Novel functional and anatomical imaging approaches for the nervous system studies

25.04.2026, Saturday, 15:15-17:00

General focus of the symposium:

The symposium will showcase complementary imaging strategies that together enable a multi-scale view of the nervous system, from cellular activity to whole-brain organisation. The common theme is the integration of advanced optical and magnetic resonance imaging (MRI) methods with robust analysis pipelines that make these techniques reproducible and transferable between laboratories.

The first part will focus on functional imaging using genetically encoded calcium indicators such as GCaMP, combined with high-speed optical microscopy. These approaches allow monitoring of activity in defined neuronal populations with high temporal resolution and can be aligned with behavioural paradigms and disease models to dissect circuit-level mechanisms. In parallel, light-sheet microscopy, including mesoSPIM implementations, will be presented as a powerful tool for three-dimensional anatomical mapping in cleared brain and spinal cord tissue, enabling quantitative analysis of connectivity, vascular architecture and activity markers across large volumes.

The second part will highlight recent developments in MRI-based analysis frameworks for neuroscience. Particular emphasis will be placed on methods for accurate co-registration of MRI with optical imaging readouts and spatially resolved omics data. Such integration is crucial for linking macroscopic measures of structure, perfusion and function to the underlying cellular and molecular landscape.

Finally, the symposium will address automation and standardisation of image analysis across modalities. This includes the design of reusable, containerised workflows, harmonised quality control procedures and FAIR-compliant data management strategies that facilitate data sharing and meta-analyses.

Together, the contributions will provide a coherent overview of how novel functional and anatomical imaging approaches can be combined into unified experimental pipelines, and will be relevant for basic, preclinical and translational researchers aiming to implement state-of-the-art imaging in their studies of the nervous system.

15:15 Diana Cash, Ph.D

BRAIN Centre, King’s College London and L&M DataScience GmbH

"Preclinical fingerprinting of drugs with pharmacological fMRI"

This joint talk will present an integrated view of pharmacological functional MRI (ph-fMRI) and cloud-based analysis platforms as complementary tools for translational neuroscience and CNS drug development. Drawing on experience from the BRAIN Centre’s preclinical imaging programmes, the first part will outline how ph-fMRI can be used to map systems-level drug effects across circuits implicated in neurodegeneration, pain, neuroinflammation, trauma and stroke. The emphasis will be on study-design principles (choice of paradigms, anaesthesia versus awake imaging, dosing strategies and longitudinal readouts), as well as on the derivation of robust imaging biomarkers that support mechanism-of-action studies and back-translation between animal models and human cohorts.

The second part will focus on the standardisation and scaling-up of such imaging approaches using cloud-native infrastructures. Using the neuroPhINDr platform as an example, the talk will describe how MRI and corroborative modalities (such as EEG and optical readouts) can be embedded in harmonised, AWS-hosted workflows that combine automated preprocessing, multimodal registration and integrated quality-control metrics. Project-level dashboards for tracking recruitment, data acceptance and pipeline performance will be highlighted as practical tools for managing complex, multi-centre studies.

Throughout the presentation, particular attention will be paid to reproducibility and FAIR data principles, including the use of containerised analysis environments, standardised data structures and transparent reporting. By linking sophisticated ph-fMRI methodology with rigorously engineered, cloud-based pipelines, the talk will illustrate a realistic route towards more consistent, sharable and clinically relevant neuroimaging in both academic and industry settings.

15:38 Marija Petrinovic, PhD

King’s College London

"Of Mice and Mechanisms: Convergent Aggression, Divergent Biology"

Aggression is a common feature of neurodevelopmental conditions, but
whether it reflects shared or distinct underlying mechanisms remains unclear. We
show that similar increases in aggression can arise from divergent circuit-level
activity. Importantly, these differences are accompanied by selective sensitivity to
pharmacological interventions. Together, these findings demonstrate that convergent
behavioral phenotypes can mask fundamentally distinct biological substrates, with
important implications for targeted therapeutic strategies.

16:00 Julia Niemczycka & Mateusz W. Kucharczyk, Ph.D

Cancer Neurophysiology Group, Polish Centre for Technology Development, Łukasiewicz-PORT, Wrocław, Poland

Biophotonics and Electrophysiology Lab, Imaging Laboratories, Center for the Development of Therapies for Civilization and Age-Related Diseases, Jagiellonian University Medical College, Kraków, Poland

Wolfson Sensory, Pain and Regeneration Centre, King’s College London, UK

"High-Throughput Optical Microscopy Approaches for Functional Imaging and 3D Reconstruction of Neural Tissue"

Dr. Kucharczyk will introduce a practical overview of high-throughput linear and non-linear fluorescence microscopy approaches enabling (A) rapid functional sampling and (B) large-scale anatomical reconstruction across the peripheral and central nervous systems. A) Functional imaging pipelines will be demonstrated across custom-built and commercial wide-field, confocal, confocal endoscopy, and two-photon microscopes, illustrating how genetically encoded calcium indicators permit in vivo monitoring of neuronal activity in both the PNS and CNS. B) High-resolution volumetric imaging strategies will be showcased using confocal, two-photon, and light-sheet modalities, including the first Polish benchtop mesoSPIM platform developed in our laboratory.

Subsequently, PhD student Julia Niemczycka will present original datasets, including whole-organ 3D reconstructions of peripheral innervation paired with molecular profiling. A case study will revisit the specificity of transgenic reporter lines, revealing the anatomy and physiology of deep-tissue peptidergic afferents. Examples will include in vivo sensory-neuron calcium imaging in health and disease using wide-field and two-photon microscopy, as well as mesoscale reconstructions and practical experience gained during the construction of the mesoSPIM light-sheet microscope.

16:15 Karolina Nowalińska & Bartosz Pomierny, Ph.D

BioImaging Laboratory, Center for the Development of Therapies for Civilization and Age-Related Diseases, Jagiellonian University Medical College, Kraków, Poland

"Systemic zolpidem administration reduces KCl-induced CSD frequency in naïve Sprague-Dawley rats"

This talk will outline an MRI-based framework for visualising cortical spreading depolarisation/depression (CSD) in vivo in rat models. It will describe how CSD waves, triggered by topical KCl, electrical stimulation or arising spontaneously during experimental stroke, can be captured using multi-parametric MRI. The emphasis will be on diffusion- and perfusion-weighted imaging, susceptibility-based contrasts and, where applicable, functional MRI readouts to map the spatiotemporal signature of CSD and its haemodynamic correlates. The presentation will also highlight practical aspects of protocol design, co-registration with anatomical and molecular endpoints, and integration with electrophysiological recordings. Together, these elements illustrate how MRI can be used as a non-invasive tool to study CSD-related secondary injury and penumbra dynamics.

16:30 Kornelia Kliś, PhD, MD

Department of Neurosurgery and Neurotraumatology Jagiellonian University Medical College

" The Brain Beyond the Aneurysm: from morphometry and networks to treatment Effects and translational bridges"

16:45 Zofia Sikorska

SANO - Centre for Computational Personalised Medicine

"Optimizing CNNs for Detecting Low Amplitude Eye Movements from fMRI"

The fMRI signal from the eyes provides a proxy for gaze position, enabling retrospective analysis of gaze behavior and brain activity in open datasets. While prior work has successfully decoded gaze during large-scale tasks such as smooth pursuit and free viewing, detecting small-scale eye movements that are common in open-eyes resting-state (RS) paradigms remains challenging. The subtle movements during RS carry clinically relevant information but have much lower amplitudes than large-scale gaze tasks.

In this project, we aim to develop a model to detect low-amplitude eye movements from fMRI data.

We used simultaneous fMRI and eye-tracking data from 29 participants during a resting-state fixation scan. Mean gaze amplitude was ±1°, lower than large-scale tasks (±10°). We evaluated four strategies using a CNN model pretrained on large-scale gaze tasks: (1) directly applying the pretrained model to the RS dataset; (2) augmenting the original training set with RS data; (3) training a new model using only the RS dataset; and (4) transfer learning by freezing the pretrained backbone and fine-tuning on RS data using a composite loss (Smooth L1/Huber, δ=1.0; Pearson weight=0.01; learning rate=1e-4; 40 epochs) to balance robust error minimization and trend capture.

Models pretrained on large-scale tasks systematically overestimated gaze amplitudes of RS data due to the amplitude mismatch. Training from scratch on the small RS dataset failed because of limited sample size and imbalance. The proposed transfer learning approach significantly improved performance, achieving a median Pearson correlation of 0.36±0.25 between predicted and true RS gaze data.

These results demonstrate that decoding small-scale spontaneous eye movements from resting-state fMRI is feasible but requires careful domain adaptation. Correlation-aware transfer learning effectively bridges amplitude differences, unlocking the potential to extract oculomotor information from existing resting-state fMRI datasets