arXiv subjects

Giulia Frigerio

Publications and source records attributed to Giulia Frigerio.

4 recordsLinked to original sources

arXiv2026

Towards Reliable and Reproducible Fetal Brain Biometry: A Deep Learning Approach Using MRI

Fetal brain biometry is essential for quantitative assessment of brain development, supporting gestational age estimation, developmental monitoring, and detection of abnormalities. In clinical practice, measurements are manually performed, making them time-consuming and prone to variability. While automated approaches have been proposed, reproducible methods remain limited, particularly those providing anatomically interpretable landmark localization. We present a fully automated deep learning-based framework for reliable and reproducible brain biometry from 3D super-resolution-reconstructed fetal brain MRI. The proposed four-step pipeline derives biometric parameters by jointly estimating linear measurements and their corresponding anatomical landmarks. A 3D convolutional neural network is trained to regress landmark coordinates from brain segmentation label maps, followed by measurement-specific geometric optimization to refine landmark positions and compute measurements. The pipeline is evaluated on two publicly available fetal MRI datasets comprising 150 volumes (gestational age range: 20-37 weeks) acquired across different scanners and protocols, assessing five key biometric measurements across varying acquisition settings and providing a comprehensive evaluation of both measurement accuracy and landmark localization using quantitative metrics and visual assessment. Compared with the only available automated pipeline, the proposed method achieves comparable or improved accuracy for most measurements. In conclusion, we introduce a straightforward pipeline for reliable biometry estimations, with efficiency, interpretability and scalability that support integration into clinical workflows.

cs.CV↗

arXiv2026

Discovering structural, electronic and excitonic properties of bulk, nanostructured and doped C3N4 in diamond- and graphitic-like phases

In this systematic density functional theory study, we compare a standard gradient corrected functional (PBE) with a long-range hybrid functional (HSE06), with and without correction for the dispersion forces, relative to their ability to correctly reproduce structural and electronic properties of different bulk 3D C3N4 phases, encompassing diamond- and graphitic-like models. Corrugation is found to provide further stabilization to the layered structures with all methods. We observe that HSE06-D3 method provides results in good agreement with experimental data and with more sophisticated G0W0 calculations. Based on that, we exploited the method to investigate the nature of the bulk triplet excitons in these C3N4 structures to evaluate the S0-T1 energy difference, the self-trapping triplet exciton energy and the photoluminescence emission energy since this is a promising visible-light photocatalyst. Nanostructuring (0D and 2D) is another relevant aspect of these materials in practical applications; therefore, we have considered the effect of single- or multilayer exfoliation or space confinement in nanoparticles. Finally, we also discuss how the introduction of extrinsic dopants (e.g., S atoms) in the nanostructures modifies the atomic and electronic structure.

cond-mat.mtrl-sci↗

arXiv2022

Multi-scale modeling of folic acid-functionalized TiO$_{2}$ nanoparticles for active targeting of tumor cells

Strategies based on the active targeting of tumor cells are emerging as smart and efficient nanomedical procedures. Folic acid (FA) is a vitamin and a well-established tumor targeting agent because of its strong affinity for the folate receptor (FR), which is an overexpressed protein on the cell membranes of the tumor cells. FA can be successfully anchored to several nanocarriers, including inorganic nanoparticles (NPs) based on transition metal oxides. Among them, TiO$_{2}$ is extremely interesting because of its excellent photoabsorption and photocatalytic properties, which can be exploited in photodynamic therapy. However, it is not yet clear in which respects direct anchoring of FA to the NP or the use of spacers, based on polyethylene glycol (PEG) chains, are different and whether one approach is better than the other. In this work, we combine Quantum Mechanics (QM) and Classical Molecular Dynamics (MD) to design and optimize the FA functionalization on bare and PEGylated TiO$_{2}$ models and to study the dynamical behavior of the resulting nanoconjugates in pure water environment and in physiological conditions. We observe that they are chemically stable, even under the effect of increasing temperature (up to 500 K). Using the results from long MD simulations (100 ns) and from free energy calculations, we determine how the density of FA molecules on the TiO$_{2}$ NP and the presence of PEG spacers impact on the actual exposure of the ligands, especially by affecting the extent of FA-FA intermolecular interactions, which are detrimental for the targeting ability of FA towards the folate receptor. This analysis provides a solid and rational basis for experimentalists to define the optimal FA density and the more appropriate mode of anchoring to the carrier, according to the final purpose of the nanoconjugate.

cond-mat.mtrl-sci↗

arXiv2022

Molecular dynamics simulations of cRGD-conjugated PEGylated TiO$_2$ nanoparticles for targeted photodynamic therapy

The conjugation of high-affinity cRGD-containing peptides is a promising approach in nanomedicine to efficiently reduce off-targeting effects and enhance the cellular uptake by integrin-overexpressing tumor cells. Herein we utilize atomistic molecular dynamics simulations to evaluate key structural-functional parameters of these targeting ligands for an effective binding activity towards $α_Vβ_3$ integrins. An increasing number of cRGD ligands is conjugated to PEG chains grafted to highly curved TiO$_2$ nanoparticles to unveil the impact of cRGD density on the ligand's presentation, stability, and conformation in an explicit aqueous environment. We find that a low density leads to an optimal spatial presentation of cRGD ligands out of the "stealth" PEGylated layer around the nanosystem, favoring a straight upward orientation and spaced distribution of the targeting ligands in the bulk-water phase. On the contrary, high densities favor clustering and internalization of cRGD ligands in the inner region of the PEGylated layer, driven by a concerted mechanism of enhanced ligand-ligand interactions and reduced water accessibility over the ligand's molecular surface. These findings strongly suggest that the ligand density modulation is a key factor in the design of cRGD-targeting nanodevices to maximize their binding efficiency into over-expressed $α_Vβ_3$ integrin receptors.

cond-mat.soft↗