arXiv subjects

Evaristo Cisbani

Publications and source records attributed to Evaristo Cisbani.

4 recordsLinked to original sources

arXiv2026

Optimal stimulation sites are not the most affected: personalised models of resting-state fMRI in Alzheimer's disease

Resting-state functional connectivity (FC) is altered in Alzheimer's disease (AD), widely regarded as a distributed network process; whether its signature reduces to a few focal sites has not been tested causally, a question central to targeted neuromodulation. We fit subject-specific, cross-subject-identifiable models whose free-running dynamics reproduce those of each individual patient. The fitted model parameters classify AD from controls at modest accuracy, below that of structural atrophy; we build on the functional model nonetheless, because dynamics, not tissue loss, are what stimulation can act on. Changing a virtual patient's model connectivity toward the control template reverts its AD classification, establishing in silico that the disease signature is correctable, yet the required correction is intrinsically distributed: a coordinated, multi-site change of the model connectivity that no single-node edit reproduces. Where, then, should a physically realisable focal drive act? A single-site drive at the node whose connectivity is most altered fails to revert the classification even at supra-physiological amplitudes, whereas selecting each patient's site by its effect on the disease discriminant achieves complete, individualised reclassification from one site, and a real-time closed-loop controller reaches comparable efficacy at lower dose using only causally available information. Optimal targets are cortical and heterogeneous: the site to stimulate is not where connectivity is most altered but where the network is most therapeutically responsive.

q-bio.NC↗

arXiv2026

Adversarial robustness of a U-Net-based model observer for CT protocol optimization

Artificial intelligence is increasingly used in medical imaging, yet its robustness to input perturbations remains a critical concern for a wide clinical adoption. To this end, we used adversarial examples to systematically probe vulnerabilities of a U-Net-based model observer for computed tomography protocol optimization, performing detection and localization of low-contrast objects in a phantom dataset. Adversarial attacks were generated using both gradient-based and optimization-based white-box methods. Fast gradient perturbations produced high misclassification rates, reaching up to 75% at intermediate perturbation levels while remaining visually imperceptible. Localization was more robust, with success rates of about 25% for small perturbations and 42% at moderate levels. In contrast, optimization-based attack achieved success rates close to 50% for both tasks. To mitigate these vulnerabilities, dynamic adversarial training was implemented. This reduced the success rate of optimization-based attacks to 7% for classification and 13% when including localization-specific training, demonstrating a substantial robustness improvement without compromising task performances, confirmed by localization receiver operating characteristic analysis. To further interpret model behavior, radiomic texture analysis was performed on original and adversarial images. While most global image statistics remain stable, specific texture-related features exhibit consistent changes in successful attacks, highlighting the model's sensitivity to subtle local intensity patterns. Overall, adversarial training improves robustness without degrading performance, while radiomic analysis reveals interpretable links between texture alterations and prediction failures, supporting more reliable and explainable AI systems for medical imaging.

physics.med-ph↗

arXiv2018

Calibration of highly segmented, compact gamma camera for Molecular Breast Imaging

Breast cancers is the second leading cause of cancer mortality in women; early diagnosis increase the probability of a successful therapy; any marginal improvement in this direction helps sparing lives. In this context functional imaging techniques such as Molecular Breast Imaging (MBI) represents an important supplemental screening, especially in the more questionable cases. In order to further extend the MBI performances an innovative asymmetric dual detector device, with mixed optics has been recently proposed and prototyped; the sensors are highly segmented with a correspondingly large number of independent, configurable, electronic readout channels with self-triggering capability. This flexible electronics architecture has different advantages in addition to those related to the adopted asymmetric dual detector geometry: real-time event selection based on the adjustable gain and discriminator threshold at single channel (or group of channels) level; repeatable, quick hardware and software channel response equalization; configurable list mode acquisition for versatile offline image processing. These benefits come at the expenses of more complex calibration methods and optimization procedures, which are detailed in the present paper.

physics.ins-det↗

arXiv2008

An empirical study of large, naturally occurring starling flocks: a benchmark in collective animal behaviour

Bird flocking is a striking example of collective animal behaviour. A vivid illustration of this phenomenon is provided by the aerial display of vast flocks of starlings gathering at dusk over the roost and swirling with extraordinary spatial coherence. Both the evolutionary justification and the mechanistic laws of flocking are poorly understood, arguably because of a lack of data on large flocks. Here, we report a quantitative study of aerial display. We measured the individual three-dimensional positions in compact flocks of up to 2700 birds. We investigated the main features of the flock as a whole - shape, movement, density and structure - and discuss these as emergent attributes of the grouping phenomenon. We find that flocks are relatively thin, with variable sizes, but constant proportions. They tend to slide parallel to the ground and, during turns, their orientation changes with respect to the direction of motion. Individual birds keep a minimum distance from each other that is comparable to their wingspan. The density within the aggregations is non-homogeneous, as birds are packed more tightly at the border compared to the centre of the flock. These results constitute the first set of large-scale data on three-dimensional animal aggregations. Current models and theories of collective animal behaviour can now be tested against these results.

q-bio.QM↗