Raffaella Fiamma Cabini

Objective: Radiomics extracts quantitative features from medical images that may provide non-invasive biomarkers of tumor biology and heterogeneity. However, few studies have explored radiomics on micro-computed tomography (micro-CT) or its transferability to clinical CT. This study investigates the reproducibility and the potential biological correlation of radiomic features across in vivo CT, ex vivo CT, and ex vivo micro-CT in non-small-cell lung cancer (NSCLC). Materials and methods: A retrospective cohort of 60 NSCLC patients with preoperative CT scans and surgical specimens was analyzed. Ex vivo CT and micro-CT images were acquired and compared with in vivo CT using standardized preprocessing, including normalization, resampling, and registration. Tumor regions of interest (ROIs) were semi-automatically segmented, and 93 radiomic features were extracted with Pyradiomics. Cross-modality reproducibility and associations with histotype were evaluated using Spearman correlation, intraclass correlation coefficient, Bland–Altman analysis, and Mann–Whitney U test. Results: Normalization and volume matching improved feature repeatability, with the highest agreement observed in normalized datasets using subsampled ROIs. Four texture-based features consistently showed high reproducibility and potential biological correlation, likely reflecting intratumoral heterogeneity and variation in cell density: DependenceNonUniformity, GrayLevelNonUniformity (GLDM and GLRLM), and Coarseness (NGTDM). These features reliably distinguished adenocarcinoma from squamous cell carcinoma across modalities, confirming their stability and potential as candidate features for transferable biomarkers. Conclusion: Harmonization and volume standardization are essential for cross-modality radiomic analyses. Stable texture features provide a potential transferable signature bridging micro-CT and clinical CT imaging in NSCLC, supporting their role as candidate features for further investigation in biologically oriented studies. Key Points: Question Micro-CT serves as an intermediate imaging scale enabling radiomic feature extraction from ex vivo NSCLC tumors, allowing direct comparison with in vivo and ex vivo CT. Findings Intensity normalization and volume-matched subsampling between in vivo CT, ex vivo CT, and micro-CT reduced cross-modality feature variability, improving feature reproducibility. Relevance statement Normalization, volume matching, and cross-modality analysis support the identification of candidate texture features with consistent cross-scale behavior and exploratory associations with NSCLC histotype.

The classification of microscopy videos capturing complex cellular behaviors is crucial for understanding and quantifying the dynamics of biological processes over time. However, it remains a frontier in computer vision, requiring approaches that effectively model the shape and motion of objects without rigid boundaries, extract hierarchical spatiotemporal features from entire image sequences rather than static frames, and account for multiple objects within the field of view. To this end, we organized the Cell Behavior Video Classification Challenge (CBVCC), benchmarking 35 methods based on three approaches: classification of tracking-derived features, end-to-end deep learning architectures to directly learn spatiotemporal features from the entire video sequence without explicit cell tracking, or ensembling tracking-derived with image-derived features. We discuss the results achieved by the participants and compare the potential and limitations of each approach, serving as a basis to foster the development of computer vision methods for studying cellular dynamics.

Objective. Lung cancer remains the leading cause of cancer-related mortality worldwide, with Non-Small Cell Lung Cancer (NSCLC) accounting for approximately 85% of all cases. Programmed cell Death Ligand-1 (PD-L1) is a well-established biomarker that guides immunotherapy in advanced-stage NSCLC, currently evaluated via invasive biopsy procedures. This study aims to develop and validate a non-invasive pipeline for stratifying PD-L1 expression using quantitative analysis of IVIM parameter maps—diffusion (D), pseudo-diffusion (D*), perfusion fraction (pf)—and T1-VIBE MRI acquisitions. Approach. MRI data from 43 NSCLC patients were analysed and labelled as PD-L1 positive (≥1%) or negative (Main Results. Logistic Regression generally demonstrated the highest classification performance, with AUC values ranging from 0.78 to 0.92 across all feature sets. Fusion models outperformed or matched the performance of the best standalone radiomics or deep learning model. Among multisequence MRI, the IVIM-D fusion features yielded the best performance with an AUC of 0.92, followed by IVIM-D* radiomic features that showed a similar AUC of 0.91. For IVIM-pf and T1-VIBE derived features, the fusion model yielded the best AUC values of 0.87 and 0.90, respectively. Significance. The obtained results highlight the potential of a combined radiomic-deep learning approach to effectively detect PD-L1 expression from MRI acquisitions, paving the way for a non-invasive PD-L1 evaluation procedure.

Understanding the spatiotemporal dynamics of immune cells in living organisms is a major goal in bioimaging. Intravital microscopy enables direct observation of cellular behavior over time with tissue-to-subcellular resolution, making it essential for investigating immune responses across tissues, conditions, and disease contexts. However, most intravital microscopy data remain siloed in individual labs, limiting reuse, standardization, and large-scale analysis. To address these limitations, we present Immunemap, an open-data platform and interactive atlas of immune cell motility. Immunemap currently provides access to over 58,000 curated single-cell tracks and more than 1,049,000 cell-centroid annotations from 400 intravital microscopy videos in murine models, spanning diverse tissues and conditions. The platform supports both exploratory and quantitative research. We show here how unsupervised learning identifies distinct motility patterns, and how large-scale mapping enables comparisons across stimuli, imaging setups, and organs. Its cloud-based architecture offers an interactive web interface and public APIs for integration with computational pipelines. By adhering to FAIR principles (Findability, Accessibility, Interoperability, and Reuse) and fostering cross-disciplinary studies, Immunemap supports reproducible research and provides a benchmark for bioimage analysis and tool development in intravital imaging.

A fundamental characteristic of living organisms is their capacity to change over time. These changes manifest across various scales: from molecular dynamics to modifications in the shape or arrangement of cells and, ultimately, changes at the level of tissues and organs. Imaging techniques to record biomedical videos have become integral to studying these processes in motion, but are notoriously difficult to analyze. While exciting advances in computer vision technology hold the promise to solve open challenges in quantifying these videos, their application to the field of biomedical imaging is far from trivial. This book chapter summarizes the state-of-the-art in analyzing motion from biomedical videos, and highlights opportunities and challenges for the translation of computer vision techniques to this field.

Live-cell imaging techniques enable the acquisition of videos capturing complex cellular morphodynamics over time. However, the classification of these videos presents unmet technological needs, including methods to effectively analyze the shape and motion of objects without a rigid boundary, derive hierarchical spatiotemporal features from entire video sequences rather than static images, and potentially account for multiple other objects in the field of view. To this end, we organized the Cell Behavior Video Classification Challenge (CBVCC), benchmarking 35 different computer vision methods that can be grouped into three approaches: based on classification of tracking-derived features, based on end-to-end deep learning architectures to directly learn spatiotemporal features from the entire video sequence without explicit cell tracking, or ensembling tracking-derived with image-derived features. Here we compare the methods and discuss the potential and limitations of each approach, serving as a basis to foster the development of computer vision methods for studying cellular dynamics.

In this article, we extend a recently introduced kinetic model for consensus-based segmentation of images. In particular, we will interpret the set of pixels of a 2D image as an interacting particle system that evolves in time in view of a consensus-type process obtained by interactions between pixels and external noise. Thanks to a kinetic formulation of the introduced model, we derive the large time solution of the model. We will show that the parameters defining the segmentation task can be chosen from a plurality of loss functions that characterize the evaluation metrics.

Background: Body composition scores allow for quantifying the volume and physical properties of specific tissues. However, their manual calculation is time-consuming and prone to human error. This study aims to develop and validate CompositIA, an automated, open-source pipeline for quantifying body composition scores from thoraco-abdominal computed tomography (CT) scans. Methods: A retrospective dataset of 205 contrast-enhanced thoraco-abdominal CT examinations was used for training, while 54 scans from a publicly available dataset were used for independent testing. Two radiology residents performed manual segmentation, identifying the centers of the L1 and L3 vertebrae and segmenting the corresponding axial slices. MultiResUNet was used to identify CT slices intersecting the L1 and L3 vertebrae, and its performance was evaluated using the mean absolute error (MAE). Two U-nets were used to segment the axial slices, with performance evaluated through the volumetric Dice similarity coefficient (vDSC). CompositIA’s performance in quantifying body composition indices was assessed using mean percentage relative error (PRE), regression, and Bland–Altman analyses. Results: On the independent dataset, CompositIA achieved a MAE of about 5 mm in detecting slices intersecting the L1 and L3 vertebrae, with a MAE < 10 mm in at least 85% of cases and a vDSC greater than 0.85 in segmenting axial slices. Regression and Bland–Altman analyses demonstrated a strong linear relationship and good agreement between automated and manual scores (p values < 0.001 for all indices), with mean PREs ranging from 5.13% to 15.18%. Conclusion: CompositIA facilitated the automated quantification of body composition scores, achieving high precision in independent testing. Relevance statement: CompositIA is an automated, open-source pipeline for quantifying body composition indices from CT scans, simplifying clinical assessments, and expanding their applicability. Key Points: Manual body composition assessment from CTs is time-consuming and prone to errors. CompositIA was trained on 205 CT scans and tested on 54 scans. CompositIA demonstrated mean percentage relative errors under 15% compared to manual indices. CompositIA simplifies body composition assessment through an artificial intelligence-driven and open-source pipeline. Graphical

Fungiform papillae (FPs) are fundamental for taste perception, as they contain the taste sensory cells responsible for detecting taste stimuli. Variations in the number and functionality of FPs among individuals lead to differences in taste perception, impacting the ability to identify nutrient-rich foods, health, and the joy of consuming tasty foods. Detecting FPs is a complex and time-consuming task, and there is no consensus on manual and automated methods for their identification and analysis. Objectives: This work aimed to provide an efficient, reliable, and automatic method for FP identification on the tongue, considering the physiological variations in morphology and distribution among subjects. Methods: We used three different Convolutional Neural Networks as a regression task on 175 images of the tongue, the Classic U-Net, the MultiResUNet, and the Optimized U-Net, designed to enhance the performance also when it must identify FPs in challenging input images. Results: The Optimized U-Net showed the best performance by achieving the lowest errors and the highest similarity between Ground Truths and prediction values, and the more balanced detection of True Positives, Untrue Negatives, and Untrue Positives. Conclusions: Our results show that the Optimized U-Net achieved the highest stability, accuracy, and robustness in learning and prediction of FPs with challenging morphologies. The ability to automatically detect FPs has important implications for understanding individual differences in taste perception, which could eventually help in diagnosing taste disorders or guiding personalized nutrition plans.