A5016940537- Radiomics and Machine Learning in Medical Imaging
- Radiation Detection and Scintillator Technologies
- Radiation Therapy and Dosimetry
- Particle Detector Development and Performance
- Brain Tumor Detection and Classification
- Particle physics theoretical and experimental studies
- Colorectal Cancer Screening and Detection
- Gastric Cancer Management and Outcomes
- Advanced Radiotherapy Techniques
- Nuclear Physics and Applications
- Advanced Data Storage Technologies
- Helicobacter pylori-related gastroenterology studies
- Embedded Systems Design Techniques
- Anomaly Detection Techniques and Applications
- Explainable Artificial Intelligence (XAI)
- CCD and CMOS Imaging Sensors
- Blind Source Separation Techniques
- Parallel Computing and Optimization Techniques
- Data Visualization and Analytics
- Advanced Neural Network Applications
- Computational Physics and Python Applications
- Colorectal Cancer Surgical Treatments
- Endometrial and Cervical Cancer Treatments
- Interconnection Networks and Systems
- Medical Imaging Techniques and Applications
Sapienza University of Rome
2019-2025
Istituto Superiore di Sanità
2025
Istituto Nazionale di Fisica Nucleare, Sezione di Roma I
2020-2024
Istituto Nazionale di Fisica Nucleare
2019
Abstract Background Endometrial cancer (EC) is one of the most common gynecological malignancies and second malignancy cause death in women. Heterogeneous tissues with different grades complexity diffusion properties characterize EC. Several magnetic resonance imaging (DMRI) protocols have been used to perform a non‐invasive global evaluation EC for diagnostic prognostic purposes. However, association single value coefficient an tissue could be severe limit developing DMRI virtual histology...
APEIRON is a framework encompassing the general architecture of distributed heterogeneous processing platform and corresponding software stack, from low level device drivers up to high programming model. Developers can define scalable applications that be deployed on multi-FPGA system coding at level: communication IPs allow low-latency between tasks FPGAs, even if hosted different computing nodes. Thanks use High Level Synthesis tools, are described in language (C/C++) while expressed...
FPGA-RICH is an FPGA-based online partial particle identification system for the NA62 experiment employing AI techniques. Integrated between readout of Ring Imaging Cherenkov detector (RICH) and low-level trigger processor (L0TP+), implements a fast pipeline to process in real-time RICH raw hit data stream, producing primitives containing elaborate physics information—e.g., number charged particles event—that L0TP+ can use improve decision efficiency. Deployed on single FPGA, combines...
Spread through air spaces (STAS) has been reported as a negative prognostic factor in patients with lung cancer undergoing sublobar resection. Radiomics recently proposed to predict STAS using preoperative computed tomography (CT). However, limitations of previous studies included the strict selection imaging acquisition protocols, leading results hardly applicable daily clinical practice. The aim this study is test radiomics-based prediction model practice-based dataset.A training cohort 99...
Aims Gastric cancer (GC) is a significant healthcare concern and the recognition of high-risk patients crucial [1]. The common mechanism for progression GC Correa cascade, which outlines multi-step from chronic gastritis through atrophic gastritis, intestinal metaplasia (IM), dysplasia finally GC. detection IM in stomach not always easy, especially when it focal. While electronic chromoendoscopy has demonstrated high diagnostic accuracy [2], its limited availability, variable expertise...
High Energy Physics (HEP) Trigger and Data Acquisition systems (TDAQs) need ever increasing throughput real-time data analytics capabilities either to improve particle identification accuracy further suppress background events in trigger or perform an efficient online reduction for trigger-less ones. As the requirements imposed by HEP TDAQs applications class of dataflow processing, FPGA devices are a good fit inasmuch they can not only provide adequate compute, memory I/O resources but also...
Graphs are versatile structures for the representation of many real-world data. Deep Learning on graphs is currently able to solve a wide range problems with excellent results. However, both generation and handling large still remain open challenges. This work aims introduce techniques generating test approach complex problem such as calculation dose distribution in oncological radiotherapy applications. To this end, we introduced pooling technique (ReNN-Pool) capable sampling nodes that...
In recent years Artificial Intelligence has emerged as a fundamental tool in medical applications. Despite this rapid development, deep neural networks remain black boxes that are difficult to explain, and represents major limitation for their use clinical practice. We focus on the segmentation of images task, where most explainability methods proposed so far provide visual explanation terms an input saliency map. The aim work is extend, implement test instead influence-based algorithm,...
Gastric cancer (GC) is a significant healthcare concern, and the identification of high-risk patients crucial. Indeed, gastric precancerous conditions present diagnostic challenges, particularly early intestinal metaplasia (IM) detection. This study developed deep learning system to assist in IM detection using image patches from corpus examined virtual chromoendoscopy Western country. Utilizing retrospective dataset endoscopic images Sant'Andrea University Hospital Rome, collected between...
Introduction External beam radiotherapy (RT) is one of the most common treatments against cancer, with photon-based RT and particle therapy being commonly employed modalities. Very high energy electrons (VHEE) have emerged as promising candidates for novel treatments, particularly in exploiting FLASH effect, offering potential advantages over traditional Methods This paper introduces a Deep Learning model based on graph convolutional networks to determine dose distributions therapeutic VHEE...
The L0TP+ initiative is aimed at the upgrade of FPGA-based Level-0 Trigger Processor (L0TP) NA62 experiment CERN for post-LS2 data taking, which expected to happen 100% design beam intensity, corresponding about 3.3 × 10 12 protons per pulse on beryllium target used produce kaons beam. Although tests performed end 2018 showed a substantial robustness L0TP system also full there are several reasons motivate such an upgrade: i) avoid FPGA platform obsolescence, ii) make room improvements in...
Abstract A new FPGA-based low-level trigger processor has been installed at the NA62 experiment. It is intended to extend features of its predecessor due a faster interconnection technology and additional logic resources available on platform. With aim improving selectivity exploring architectures for complex computation, GPU system developed neural network FPGA in progress. They both process data streams from ring imaging Cherenkov detector experiment extract real time high level logic....
We present a Deep Learning generative model specialized to work with graphs regular geometry. It is build on Variational Autoencoder framework and employs Graph convolutional layers in both encoding decoding phases. also introduce pooling technique (ReNN-Pool), used the encoder, that allows downsample graph nodes spatially uniform highly interpretable way. In decoder, symmetrical un-pooling retrieve original dimensionality of graphs. Performance are tested standard Sprite benchmark dataset,...
APEIRON is a framework encompassing the general architecture of distributed heterogeneous processing platform and corresponding software stack, from low level device drivers up to high programming model. The designed be efficiently used for studying, prototyping deploying smart trigger data acquisition (TDAQ) systems energy physics experiments.