arXiv subjects

Matteo Negro

Publications and source records attributed to Matteo Negro.

7 recordsLinked to original sources

arXiv2026

RamanSeg: Interpretability-driven Deep Learning on Raman Spectra for Cancer Diagnosis

Histopathology, the current gold standard for cancer diagnosis, involves the manual examination of tissue samples after chemical staining, a time-consuming process requiring expert analysis. Raman spectroscopy is an alternative, stain-free method of extracting information from samples. Using nnU-Net, we trained a segmentation model on a novel dataset of spatial Raman spectra aligned with tumour annotations, achieving a mean foreground Dice score of 80.9%, surpassing previous work. Furthermore, we propose a novel, interpretable, prototype-based architecture called RamanSeg. RamanSeg classifies pixels based on discovered regions of the training set, generating a segmentation mask. Two variants of RamanSeg allow a trade-off between interpretability and performance: one with prototype projection and another projection-free version. The projection-free RamanSeg outperformed a U-Net baseline with a mean foreground Dice score of 67.3%, offering a meaningful improvement over a black-box training approach.

eess.IV↗

arXiv2025

Relational Causal Discovery with Latent Confounders

Estimating causal effects from real-world relational data can be challenging when the underlying causal model and potential confounders are unknown. While several causal discovery algorithms exist for learning causal models with latent confounders from data, they assume that the data is independent and identically distributed (i.i.d.) and are not well-suited for learning from relational data. Similarly, existing relational causal discovery algorithms assume causal sufficiency, which is unrealistic for many real-world datasets. To address this gap, we propose RelFCI, a sound and complete causal discovery algorithm for relational data with latent confounders. Our work builds upon the Fast Causal Inference (FCI) and Relational Causal Discovery (RCD) algorithms and it defines new graphical models, necessary to support causal discovery in relational domains. We also establish soundness and completeness guarantees for relational d-separation with latent confounders. We present experimental results demonstrating the effectiveness of RelFCI in identifying the correct causal structure in relational causal models with latent confounders.

cs.LG↗

arXiv2022

Efficient Coherent XUV Generation and Manipulation in Microfluidic Glass Devices

The development of compact and bright XUV and soft X-ray sources based on high-order harmonic generation is boosting advances towards understanding the behavior of matter with extreme temporal and spatial resolutions. Here, we report efficient XUV generation inside microfluidic devices fabricated by femtosecond laser irradiation followed by chemical etching. Our microfluidic approach allows one to control and manipulate the generation conditions in gas on a micro-meter scale with unprecedented flexibility, thus enabling a high photon-flux and broadband harmonics spectra up to 200 eV.

physics.optics↗

arXiv2018

Impulsive laser-induced alignment of OCS molecules at FERMI

We demonstrate the experimental realization of impulsive alignment of carbonyl sulfide (OCS) molecules at the Low Density Matter Beamline (LDM) at the free-electron laser FERMI. OCS molecules in a molecular beam were impulsively aligned using 200 fs pulses from a near-infrared laser. The alignment was probed through time-delayed ionization above the sulphur 2p edge, resulting in multiple ionization via Auger decay and subsequent Coulomb explosion of the molecules. The ionic fragments were collected using a time-of-flight mass spectrometer and the analysis of ion-ion covariance maps confirmed the correlation between fragments after Coulomb explosion. The analysis of the CO+ and S+ channels allowed us to extract the rotational dynamics, which is in agreement with our theoretical description as well as with previous experiments. This result opens the way for a new class of experiments at LDM within the field of coherent control of molecules with the possibilities that a precisely synchronized optical-pump XUV-probe laser setup like FERMI can offer.

physics.chem-ph↗

arXiv2017

Laser induced strong-field ionization gas jet tomography

We introduce a novel in-situ strong field ionization tomography approach for characterizing the spatial density distribution of gas jets. We show that for typical intensities in high harmonic generation experiments, the strong field ionization mechanism used in our approach provides an improvement in the resolution close to factor of 2 (resolving about 8 times smaller voxel volume), when compared to linear/single-photon imaging modalities. We find, that while the depth of scan in linear tomography is limited by resolution loss due to the divergence of the driving laser beam, in the proposed approach the depth of focus is localized due to the inherent physical nature of strong-field interaction and discuss implications of these findings. We explore key aspects of the proposed method and compare it with commonly used single- and multi-photon imaging mechanisms. The proposed method will be particularly useful for strong field and attosecond science experiments.

physics.optics↗

arXiv2016

Probe of Multi-electron Dynamics in Xenon by Caustics in High Order Harmonic Generation

We investigated the giant resonance in Xenon by high-order harmonic generation spectroscopy driven by a two-color field. The addition of a non-perturbative second harmonic component parallel to the driving field breaks the symmetry between neighboring sub-cycles resulting in the appearance of spectral caustics at two distinct cut-off energies. By controlling the phase delay between the two color components it is possible to tailor the harmonic emission in order to amplify and isolate the spectral feature of interest. In this paper we demonstrate how this control scheme can be used to investigate the role of electron correlations that give birth to the giant resonance in Xenon. The collective excitations of the giant dipole resonance in Xenon combined with the spectral manipulation associated with the two color driving field allow to see features that are normally not accessible and to obtain a quantitative good agreement between the experimental results and the theoretical predictions.

physics.atom-ph↗

arXiv2014

Studying the universality of field induced tunnel ionization times via high-order harmonic spectroscopy

High-harmonics generation spectroscopy is a promising tool for resolving electron dynamics and structure in atomic and molecular systems. This scheme, commonly described by the strong field approximation, requires a deep insight into the basic mechanism that leads to the harmonics generation. Recently, we have demonstrated the ability to resolve the first stage of the process -- field induced tunnel ionization -- by adding a weak perturbation to the strong fundamental field. Here we generalize this approach and show that the assumptions behind the strong field approximation are valid over a wide range of tunnel ionization conditions. Performing a systematic study -- modifying the fundamental wavelength, intensity and atomic system -- we observed a good agreement with quantum path analysis over a range of Keldysh parameters. The generality of this scheme opens new perspectives in high harmonics spectroscopy, holding the potential of probing large, complex molecular systems.

physics.atom-ph↗