate high-quality synthetic datasets for training deep learning models. It supports a wide range of imaging modalities, including brightfield, fluorescence, darkfield, and holography, allowing users to simulate realistic experimental conditions with ease. Its modular architecture enables users to customize experimental setups, simulate a variety of objects, and incorporate optical aberrations, realistic experimental noise, and other user-defined effects, making it suitable for various research applications. DeepTrack2 is designed to be an accessible tool for researchers in fields that utilize image analysis and deep learning, as it removes the need for labor-intensive manual annotation through simulations. This helps accelerate the development of AI-driven methods for experiments by providing large volumes of data that is often required by deep learning models. DeepTrack2 has already been used for a number of applications in cell tracking, classifications tasks, segmentations and holographic reconstruction. Its flexible and scalable nature enables researchers to simulate a wide array of experimental conditions and scenarios with full control of features and parameters.
DeepTrack2 is available on GitHub, with extensive documentation, tutorials, and an active community for support and collaboration at https://github.com/DeepTrackAI/DeepTrack2.
Quantitative Digital Microscopy with Deep Learning.
Benjamin Midtvedt, Saga Helgadottir, Aykut Argun, Jesús Pineda, Daniel Midtvedt & Giovanni Volpe.
Applied Physics Reviews, volume 8, article number 011310 (2021).
DeepTrack2 Logo. (Image by J. Pineda)DeepTrack2: physics-based microscopy simulations for deep learning
Mirja Granfors, Alex Lech, Benjamin Midtvedt, Jesús Pineda, Harshith Bachimanchi, Carlo Manzo, and Giovanni Volpe BNMI 2025, 19-22 August 2025, Gothenburg, Sweden Date: 20 August 2025 Time: 15:00 – 15:15 Place: Wallenberg Conference Centre
DeepTrack2 is a flexible and scalable Python library designed to generate physics-based synthetic microscopy datasets for training deep learning models. It supports a wide range of imaging modalities, including brightfield, fluorescence, darkfield, and holography, enabling the creation of synthetic samples that accurately replicate real experimental conditions. Its modular architecture empowers users to customize optical systems, incorporate optical aberrations and noise, simulate diverse objects across various imaging scenarios, and apply image augmentations. DeepTrack2 is accompanied by a dedicated GitHub page, providing extensive documentation, examples, and an active community for support and collaboration: https://github.com/DeepTrackAI/DeepTrack2.
GAUDI leverages a hierarchical graph-convolutional variational autoencoder architecture, where an encoder progressively compresses the graph into a low-dimensional latent space, and a decoder reconstructs the graph from the latent embedding. (Image by M. Granfors and J. Pineda.)Global graph features unveiled by unsupervised geometric deep learning
Mirja Granfors, Jesús Pineda, Blanca Zufiria Gerbolés, Joana Pereira, Carlo Manzo, and Giovanni Volpe Date: 7 August 2025 Time: 2:45 PM – 3:00 PM Place: Conv. Ctr. Room 4
Graphs are used to model complex relationships, such as interactions between particles or connections between brain regions. The structural complexity and variability of graphs pose challenges to their efficient analysis and classification. Here, we propose GAUDI (Graph Autoencoder Uncovering Descriptive Information), a graph autoencoder that addresses these challenges. GAUDI is trained in an unsupervised manner to capture the most critical parameters of graphs in the latent space, thereby enabling the extraction of essential parameters characterizing the graphs. We demonstrate the performance of GAUDI across diverse graph data originating from complex systems, including the estimation of the parameters of Watts-Strogatz graphs, the classification of protein assembly structures from single-molecule localization microscopy data, the analysis of collective behaviors, and correlations between brain connections and age. This approach offers a robust framework for efficiently analyzing and interpreting complex graph data, facilitating the extraction of meaningful patterns and insights across a wide range of applications.
Memory capacity in aging. A Brain reservoir computing architecture with uniform random signals applied to all nodes. (Image from the article.)Computational memory capacity predicts aging and cognitive decline Blanca Zufiria-Gerbolés, Mite Mijalkov, Ludvig Storm, Dániel Veréb, Zhilei Xu, Anna Canal-Garcia, Jiawei Sun, Yu-Wei Chang, Hang Zhao, Emiliano Gómez-Ruiz, Massimiliano Passaretti, Sara Garcia-Ptacek, Miia Kivipelto, Per Svenningsson, Henrik Zetterberg, Heidi Jacobs, Kathy Lüdge, Daniel Brunner, Bernhard Mehlig, Giovanni Volpe, Joana B. Pereira
SPIE-ETAI, San Diego, CA, USA, 3 – 7 August 2025 Date: 7 August 2025 Time: 11:15 AM – 11:45 AM PDT Place: Conv. Ctr. Room 4
Using reservoir computing and diffusion-weighted imaging, we explored changes in brain connectivity patterns and their impact on cognition during aging. We found that whole-brain networks perform optimally at low densities, with performance decreasing as network density increases, particularly in regions with weaker connections. This decline was strongly associated with age and cognitive performance. Our results suggest that a core network of anatomical hubs is essential for optimal brain function, while peripheral connections are more vulnerable to aging. This study highlights the potential of reservoir computing for understanding age-related cognitive decline.
Reference
Mite Mijalkov, Ludvig Storm, Blanca Zufiria-Gerbolés, Dániel Veréb, Zhilei Xu, Anna Canal-Garcia, Jiawei Sun, Yu-Wei Chang, Hang Zhao, Emiliano Gómez-Ruiz, Massimiliano Passaretti, Sara Garcia-Ptacek, Miia Kivipelto, Per Svenningsson, Henrik Zetterberg, Heidi Jacobs, Kathy Lüdge, Daniel Brunner, Bernhard Mehlig, Giovanni Volpe, Joana B. Pereira, Computational memory capacity predicts aging and cognitive decline
Nature Communications 16, 2748 (2025)
Illustration of adsorption process of a polymer coated particle. A single particle is brought to a liquid-liquid interface using an optical tweezers and once the polymer shell makes contact with the interface the particle immediately jumps into the interface. (Image by M. Selin.) Mapping the adsorption dynamics of core-shell particles at liquid-liquid interfaces with optical tweezers
Martin Selin, Maret Ickler, Gerardo Campos-Villalobos, Fabrizio Camerin, Nicolas Vogel, Antonio Ciarlo, Giovanni Volpe, and Marcel Rey Date: 6 August 2025 Time: 4:30 PM – 4:45 PM PDT Place: Conv. Ctr. Room 3
Colloidal systems are integral to industries such as food and cosmetics, where liquid-liquid interfaces—like oils dispersed in water—are common. Whether colloidal particles adsorb to these interfaces depends on multiple factors such as particle surface chemistry, pH and salinity.
Here, we investigate how core–shell particles breach a liquid-liquid interface by using optical tweezers to gently push the particles into dodecane-water interfaces formed by microbubbles. Our core–shell particles feature a silica core and a PDMAEMA shell and by varying the amount of monomer added during synthesis the size of the shell can be tuned. Using the tweezers we measure the extent of the polymer shell. Importantly, we find that uncoated silica particles do not adsorb in pure water, whereas polymer coated particles absorb rapidly once the polymer layer contacts the interface, also when the core itself remains microns away. The longer the polymer the greater the distance from which the particle absorbs.
We also observe similar adsorption other polymer shells like PNIPAM and PVP, indicating that the presence of a polymer coating, rather than its specific chemical composition, is the key factor governing adsorption. At low and high pH the polymer shell contracts, also the binding energy becomes weaker making the absorption slower. In very acidic conditions the binding is so weak that the optical tweezers can pull particles out from the interface, allowing us to directly observe individual polymers detaching. These findings provide new insight into how polymer coatings dictate particle-interface interactions, paving the way for improved control of colloidal behavior.
Automated segnmentation of bacterial structures within a droplet. The image shows a bright-field microscopy view where a large biofilm region (green, outlined in blue) has been segmented from surrounding features. Small aggregates (yellow contours) are also highlighted. This segmentation enables structural differentiation of biofilm components for downstream quantitative analysis. (Image by D. Pérez Guerrero.)Quantitative analysis of dynamic biofilm structures via time-resolved droplet microfluidics and artificial intelligence
Daniela Pérez Guerrero, Jesús Manuel Antúnez Domínguez, Aurélie Vigne, Daniel Midtvedt, Wylie Ahmed, Lisa Muiznieks, Giovanni Volpe and Caroline Beck Adiels.
SPIE-ETAI, San Diego, CA, USA, 03 – 07 August 2025 Date: 6 August 2025 Time: 2:30 PM – 2:45 PM PDT Place: Conv. Ctr. Room 4
Biofilms are structured communities of microorganisms that play a crucial role in medicine, biotechnology, and ecology, contributing to microbial adaptation to any environment. Despite their significance, understanding their formation, development, and behavior remains a challenge for the community. We utilize high-throughput droplet microfluidics to enable biofilm growth in miniaturized environments, generating extensive time-lapse bright-field microscopy images. To overcome experimental constraints, including dense structural heterogeneity and skewed illumination, we developed a deep learning-based segmentation approach capable of identifying biofilm structures in complex imaging conditions. Our method operates in an unsupervised manner, reducing the need for ground truth annotations and mitigating the introduced bias of manual segmentation approaches.
Our unsupervised model effectively detects and quantifies biofilm structures, even in late-stage growth, where traditional segmentation techniques fail. The neural network demonstrates robust performance across the development cycle, distinguishing biofilm boundaries and bacteria aggregates separated from the main biofilm structure despite imaging inconsistencies. Additionally, our approach reduces manual intervention, streamlining the analysis of high-throughput biofilm imaging data.
This AI-powered segmentation technique provides a reliable and scalable tool for biofilm analysis, addressing key limitations of conventional methods. By bridging the gap between microbiology research and automated image analysis, our approach facilitates more efficient and reproducible biofilm studies.
BRAPH 2 Genesis enables swift creation of custom, reproducible software distributions—tackling the growing complexity of neuroscience by streamlining analysis across diverse data types and workflows. (Image by B. Zufiria-Gerbolés and Y.-W. Chang.)BRAPH 2: a flexible, open-source, reproducible, community-oriented, easy-to-use framework for replicable network analysis in neurosciences Yu-Wei Chang, Blanca Zufiria Gerbolés, Joana B Pereira, Giovanni Volpe Date: 6 August 2023 Time: 11:00 AM PDT
As network analyses in neuroscience continue to grow in both complexity and size, flexible methods are urgently needed to provide unbiased, reproducible insights into brain function. BRAPH 2 is a versatile, open-source framework that meets this challenge by offering streamlined workflows for advanced statistical models and deep learning in a community-oriented environment. Through its Genesis compiler, users can build specialized distributions with custom pipelines, ensuring flexibility and scalability across diverse research domains. These powerful capabilities will ensure reproducibility and accelerate discoveries in neuroscience.
