2026
Other
Restricted
Relazione CNR-ISTI / SIMeM CNR-ISTI / SIMeM. Accordo di collaborazione scientifica - Relazione conclusiva sulle attività svolte e sui risultati di disseminazione
Martinelli Massimo
La presente relazione descrive le attività svolte nell'ambito della collaborazione scientifica tra il Consiglio Nazionale delle Ricerche - Istituto di Scienza e Tecnologie dell'Informazione "A. Faedo" (CNR-ISTI) e la Società Italiana di Medicina di Montagna (SIMeM), con riferimento al periodo aprile 2020 - aprile 2026. Sono inoltre richiamate, come contesto scientifico e tecnologico, alcune attività propedeutiche avviate nel periodo 2016-2019 e risultati collegati alla sinergia tra i rispettivi gruppi di ricerca. La collaborazione ha integrato competenze di medicina di montagna, telemedicina, sistemi informativi, gestione di dati clinico-epidemiologici, supporto alla decisione e metodi di intelligenza artificiale.
Martinelli Massimo
La presente relazione descrive le attività svolte nell'ambito della collaborazione scientifica tra il Consiglio Nazionale delle Ricerche - Istituto di Scienza e Tecnologie dell'Informazione "A. Faedo" (CNR-ISTI) e la Società Italiana di Medicina di Montagna (SIMeM), con riferimento al periodo aprile 2020 - aprile 2026. Sono inoltre richiamate, come contesto scientifico e tecnologico, alcune attività propedeutiche avviate nel periodo 2016-2019 e risultati collegati alla sinergia tra i rispettivi gruppi di ricerca. La collaborazione ha integrato competenze di medicina di montagna, telemedicina, sistemi informativi, gestione di dati clinico-epidemiologici, supporto alla decisione e metodi di intelligenza artificiale.
2026
Other
Restricted
Spoke 9 - AGRITECH 36-MONTH REPORT
Pucci Laura, Tomassi Elena, Arouna Nafiou, Gabriele Morena, Peres Fabbri Laryssa, Pozzo Luisa, Conte Giuseppe, Cremonesi Paola, Castiglioni Bianca, Moroni Davide, Martinelli Massimo
This document represents the 36-month report on products of animal origin intended for human consumption.Project(s): Spoke 9 AGRITECH
Pucci Laura, Tomassi Elena, Arouna Nafiou, Gabriele Morena, Peres Fabbri Laryssa, Pozzo Luisa, Conte Giuseppe, Cremonesi Paola, Castiglioni Bianca, Moroni Davide, Martinelli Massimo
This document represents the 36-month report on products of animal origin intended for human consumption.Project(s): Spoke 9 AGRITECH
2026
Contribution to journal
Restricted
Blood pressure measurement in Italian mountain shelters: a world hypertension day initiative
Bilo Grzegorz, Zanotti Lucia, Croce Alessandro, D’angelo Carla, Faini Andrea, Martinelli Massimo, Muiesan Maria Lorenza, Pengo Martino F., Pratali Lorenza, Vega Deceno Jose Ivan, Soranna Davide, Zambon Antonella, Virdis Agostino, Strapazzon Giacomo, Agazzi Giancelso, Parati Gianfranco
World Hypertension Day and May Measurement Month are large-scale population initiatives aimed at increasing awareness of hypertension and providing opportunistic blood pressure (BP) screening. In Italy, “Blood Pressure Measurement in Mountain Shelters” has been conducted since 2016 as a joint initiative of Italian Society of Hypertension, Italian Alpine Club and Italian Society of Mountain Medicine, with the aim of providing advice and education on hypertension and cardiovascular risk among visitors of mountain shelters. Methods. Trained volunteers provided free advice to visitors of mountain shelters throughout Italy. Basic demographic, lifestyle and clinical information was collected with anonymous questionnaire. Seated BP was obtained using manual or oscillometric devices as the average of three measurements. The initiative took place in summer when mountain attendance is highest. The present analysis of merged data collected in 2019, 2022, 2023, 2024 and 2025 editions focuses on describing participant characteristics, including BP, stratified by altitude of data collection. Results. After exclusion of incomplete records (missing age, sex or BP), data of 12,317 participants from 140 mountain shelters were analysed. No relevant differences were observed across the different years in terms of participant characteristics or BP levels. Participant characteristics are reported in the Table. Median BP and body mass index were within normal limits, and the prevalence of diabetes, hypercholesterolaemia and hypertension was lower than that in the general Italian population. Fewer than 15% of participants reported taking at least one antihypertensive medication. Higher altitudes of data collection were associated with higher male sex prevalence and heart rate and with lower age, BMI, SpO2, prevalence of hypercholesterolemia and hypertension. Only minor differences were observed in measured BP. Conclusions. Blood Pressure Measurement in Mountain Shelters is a unique initiative, willingly attended by visitors to mountain areas. Participants generally displayed a favourable cardiovascular risk profile; however, a substantial proportion, including those assessed at higher altitudes, presented at least one cardiovascular risk factor. This screening and educational initiative represents a valuable opportunity to identify individuals at risk and to provide counselling on cardiovascular prevention among visitors of mountain shelters.Source: JOURNAL OF HYPERTENSION, vol. 44(Suppl 1), pp. 70-71
DOI: 10.1097/01.hjh.0001195860.36068.a1
Metrics:
Bilo Grzegorz, Zanotti Lucia, Croce Alessandro, D’angelo Carla, Faini Andrea, Martinelli Massimo, Muiesan Maria Lorenza, Pengo Martino F., Pratali Lorenza, Vega Deceno Jose Ivan, Soranna Davide, Zambon Antonella, Virdis Agostino, Strapazzon Giacomo, Agazzi Giancelso, Parati Gianfranco
World Hypertension Day and May Measurement Month are large-scale population initiatives aimed at increasing awareness of hypertension and providing opportunistic blood pressure (BP) screening. In Italy, “Blood Pressure Measurement in Mountain Shelters” has been conducted since 2016 as a joint initiative of Italian Society of Hypertension, Italian Alpine Club and Italian Society of Mountain Medicine, with the aim of providing advice and education on hypertension and cardiovascular risk among visitors of mountain shelters. Methods. Trained volunteers provided free advice to visitors of mountain shelters throughout Italy. Basic demographic, lifestyle and clinical information was collected with anonymous questionnaire. Seated BP was obtained using manual or oscillometric devices as the average of three measurements. The initiative took place in summer when mountain attendance is highest. The present analysis of merged data collected in 2019, 2022, 2023, 2024 and 2025 editions focuses on describing participant characteristics, including BP, stratified by altitude of data collection. Results. After exclusion of incomplete records (missing age, sex or BP), data of 12,317 participants from 140 mountain shelters were analysed. No relevant differences were observed across the different years in terms of participant characteristics or BP levels. Participant characteristics are reported in the Table. Median BP and body mass index were within normal limits, and the prevalence of diabetes, hypercholesterolaemia and hypertension was lower than that in the general Italian population. Fewer than 15% of participants reported taking at least one antihypertensive medication. Higher altitudes of data collection were associated with higher male sex prevalence and heart rate and with lower age, BMI, SpO2, prevalence of hypercholesterolemia and hypertension. Only minor differences were observed in measured BP. Conclusions. Blood Pressure Measurement in Mountain Shelters is a unique initiative, willingly attended by visitors to mountain areas. Participants generally displayed a favourable cardiovascular risk profile; however, a substantial proportion, including those assessed at higher altitudes, presented at least one cardiovascular risk factor. This screening and educational initiative represents a valuable opportunity to identify individuals at risk and to provide counselling on cardiovascular prevention among visitors of mountain shelters.Source: JOURNAL OF HYPERTENSION, vol. 44(Suppl 1), pp. 70-71
DOI: 10.1097/01.hjh.0001195860.36068.a1
Metrics:
See at:
CNR IRIS 
| CNR IRIS | CNR IRIS | journals.lww.com
2026
Other
Metadata Only Access
AI4GO: AI Framework for Gamma-Oryzanol-rich oil.
Massimo Martinelli
A Collaborative Framework for Non-Invasive Monitoring and Predictive Quality Control and Zero-Waste Efficiency in Rice Bran Oil.
Massimo Martinelli
A Collaborative Framework for Non-Invasive Monitoring and Predictive Quality Control and Zero-Waste Efficiency in Rice Bran Oil.
See at:
CNR IRIS
2026
Conference article
Open Access
Advancements in Artificial Intelligence and Computer Vision for Biomedical Applications @SI-Lab
Marco Cafiso, Andrea Carboni, Claudia Caudai, Sara Colantonio, Francesco Conti, Mario D’acunto, Said Daoudagh, Giulio Del Corso, Riccardo Del Gratta, Danila Germanese, Giacomo Ignesti, Gianmarco Lazzini, Giuseppe Riccardo Leone, Barbara Leporini, Massimo Magrini, Massimo Martinelli, Davide Moroni, Ali Reza Omrani, Francesca Pardini, Maria Antonietta Pascali, Paolo Paradisi, Laura Sebastiaini, Marco Tampucci, Valeria Tateo, Federico Volpini
This paper summarizes recent research at the Signals and Images Lab (ISTI-CNR) leveraging artificial intelligence, machine learning, and computer vision to address complex biomedical challenges. We highlight advancements in trustworthy, explainable algorithmic foundations alongside their practical clinical applications across diagnostic imaging, physiological signal analysis and neuromotor rehabilitation.
Marco Cafiso, Andrea Carboni, Claudia Caudai, Sara Colantonio, Francesco Conti, Mario D’acunto, Said Daoudagh, Giulio Del Corso, Riccardo Del Gratta, Danila Germanese, Giacomo Ignesti, Gianmarco Lazzini, Giuseppe Riccardo Leone, Barbara Leporini, Massimo Magrini, Massimo Martinelli, Davide Moroni, Ali Reza Omrani, Francesca Pardini, Maria Antonietta Pascali, Paolo Paradisi, Laura Sebastiaini, Marco Tampucci, Valeria Tateo, Federico Volpini
This paper summarizes recent research at the Signals and Images Lab (ISTI-CNR) leveraging artificial intelligence, machine learning, and computer vision to address complex biomedical challenges. We highlight advancements in trustworthy, explainable algorithmic foundations alongside their practical clinical applications across diagnostic imaging, physiological signal analysis and neuromotor rehabilitation.
2026
Conference article
Open Access
See at:
CNR IRIS 
| sites.google.com | CNR IRIS
2026
Conference article
Restricted
Reliable and trustworthy learning prototype: insight from POCUS
Ignesti G., D’angelo G., Pratali L., Moroni D., Martinelli M.
Deep learning models often lack the interpretability and trustworthiness required for clinical use. This paper proposes a prototype-regularised training method to analyse 1,208 lung ultrasound images, focusing on B-line artefacts. A ConvNeXt- Tiny architecture is used, adding a novel reconstruction loss to the standard classification loss. The model is guided to extract meaningful prototypes and uses them to classify the ultrasound images. To prevent these constraints from hindering generalisation, it is used in pairs with the proposed reconstruction loss, a set of plausible data augmentation of the ideal researched prototypes, and a geometry-aware network, a spatial transformer network, to measure which solutions help the network towards outputting the most reliable outcomes. The resulting models are precise, lightweight and interpretable, indicating that the proposed solution can be embedded in an ultrasound device to assist healthcare specialists in point-of-care applications.Project(s): TiAssisto
Ignesti G., D’angelo G., Pratali L., Moroni D., Martinelli M.
Deep learning models often lack the interpretability and trustworthiness required for clinical use. This paper proposes a prototype-regularised training method to analyse 1,208 lung ultrasound images, focusing on B-line artefacts. A ConvNeXt- Tiny architecture is used, adding a novel reconstruction loss to the standard classification loss. The model is guided to extract meaningful prototypes and uses them to classify the ultrasound images. To prevent these constraints from hindering generalisation, it is used in pairs with the proposed reconstruction loss, a set of plausible data augmentation of the ideal researched prototypes, and a geometry-aware network, a spatial transformer network, to measure which solutions help the network towards outputting the most reliable outcomes. The resulting models are precise, lightweight and interpretable, indicating that the proposed solution can be embedded in an ultrasound device to assist healthcare specialists in point-of-care applications.Project(s): TiAssisto
2026
Journal article
Open Access
Oxy-inflammatory profile of finishers and non-finishers in an extreme ultra-endurance trail race: the 866-km Transpyrénéa
Mrakic-Sposta Simona, Gussoni Maristella, Mrakic-Sposta Federica, Giardini Guido, Pratali Lorenza, Montorsi Michela, Tonacci Alessandro, Dellanoce Cinzia, Martinelli Massimo, Vezzoli Alessandra
This study investigates the bio-physiological responses occurring under extreme stress conditions and the characterization of the oxy-inflammatory profile of Finishers (FRs) and NoFinishers (NFRs) athletes during the time course and following the Transpyrénéa, an 866 km extreme ultra-race across the French Pyrenees with an altitude difference of 52,900+ m ascent. Thirty-nine experienced ultra-marathon runners (age 43.5 ± 9.1 years; weight 72.1 ± 11.1 kg; BMI 23.3 ± 2.6 kg/m2) were studied using minimally invasive methods on capillary blood and urine samples obtained at baseline (T0), during (T1, 2, 3) and at the end (T4) of the race. Reactive Oxygen Species (ROS) production, total antioxidant capacity (TAC), oxidative damage (8-hydroxy-2-deoxy Guanosine: 8-OH-dG and 8-isoprostane: 8-isoPGF2α), inflammatory (IL-6), nitric oxide pathway (NOx and 3-NT), neopterin, and hematologic (lactate, and hematocrit) biomarkers were assessed. In both FR and NFR athletes a marked systemic increase in ROS, oxidative and nitrosative damage, inflammation, transient immune-renal dysfunction and lactate release were detected throughout the race. Compared to FRs, NFRs displayed significant differences concerning ROS production at T0, 8-isoPGF2-α at T0, T1 and T2, and perceived exertion (RPE score) at T2. These data potentially reflect enhanced adaptative responses to training and metabolic efficacy in FRs, allowing them to better tolerate extreme physiological stress.Source: INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES, vol. 27 (issue 10)
DOI: 10.3390/ijms27104295
Metrics:
Mrakic-Sposta Simona, Gussoni Maristella, Mrakic-Sposta Federica, Giardini Guido, Pratali Lorenza, Montorsi Michela, Tonacci Alessandro, Dellanoce Cinzia, Martinelli Massimo, Vezzoli Alessandra
This study investigates the bio-physiological responses occurring under extreme stress conditions and the characterization of the oxy-inflammatory profile of Finishers (FRs) and NoFinishers (NFRs) athletes during the time course and following the Transpyrénéa, an 866 km extreme ultra-race across the French Pyrenees with an altitude difference of 52,900+ m ascent. Thirty-nine experienced ultra-marathon runners (age 43.5 ± 9.1 years; weight 72.1 ± 11.1 kg; BMI 23.3 ± 2.6 kg/m2) were studied using minimally invasive methods on capillary blood and urine samples obtained at baseline (T0), during (T1, 2, 3) and at the end (T4) of the race. Reactive Oxygen Species (ROS) production, total antioxidant capacity (TAC), oxidative damage (8-hydroxy-2-deoxy Guanosine: 8-OH-dG and 8-isoprostane: 8-isoPGF2α), inflammatory (IL-6), nitric oxide pathway (NOx and 3-NT), neopterin, and hematologic (lactate, and hematocrit) biomarkers were assessed. In both FR and NFR athletes a marked systemic increase in ROS, oxidative and nitrosative damage, inflammation, transient immune-renal dysfunction and lactate release were detected throughout the race. Compared to FRs, NFRs displayed significant differences concerning ROS production at T0, 8-isoPGF2-α at T0, T1 and T2, and perceived exertion (RPE score) at T2. These data potentially reflect enhanced adaptative responses to training and metabolic efficacy in FRs, allowing them to better tolerate extreme physiological stress.Source: INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES, vol. 27 (issue 10)
DOI: 10.3390/ijms27104295
Metrics:
See at:
CNR IRIS | www.mdpi.com | CNR IRIS
2026
Journal article
Open Access
SWIM (smart weed impact management): A decision support system (DSS) for the site-specific weed control in maize cultivation
Leonardo Ercolini, Nicola Grossi, Davide Moroni, Andrea Berton, Massimo Martinelli, Nicola Silvestri
Purpose Site-specific weed management (SSWM) represents a promising solution for reducing herbicide use while maintaining effective weed control. To encourage the adoption of these techniques, SWIM (Smart Weed Impact Management), a decision support system for generating prescription maps based on RGB imagery, is proposed. Two options are available: a Simple Method (SM) for farmers with limited digital expertise and an Improved Method (IM) for more advanced applications. Methods SWIM operates through three sequential phases: (i) weed detection, aimed at estimating Weed Green Cover (WGC) by subtracting Maize Green Cover (MGC) from Total Green Cover (TGC) from aerial images acquired by drone; (ii) potential damage, aimed at determining the economic intervention threshold based on maize yield losses due to weed competition; (iii) prescription map generation, based on the creation of prescription maps for Patch Spraying or Variable Rate Applications (PSA and VRA). In SM, MGC estimation relies on fixed values of number of plants per unit area and cover of a single plant for the entire field obtained from calibration plots. These values are also partially used by IM, which additionally integrates information derived from the spatial variability of plant density. For maize yield loss assessment, SM uses predefined and/or literature-based parameters, whereas IM relies on site-specific data collected in previous years. Results SWIM showed that both methods provided accurate WGC estimates, with root mean square error values below 0.10 and concordance correlation coefficients above 0.91. However, IM outperformed SM in capturing spatial variability and in defining the economic intervention threshold, as SM tended to overestimate the EIT when applying the Goldsmith model. These differences led to two PSA maps with different potential herbicide savings (44% for SM and 28% for IM). Conclusion The ease of use and cost-effectiveness of SWIM may promote its adoption at farm level and contribute to a reduction in herbicide use in arable cropping systems. Highlights SWIM is a DSS for site-specific weed management (SSWM) in maize cultivation, offering two application levels: a Simple Method (SM) for farmers with limited digital skills and an Improved Method (IM) for more experienced users SWIM simplifies the creation of prescription maps aimed at significantly reducing herbicide use by requiring only a straightforward comparison between remotely sensed data and model-defined thresholds The use of SWIM requires the active involvement of farmers, both in providing field-specific information and in selecting the level of economic risk they are willing to assume Impact SWIM is a cost-effective and user-friendly RGB image-based Decision Support System for site-specific weed management in maize, designed to optimise post-emergence herbicide use. Although further experimental testing is needed to confirm the benefits on the field scale, the proposed model provided reliable estimates of both weed competition levels and the economic intervention threshold, above which weed control becomes economically justified. For these reasons, SWIM can facilitate the wider adoption of digital sensing and precision agriculture techniques, even among farmers without advanced IT skills.Source: PRECISION AGRICULTURE, vol. 27 (issue 89), pp. 1-24
DOI: 10.1007/s11119-026-10392-z
Metrics:
Leonardo Ercolini, Nicola Grossi, Davide Moroni, Andrea Berton, Massimo Martinelli, Nicola Silvestri
Purpose Site-specific weed management (SSWM) represents a promising solution for reducing herbicide use while maintaining effective weed control. To encourage the adoption of these techniques, SWIM (Smart Weed Impact Management), a decision support system for generating prescription maps based on RGB imagery, is proposed. Two options are available: a Simple Method (SM) for farmers with limited digital expertise and an Improved Method (IM) for more advanced applications. Methods SWIM operates through three sequential phases: (i) weed detection, aimed at estimating Weed Green Cover (WGC) by subtracting Maize Green Cover (MGC) from Total Green Cover (TGC) from aerial images acquired by drone; (ii) potential damage, aimed at determining the economic intervention threshold based on maize yield losses due to weed competition; (iii) prescription map generation, based on the creation of prescription maps for Patch Spraying or Variable Rate Applications (PSA and VRA). In SM, MGC estimation relies on fixed values of number of plants per unit area and cover of a single plant for the entire field obtained from calibration plots. These values are also partially used by IM, which additionally integrates information derived from the spatial variability of plant density. For maize yield loss assessment, SM uses predefined and/or literature-based parameters, whereas IM relies on site-specific data collected in previous years. Results SWIM showed that both methods provided accurate WGC estimates, with root mean square error values below 0.10 and concordance correlation coefficients above 0.91. However, IM outperformed SM in capturing spatial variability and in defining the economic intervention threshold, as SM tended to overestimate the EIT when applying the Goldsmith model. These differences led to two PSA maps with different potential herbicide savings (44% for SM and 28% for IM). Conclusion The ease of use and cost-effectiveness of SWIM may promote its adoption at farm level and contribute to a reduction in herbicide use in arable cropping systems. Highlights SWIM is a DSS for site-specific weed management (SSWM) in maize cultivation, offering two application levels: a Simple Method (SM) for farmers with limited digital skills and an Improved Method (IM) for more experienced users SWIM simplifies the creation of prescription maps aimed at significantly reducing herbicide use by requiring only a straightforward comparison between remotely sensed data and model-defined thresholds The use of SWIM requires the active involvement of farmers, both in providing field-specific information and in selecting the level of economic risk they are willing to assume Impact SWIM is a cost-effective and user-friendly RGB image-based Decision Support System for site-specific weed management in maize, designed to optimise post-emergence herbicide use. Although further experimental testing is needed to confirm the benefits on the field scale, the proposed model provided reliable estimates of both weed competition levels and the economic intervention threshold, above which weed control becomes economically justified. For these reasons, SWIM can facilitate the wider adoption of digital sensing and precision agriculture techniques, even among farmers without advanced IT skills.Source: PRECISION AGRICULTURE, vol. 27 (issue 89), pp. 1-24
DOI: 10.1007/s11119-026-10392-z
Metrics:
See at:
CNR IRIS | link.springer.com | CNR IRIS
2026
Conference article
Restricted
What Improves UAV RGB Segmentation of Maize and Weeds?
An Ablation Study of U-Net Components
Massimo Martinelli, Leonardo Ercolini, Nicola Grossi, Nicola Silvestri, Andrea Berton, Davide Moroni
The use of RGB images acquired by unmanned aerial vehicles (UAVs) is increasingly common in precision agriculture, for monitoring crops, weeds, and field conditions. However, large variations in plant size, illumination, shadows, and background appearance make automatic segmentation difficult. A major challenge is discriminating crop plants from weeds that may compete for resources and identifying conditions that may favor the development of plant diseases. In this study, we evaluated U-Net-based semantic segmentation models on 142 UAV RGB images from two maize fields using grouped 5-fold cross-validation in order to segment maize, weeds, and bare soil. We conducted an ablation study on a comprehensive baseline U-Net model (incorporating a ResNet-101 encoder, attention mechanisms, deep supervision, and physics-inspired regularization) by iteratively removing individual components to isolate their impact. The best overlap-based performance was obtained without deep supervision, while the lowest boundary error, measured by the 95th percentile Hausdorff Distance (HD95), was achieved without attention. We saw that using the Excess Green parameter made little or no difference compared to the standard method, while using a fixed threshold reconstruction always worsened performance. However, some examples sometimes contained isolated mislabeled pixels for weeds close to the image borders in the reference mask: these outliers can significantly increase the distance-based metrics, even when the prediction correctly ignores them. For weeds, overlap metrics and visual inspection were more informative than HD95, which was too sensitive to small annotation errors located near the edges of the image. Overall, our results suggest that a simpler segmentation pipeline is preferable for this dataset, that hard-threshold reconstruction is unreliable, and that the quality of annotation is a critical factor when interpreting distance-based metrics for weeds. The study does not propose a new architecture; instead, it provides a controlled ablation protocol and practical evidence that additional U-Net components may not improve small-data UAV RGB segmentation.
Massimo Martinelli, Leonardo Ercolini, Nicola Grossi, Nicola Silvestri, Andrea Berton, Davide Moroni
The use of RGB images acquired by unmanned aerial vehicles (UAVs) is increasingly common in precision agriculture, for monitoring crops, weeds, and field conditions. However, large variations in plant size, illumination, shadows, and background appearance make automatic segmentation difficult. A major challenge is discriminating crop plants from weeds that may compete for resources and identifying conditions that may favor the development of plant diseases. In this study, we evaluated U-Net-based semantic segmentation models on 142 UAV RGB images from two maize fields using grouped 5-fold cross-validation in order to segment maize, weeds, and bare soil. We conducted an ablation study on a comprehensive baseline U-Net model (incorporating a ResNet-101 encoder, attention mechanisms, deep supervision, and physics-inspired regularization) by iteratively removing individual components to isolate their impact. The best overlap-based performance was obtained without deep supervision, while the lowest boundary error, measured by the 95th percentile Hausdorff Distance (HD95), was achieved without attention. We saw that using the Excess Green parameter made little or no difference compared to the standard method, while using a fixed threshold reconstruction always worsened performance. However, some examples sometimes contained isolated mislabeled pixels for weeds close to the image borders in the reference mask: these outliers can significantly increase the distance-based metrics, even when the prediction correctly ignores them. For weeds, overlap metrics and visual inspection were more informative than HD95, which was too sensitive to small annotation errors located near the edges of the image. Overall, our results suggest that a simpler segmentation pipeline is preferable for this dataset, that hard-threshold reconstruction is unreliable, and that the quality of annotation is a critical factor when interpreting distance-based metrics for weeds. The study does not propose a new architecture; instead, it provides a controlled ablation protocol and practical evidence that additional U-Net components may not improve small-data UAV RGB segmentation.
See at:
2026.fedcsis.org | CNR IRIS | CNR IRIS
2026
Conference article
Open Access
Fusing parametric and spatial constraints in continual learning
Ignesti Giacomo, D'Angelo Gennaro, Pratali Lorenza, Martinelli Massimo
The deployment of powerful foundation models in medical imaging is severely hampered by the fact that neural networks cannot learn sequentially from decentralised data with- out catastrophic forgetting. Early continual learning frameworks, such as Elastic Weight Consolidation (EWC), offer a parametric defence against forgetting but treat the network as a black box and neglect to preserve learned spatial representations. To solve this problem, we propose a novel continual learning framework that anchors EWC within a stable reconstruction-classification training paradigm using batch-trained prototypes. Our approach fundamentally constrains representation drift by enforcing spa- tial alignment between the network’s dynamically generated saliency maps and static visual class prototypes. Our approach combines weight consolidation with spatial reconstruction penal- ties to explicitly prevent the prior classification state-space from being distorted during novel task updates. Empirical evaluations show that the fusion not only retains high classification efficacy on historical tasks but also successfully maintains the network’s visual interpretability. Ultimately, this framework establishes a new direction for continual learning in medical imaging, shifting the focus from purely weight-based regularisation to the holistic preservation of both classifier stability and interpretable feature representations.
Ignesti Giacomo, D'Angelo Gennaro, Pratali Lorenza, Martinelli Massimo
The deployment of powerful foundation models in medical imaging is severely hampered by the fact that neural networks cannot learn sequentially from decentralised data with- out catastrophic forgetting. Early continual learning frameworks, such as Elastic Weight Consolidation (EWC), offer a parametric defence against forgetting but treat the network as a black box and neglect to preserve learned spatial representations. To solve this problem, we propose a novel continual learning framework that anchors EWC within a stable reconstruction-classification training paradigm using batch-trained prototypes. Our approach fundamentally constrains representation drift by enforcing spa- tial alignment between the network’s dynamically generated saliency maps and static visual class prototypes. Our approach combines weight consolidation with spatial reconstruction penal- ties to explicitly prevent the prior classification state-space from being distorted during novel task updates. Empirical evaluations show that the fusion not only retains high classification efficacy on historical tasks but also successfully maintains the network’s visual interpretability. Ultimately, this framework establishes a new direction for continual learning in medical imaging, shifting the focus from purely weight-based regularisation to the holistic preservation of both classifier stability and interpretable feature representations.
See at:
attend.ieee.org | CNR IRIS | CNR IRIS
2026
Conference article
Open Access
AI, complexity and research-driven innovation in food safety
Martinelli Massimo, Moroni Davide, Yingthawornsuk Thaweesak, Tuntrakool Sunti
In the twenty-first century, food safety is a challenge requiring increasingly capable systems for detecting hazards that often remain invisible to sight, including microbial contamination, oxidative deterioration, small changes in composition, adulterations hidden by globalized and fragmented supply chains. Ensuring the quality and integrity of food today requires both regulatory compliance and real-time monitoring. In this complex environment, Artificial Intelligence (AI) can support us effectively through controls based on past problems and errors and by predicting risks in advance. Methodologically, this article relies on a narrative review of peer-reviewed scientific works, synthesizing broad international research as well as key studies conducted at the Signals and Images Laboratory of the National Research Council (CNR-ISTI). The analysis covers advancements in precision agriculture related to wheat quality, image classification and segmentation, visible and hyperspectral assessment of milk quality, and the potential of extending these methodologies to rice bran oil authenticity. The results describe five main points: 1) the significant contribution of AI to prediction, anomaly detection, and multimodal information interpretation; 2) how modern methodologies use computer vision with hyperspectral imaging, spectroscopy, and machine learning methods; 3) concrete cases demonstrating that computational systems can detect information invisible to sight and unaided human perception, though they remain dependent on domain-specific data and expert validation; 4) AI allows for the analysis of huge amounts of data that a human could not handle quickly, however, it should not be used as an autonomous solution, but as a decision support tool that only works when integrated into a well-organized system of laboratory tests, sensors, and clear rules; 5) the final decision about using AI for new problems, must be made by expert researchers and scientists who have to study it to ensure it is seeing and learning information correctly, and that are essential to validate the integrity of automatic results.
Martinelli Massimo, Moroni Davide, Yingthawornsuk Thaweesak, Tuntrakool Sunti
In the twenty-first century, food safety is a challenge requiring increasingly capable systems for detecting hazards that often remain invisible to sight, including microbial contamination, oxidative deterioration, small changes in composition, adulterations hidden by globalized and fragmented supply chains. Ensuring the quality and integrity of food today requires both regulatory compliance and real-time monitoring. In this complex environment, Artificial Intelligence (AI) can support us effectively through controls based on past problems and errors and by predicting risks in advance. Methodologically, this article relies on a narrative review of peer-reviewed scientific works, synthesizing broad international research as well as key studies conducted at the Signals and Images Laboratory of the National Research Council (CNR-ISTI). The analysis covers advancements in precision agriculture related to wheat quality, image classification and segmentation, visible and hyperspectral assessment of milk quality, and the potential of extending these methodologies to rice bran oil authenticity. The results describe five main points: 1) the significant contribution of AI to prediction, anomaly detection, and multimodal information interpretation; 2) how modern methodologies use computer vision with hyperspectral imaging, spectroscopy, and machine learning methods; 3) concrete cases demonstrating that computational systems can detect information invisible to sight and unaided human perception, though they remain dependent on domain-specific data and expert validation; 4) AI allows for the analysis of huge amounts of data that a human could not handle quickly, however, it should not be used as an autonomous solution, but as a decision support tool that only works when integrated into a well-organized system of laboratory tests, sensors, and clear rules; 5) the final decision about using AI for new problems, must be made by expert researchers and scientists who have to study it to ensure it is seeing and learning information correctly, and that are essential to validate the integrity of automatic results.
See at:
CNR IRIS | sites.google.com | CNR IRIS
2025
Other
Open Access
SI-Lab Annual Research Report 2024
Awais Ch Muhammad, Baiamonte A., Benassi A., Berti A., Bertini G., Buongiorno R, Bulotta D., Cafiso M., Carboni A., Carloni G., Caudai C., Colantonio S., Conti F., Daoudagh S., Del Corso G., Fusco G., Galesi G., Germanese D., Gravili S., Ignesti G., Kuruoglu E. E., Lazzini G., Leone G. R., Leporini B., Magrini M., Martinelli M., Omrani Ali Reza, Pachetti E., Papini O., Paradisi P., Pardini F., Pascali M. A., Pieri G., Reggiannini M., Righi M., Salerno E., Salvetti O., Scozzari A., Sebastiani L., Straface S., Tampucci M., Tarabella L., Tonazzini A., Moroni D.
The Signal & Images Laboratory (SI-Lab) is an interdisciplinary research group in computer vision, signal analysis, intelligent vision systems and multimedia data understanding. It is part of the Institute of Information Science and Technologies (ISTI) of the National Research Council of Italy (CNR). This report accounts for the research activities of the Signal and Images Laboratory of the Institute of Information Science and Technologies during the year 2024.DOI: 10.32079/isti-ar-2025/002
Metrics:
Awais Ch Muhammad, Baiamonte A., Benassi A., Berti A., Bertini G., Buongiorno R, Bulotta D., Cafiso M., Carboni A., Carloni G., Caudai C., Colantonio S., Conti F., Daoudagh S., Del Corso G., Fusco G., Galesi G., Germanese D., Gravili S., Ignesti G., Kuruoglu E. E., Lazzini G., Leone G. R., Leporini B., Magrini M., Martinelli M., Omrani Ali Reza, Pachetti E., Papini O., Paradisi P., Pardini F., Pascali M. A., Pieri G., Reggiannini M., Righi M., Salerno E., Salvetti O., Scozzari A., Sebastiani L., Straface S., Tampucci M., Tarabella L., Tonazzini A., Moroni D.
The Signal & Images Laboratory (SI-Lab) is an interdisciplinary research group in computer vision, signal analysis, intelligent vision systems and multimedia data understanding. It is part of the Institute of Information Science and Technologies (ISTI) of the National Research Council of Italy (CNR). This report accounts for the research activities of the Signal and Images Laboratory of the Institute of Information Science and Technologies during the year 2024.DOI: 10.32079/isti-ar-2025/002
Metrics:
2025
Other
Restricted
Contents of the Digital agriculture for sustainable development. MASTER AGRITECH EU Course # 4
Martinelli M.
Contents of the Digital agriculture for sustainable development. MASTER AGRITECH EU CourseProject(s): Agritech
Martinelli M.
Contents of the Digital agriculture for sustainable development. MASTER AGRITECH EU CourseProject(s): Agritech
2025
Other
Open Access
Frostbite
Martinelli M.
Frostbite happens when your skin becomes frozen after being exposed to extremely cold temperatures.
Martinelli M.
Frostbite happens when your skin becomes frozen after being exposed to extremely cold temperatures.
2025
Other
Restricted
Contents of the Digital agriculture for sustainable development. MASTER AGRITECH EU Course # 3
Martinelli M.
Contents of the Digital agriculture for sustainable development. MASTER AGRITECH EU CourseProject(s): Agritech
Martinelli M.
Contents of the Digital agriculture for sustainable development. MASTER AGRITECH EU CourseProject(s): Agritech
2025
Other
Restricted
Contents of the Digital agriculture for sustainable development. MASTER AGRITECH EU Course # 6
Martinelli M.
Computer Vision & Applications in Agriculture Basic Techniques & Advanced ApplicationsProject(s): Agritech
Martinelli M.
Computer Vision & Applications in Agriculture Basic Techniques & Advanced ApplicationsProject(s): Agritech
2025
Other
Open Access
TiAssisto - Soluzioni per il monitoraggio clinico di pazienti in isolamento fiduciario a domicilio positivi al test per Covid-19 con associate o meno patologie croniche e situazioni di fragilità
Pratali L., Tomei A., Martinelli M.
Il poster illustra i principali obiettivi del progetto e i risultati ottenutiProject(s): TiAssisto
Pratali L., Tomei A., Martinelli M.
Il poster illustra i principali obiettivi del progetto e i risultati ottenutiProject(s): TiAssisto
2025
Other
Restricted
Contents of the Digital agriculture for sustainable development. MASTER AGRITECH EU Course # 5
Martinelli M.
Computer Vision & Applications in Agriculture - Basic Techniques & Advanced ApplicationsProject(s): Agritech
Martinelli M.
Computer Vision & Applications in Agriculture - Basic Techniques & Advanced ApplicationsProject(s): Agritech
2025
Other
Restricted
Improving weed control efficiency in maize fields: a methodological approach to site-specific weed management
Ercolini L., Grossi N., Martinelli M., Moroni D., Berton A., Silvestri N.
Methodologies to enhance precision agriculture.
Ercolini L., Grossi N., Martinelli M., Moroni D., Berton A., Silvestri N.
Methodologies to enhance precision agriculture.