Francesco Regazzoni

Federica Caforio, Francesco Regazzoni, Stefano Pagani et al.

The development of biophysical models for clinical applications is rapidly advancing in the research community, thanks to their predictive nature and their ability to assist the interpretation of clinical data. However, high-resolution and accurate multi-physics computational models are computationally expensive and their personalisation involves fine calibration of a large number of parameters, which may be space-dependent, challenging their clinical translation. In this work, we propose a new approach which relies on the combination of physics-informed neural networks (PINNs) with three-dimensional soft tissue nonlinear biomechanical models, capable of reconstructing displacement fields and estimating heterogeneous patient-specific biophysical properties. The proposed learning algorithm encodes information from a limited amount of displacement and, in some cases, strain data, that can be routinely acquired in the clinical setting, and combines it with the physics of the problem, represented by a mathematical model based on partial differential equations, to regularise the problem and improve its convergence properties. Several benchmarks are presented to show the accuracy and robustness of the proposed method and its great potential to enable the robust and effective identification of patient-specific, heterogeneous physical properties, s.a. tissue stiffness properties. In particular, we demonstrate the capability of the PINN to detect the presence, location and severity of scar tissue, which is beneficial to develop personalised simulation models for disease diagnosis, especially for cardiac applications.

Luca Muscarnera, Luigi Loreti, Giovanni Todeschini et al.

Randomness is ubiquitous in many applications across data science and machine learning. Remarkably, systems composed of random components often display emergent global behaviors that appear deterministic, manifesting a transition from microscopic disorder to macroscopic organization. In this work, we introduce a theoretical model for studying the emergence of collective behaviors in ensembles of random classifiers. We argue that, if the ensemble is weighted through the Gibbs measure defined by adopting the classification loss as an energy, then there exists a finite temperature parameter for the distribution such that the classification is optimal, with respect to the loss (or the energy). Interestingly, for the case in which samples are generated by a Gaussian distribution and labels are constructed by employing a teacher perceptron, we analytically prove and numerically confirm that such optimal temperature does not depend neither on the teacher classifier (which is, by construction of the learning problem, unknown), nor on the number of random classifiers, highlighting the universal nature of the observed behavior. Experiments on the MNIST dataset underline the relevance of this phenomenon in high-quality, noiseless, datasets. Finally, a physical analogy allows us to shed light on the self-organizing nature of the studied phenomenon.

Christian Pilato, Francesco Regazzoni

High-level synthesis (HLS) is a key component for the hardware acceleration of applications, especially thanks to the diffusion of reconfigurable devices in many domains, from data centers to edge devices. HLS reduces development times by allowing designers to raise the abstraction level and use automated methods for hardware generation. Since security concerns are becoming more and more relevant for data-intensive applications, we investigate how to abstract security properties and use HLS for their integration with the accelerator functionality. We use the case of dynamic information flow tracking, showing how classic software-level abstractions can be efficiently used to hide implementation details to the designers.

Francesco Regazzoni, Emna Amri, Samuel Burri et al.

The security of electronic devices has become a key requisite for the rapidly-expanding pervasive and hyper-connected world. Robust security protocols ensuring secure communication, device's resilience to attacks, authentication control and users privacy need to be implemented. Random Number Generators (RNGs) are the fundamental primitive in most secure protocols but, often, also the weakest one. Establishing security in billions of devices requires high quality random data generated at a sufficiently high throughput. On the other hand, the RNG should exhibit a high integration level with on-chip extraction to remove, in real time, potential imperfections. We present the first integrated Quantum RNG (QRNG) in a standard CMOS technology node. The QRNG is based on a parallel array of independent Single-Photon Avalanche Diodes (SPADs), homogeneously illuminated by a DC-biased LED, and co-integrated logic circuits for postprocessing. We describe the randomness generation process and we prove the quantum origin of entropy. We show that co-integration of combinational logic, even of high complexity, does not affect the quality of randomness. Our CMOS QRNG can reach up to 400 Mbit/s throughput with low power consumption. Thanks to the use of standard CMOS technology and a modular architecture, our QRNG is suitable for a highly scalable solution.

Madura A Shelton, Niels Samwel, Lejla Batina et al.

Since their introduction over two decades ago, side-channel attacks have presented a serious security threat. While many ciphers' implementations employ masking techniques to protect against such attacks, they often leak secret information due to unintended interactions in the hardware. We present Rosita, a code rewrite engine that uses a leakage emulator which we amend to correctly emulate the micro-architecture of a target system. We use Rosita to automatically protect masked implementations of AES, ChaCha, and Xoodoo. For AES and Xoodoo, we show the absence of observable leakage at 1,000,000 traces with less than 21% penalty to the performance. For ChaCha, which has significantly more leakage, Rosita eliminates over 99% of the leakage, at a performance cost of 64%.