Leader: Patrizia Ribino (ICAR-CNR); Other collaborator(s): M.Bochicchio (UNIBA), Giovanni Paragliola, Claudia Di Napoli, Luca Serino (ICAR-CNR), Maria Mannone (ICAR-CNR), Francesca Gasparini (UNIMIB); Davide Chicco (UNIMIB)
This task aims to provide methods for defining predictive models based on a subset of patient data and multicomponent intervention outcomes derived from data hubs and environmental IoT devices. Such models will be characterized by adaptive behaviour to deal with the requirements of customization to different patient settings.Methods will rely on AI-based software methodologies (i.e., Agent-based approaches) and machine learning algorithms (i.e., federated learning, reinforcement learning, deep learning). The development of methods will be carried out with an incremental approach where the design and prototyping phases rely on test data collected through IoT devices and software tools, and the validation phase relies on real data available from the data hubs.
Brief description of the activities and of the intermediate results
Several collaborative meetings for defining research trajectories have been organized.
- We conducted extensive research about public datasets that can be used for the exploitation of machine learning methods and the design of initial predictive. Datasets chosen for our initial study: OASIS 2, OASIS3 and ADNI.
- We initiated the conceptualization and implementation of ML models to detect the existence of dementia by leveraging both clinical and demographic information
- We have developed some predictive models based on main clustering algorithms (i.e., K-means, Mixture Gaussian Model, Hierarchical and Spectral clustering) and we have conducted some experiments on data coming from OASIS 2, containing data from Magnetic Resonance Imaging scans (i.e., eTIV, nWBV and ASF), and demographic assessments data such as the Mini-Mental State Exam (MMSE) scores, the education level, the socio-economic status of the subject, age, and gender.
- Moreover, we have conducted some research activities devoted to developing machine learning models for identifying individuals at high risk of developing heart problems. The present study introduces a novel approach that leverages Reinforcement Learning (RL) to enhance the performance of Artificial Neural Network (ANN) techniques for survival prediction by identifying the optimal configuration of model hyper-parameters.
- We are extending previous work on the OASIS 3 dataset by developing ML models for longitudinal data to evaluate cognitive decline progression
New line of research detected:
- Analysis of gene expression profiles in Alzheimer’s disease patients to advance early detection of AD
Main policy, industrial and scientific implications
Developing unsupervised longitudinal analysis to stratify the older population could unveil features and feature dependencies that can support prediction of disease progression.
We wrote several articles of the Quick Tips series to propose some guidelines, recommendations, best practices to follow in several aspects of bioinformatics and health informatics such as pangenomics, electrocardiogram signal processing, electroencephalogram, signal processing, fuzzy logic, and genetic signatures.
To achieve the objectives of the Machine Learning and AI task, weekly meetings are systematically convened with all participating researchers. Additionally, we participate in the monthly scheduled meetings of Work Package 4 (WP4), where we collaborate with researchers engaged in various tasks within WP4. This collaborative engagement facilitates the coordination of our efforts and ensures that we remain informed about the progress of the ongoing activities.
- Development of clustering models (both longitudinal and at one-time points) to identify possible predictors of dementia, mainly Alzheimer’s disease, and its progression.
We have developed a k-means-based multivariate longitudinal algorithm to cluster different trajectories of cognitive decline over time. Such an approach has been applied to electronic mental health records of Alzheimer’s patients, revealing that neuropsychiatric symptoms are predominantly associated with cognitive impairment. The preliminary results on the obtained clusters are promising, identifying subgroups of patients based on their longitudinal profiles in stable and progressing individuals. Such results will be presented at the 19th conference on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB2024) held in September in Benevento, Italy.
- Development of mathematical models of neurogenerative diseases, such as Alzheimer’s and Parkinson’s diseases, along with the development of ML approaches to understand how these diseases impact specific areas of the brain and its connectivity. In this activity, we also collaborate with researchers from Potsdam University, the University of Ca’ Foscari of Venice, and Università degli Studi di Milano.
We have developed a Brain-Network mathematical model for neurodegenerative disease. Since the human brain can be described as a multi-layer network, we modeled a neurological disease in terms of brain network alterations at different layers. Thus, we defined an operator acting on connectivity matrices and altering the weights of the connections. We conceptualized an operator, K, that acts on a healthy brain, produces a pattern of change typical for each disease, and describes the time evolution of a diseased brain. We applied our model to patients from the Parkinson’s Progression Markers Initiative (PPMI) and the Alzheimer’s Disease Neuroimaging Initiative (ADNI) datasets. We computed matrix forms of the K-operator comparing healthy control and diseased brain networks and gained insights into the disease evolution by computing the K-operator between brains from the baseline to different follow-ups. We compared our findings with the medical literature, confirming the relevance of our results. Some results of this activity have already been presented to the following conferences: (i) Dynamics Days of Europe - Bremen, July 29 - August 2, 2024; (ii) 26th International Conference on Computational Statistics COMPSTAT 2024, 27-30 August 2024, Giessen, Germany.
- Supervised machine learning for feature ranking on electronic health records to identify most predictive factors on several diseases.
We have developed a computational approach based on to Random Forests machine learning algorithm to detect the most relevant clinical factors for prognosis or diagnosis. We applied this approach to public datasets of electronic health records of patients with diabetes, sepsis and SIRS, neuroblastoma, and glioblastoma.
Weekly meetings were organized among researchers participating in Machine learning and AI activities. This periodicity made it possible to consolidate ongoing research, propose and develop new ideas, and plan scientific dissemination in top journals and participation in conferences focusing on relevant topics.
- Development of clustering models to identify possible predictors of dementia, mainly Alzheimer’s disease, and its progression. Since AD’s initial stages may not lead to a uniform cognitive decline across all cognitive domains, we conducted a study to evaluate the prognostic utility of individual domains of Clinical Dementia Rating (CDR) in predicting the progression of AD dementia over a five-year longitudinal period among an elderly cohort. Initially, a longitudinal-cluster analysis was conducted using five-point longitudinal data to categorize subjects into clusters based on the progression of CDR domains during the follow-up. Then, a statistical analysis was performed on the identified clusters to ascertain whether, at the baseline, patients exhibiting stability have different profiles about CDR domains compared to patients who converted to an AD during the whole follow-up period. Results show that the risk of AD progression was mainly related to problems with Orientation and Judgment at the baseline. Such results will be presented at the upcoming 14th International Conference on Current and Future Trends of Information and Communication Technologies in Healthcare (ICTH 2024) on October 28-30, 2024, in Leuven, Belgium.
- Development of mathematical models of neurogenerative diseases, such as Alzheimer’s and Parkinson’s diseases, along with the development of ML approaches to understand how these diseases impact specific areas of the brain and its connectivity. In this activity, we started collaborating with Università degli Studi di Milano on activities focused on the development and evaluation of a computational model aimed at operationalizing the mathematical models of neurodegenerative diseases under development. A first article reporting the preliminary results of this collaboration has been submitted to the 3rd AIxIA Workshop on Artificial Intelligence for Healthcare in Bolzano, 25-28 November.
The K-operator, studied and developed in the previous months, has been applied as a predictor for Alzheimer-Perusini’s disease and the results will be presented at the upcoming International Conference on Health and Social Care Information Systems and Technologies (HCist24), 13-15 November 2024, Madeira Portugal.
- Development of a federated longitudinal clustering approach for learning in distributed environments.
We developed a federated longitudinal clustering based on the K-means to take into consideration privacy and security concerns about patient health data. Federated Learning data systems are emerging to overcome the siloed nature of health data and the barriers to sharing it. While federated approaches have been extensively studied, especially in classification problems, clustering-oriented approaches are still relatively few and less widespread, both in formulating algorithms and in their application in eHealth domains. Hence, we developed a federated K-means-based approach for clustering tasks within the healthcare domain and exploring the impact of heterogeneous health data distributions. We have also evaluated the approach on several health-related datasets through comparison with the centralized version and by estimating the trade-off between privacy and performance. The preliminary findings of this activity will be presented to the upcoming 16th International Conference on Computation Theory and Applications NCTA24, 20-22 November Porto, Portugal.
As in the past months, we continue with weekly meetings with all participating researchers to achieve the Machine Learning and AI task objectives. Additionally, we participate in the monthly scheduled meetings of Work Package 4 (WP4), where we collaborate with researchers engaged in various tasks within WP4. This collaborative engagement facilitates the coordination of our efforts and ensures that we remain informed about the progress of the ongoing activities.
- Speech emotion recognition in an elderly population from natural conversation. Progress of the activity: 30%. Many older adults live alone in their own homes, usually isolated because of health problems or major life events that threaten to limit their social interaction. The negative impact of this isolation on mental and physical health leads to the need to develop systems that can monitor and interact naturally with older adults during their daily lives. In particular, Social Robots, as Companion Type Robots, are being developed specifically to provide companionship and cognitive support to frail people to ensure their health and psychological well-being. Such robots must be able to interact with people in a natural and realistic way, inferring their emotions and adapting their behaviour accordingly. This led in recent years to a growing interest in Speech Emotion Recognition (SER). Most of the SER classification models are based on specific domains and are not easily generalizable to other situations or use cases. For instance, most of the datasets available in the literature are acted utterances collected from English adults and thus not easily generalizable to other languages or ages. In this context, defining a SER system that can be easily generalised to new subjects or languages has become a topic of relevant importance. The main aim of our preliminary research is to analyze the challenges and limitations of using acted datasets to define a general model that could be adapted in natural conversation with italian older people.
- Development of clustering models to identify possible predictors of dementia, mainly Alzheimer’s disease, and its progression. Progress of the activity: 90%.
- In the last few months, we continued to perform longitudinal studies on electronic mental health records for understanding predictive markers of cognitive decline, especially for Alzheimer’s disease. Mainly, we have extended the functionality of the original k-means-based algorithm, previously developed, to include feature selection capabilities, further enhancing its ability to handle complex datasets. This functionality helps identify the most relevant variables for the analysis. This is particularly valuable when working with high-dimensional data, where not all features may contribute meaningfully to the clustering process. By automatically selecting a subset of features, the algorithm reduces the noise in the data, improving clustering performance and interpretability. Hence, we applied such an extended version of the algorithm to electronic mental health records derived from the OASIS-3 dataset. The study revealed an important role of some neuropsychological factors. Such first results have been reported in a paper submitted to the Workshop on Scaling Up Care for Older Adults - Scale-IT-up @Biostec2025, which will be held in February in Porto, Portogallo.
- Further analysis by adopting such a longitudinal algorithm for studying Alzheimer’s progression has identified a significant profile compatible with a syndrome known as Mild Behavioral Impairment (MBI), displaying behavioral manifestations of individuals that may precede the cognitive symptoms typically observed in AD patients. Such findings have been reported in an article submitted to the BioDataMining journal.
- Moreover, we also conducted a longitudinal study to understand the diverse trajectories of septic patients better. Sepsis, a life-threatening condition with complex and dynamic progression, demands timely and personalized treatment, particularly in vulnerable populations such as the elderly, where it can lead to severe outcomes. We analyzed sepsis-related electronic health records (EHRs), which offer rich, time-series data encompassing laboratory results, patient demographics, and underlying health conditions to gain insights into this. Our findings revealed distinct sepsis phenotypes, highlighting the variations in disease progression. We identified the Thrombin-Antigen complex and International Normalized Ratio (INR) as significant predictors of poor outcomes in septic patients. Such findings were presented at the 3rd AIxIA Workshop on Artificial Intelligence for Healthcare (HC@AIxIA2024) held on November 2024 in Bolzano, Italy. The paper that describes these findings has been obtained the Best Short Paper and Poster Award: P. Ribino, M. Mannone, C. Di Napoli, G. Paragliola, D. Chicco and F. Gasparini: Analyzing trajectories of clinical markers in patients with sepsis through multivariate longitudinal clustering, 3rd AIxIA Workshop on Artificial Intelligence For Healthcare (HC@AIxIA 2024) co-located with the 23rd International Conference of the Italian Association for Artificial Intelligence (AIxIA 2024),Bolzano, Italy, 27-28 November 2024.
- Finally, as a further result of this activity, we have consolidated the algorithm development by providing a Python-based tool named LongitProgression to be used by the research community. This tool provides researchers with an intuitive solution for analyzing complex datasets with repeated measurements over time involving multiple variables. The tool integrates time-series analysis with clustering techniques, capturing dynamic trends and variability within longitudinal datasets. It provides insights into temporal correlations across multiple variables, facilitating the identification of meaningful subgroups in heterogeneous populations. Mainly, we have designed and implemented an intuitive graphical interface that simplifies the user experience, allowing for seamless data preprocessing, clustering execution, and result analysis. This feature significantly enhances usability, enabling researchers with minimal programming experience to interact with the tool efficiently. Additionally, the tool includes built-in normalization functions like z-score normalization and Min-Max scaling to ensure the data is standardized, preventing any variable from disproportionately influencing the clustering results. LongitProgression leverages the k-means algorithm, with users able to customize parameters such as the number of clusters, centroid initialization methods, and the maximum number of iterations. The tool also supports specialized temporal distance metrics, such as Dynamic Time Warping (DTW) and softDTW, to account for the time-series nature of the data, making the clustering more accurate for longitudinal analysis. After clustering, the tool was designed to provide several functionalities for evaluating the quality of the results. The tool also offers visualization options to track the evolution of cluster centroids over time, aiding in interpreting the data's temporal dynamics. Furthermore, statistical tests like ANOVA and Kruskal-Wallis can be applied to each time point of the clusters to examine group differences. By releasing LongitProgression under the GNU General Public License on GitHub, we aim to provide the research community with an open-source, customizable, and accessible tool for longitudinal data analysis, facilitating progress in diverse fields not only in the healthcare domain but also in social sciences, and economics. The tool is accessible via the following link: https://github.com/PatriziaRibino/LongitProgression, and its implementation is currently being reviewed by the SoftwareX journal.
- Development of mathematical models of neurogenerative diseases along with the development of ML approaches to understand how these diseases impact specific areas of the brain and its connectivity. In this activity, we also collaborate with researchers from Potsdam University, the University of Ca’ Foscari of Venice, and the University of Milan. Progress of the activity: 60%.
- In the last few months, we continued to develop a Brain-Network mathematical model for neurodegenerative disease by conceptualizing an operator, K, by working on fMRI data. We extended and modified the versions of our model to patients with Alzheimer’s disease and other neurodegenerative diseases. The results of this research have been presented at the 3rd AIxIA Workshop on Artificial Intelligence for Healthcare (HC@AIxIA 2024) held on November 2024 in Bolzano, and at the NeuroMI 2024 International Meeting: “Brain Health and Prevention of Cognitive Decline” held on October 2024 in Milan, Italy. Moreover, a paper reporting the model and results of Alzheimer’s disease has been recently published in the Biomedical Signal Processing and Control 2025.
- In addition to functional magnetic resonance imaging (fMRI) data, we are also considering electroencephalograms (EEG) as new type of input data for the developed model. A preliminary study presents an approximation of the temporal evolution of operator K based on the segmentation of an invasive pre-surgical EEG from a patient with epileptic seizures. The work is currently under review for the ASPAI '25 conference in Austria. We plan to extend the study to a journal publication, including training neural networks to improve the approximation of the temporal evolution.
- We are also continuing our collaboration with the University of Milan on activities focused on developing and evaluating a computational model to operationalize the proposed mathematical models of neurodegenerative diseases under development. Maria Mannone, the TD assigned to the project, is supervising a Master's thesis in Physics at the University of Milan. The thesis focuses on the analysis of the K operator in terms of eigenvalues and eigenvectors, as well as the approximation of its temporal evolution, both for the null model and for real data.
- Development of a federated longitudinal clustering approach for learning in distributed environments. Progress of the activity: 60%We have also continued our activities on federated learning models in the last months. The findings of this activity were presented at the 16th International Conference on Computation Theory and Applications NCTA24, 20-22 November Porto, Portugal. Such a paper has been awarded as Best Poster Candidate at the conference.
- Analysis of Blood Transcriptomic Changes for MCI Stratification. Progress of the activity: 30%. In the last months, we have focused on a particular stage of Alzheimer's disease progression, the MCI. Mild cognitive impairment (MCI) is an intermediate stage in the trajectory from normal cognition to dementia. The concept of MCI is highly significant and important to the field of aging and dementia. Indeed, patients with MCI are at high risk of progression to dementia. However, among them, there are stable MCI and progressive MCI. Hence, identifying possible predictors for the stability or instability of MCI is a key challenge in the study of Alzheimer's disease progression. Hence, we have started an activity in collaboration with the Computational Data Science group (CDS@ICAR) with the aim of combining blood transcriptomic data with clinical (e.g., cognitive assessments, neuroimaging) and demographic data (e.g., age, sex, genetic factors) to improve MCI stratification and understand underlying heterogeneity in MCI patients. The preliminary results of this activity were presented in November at the 19th annual edition of the Bioinformatics and Computational Biology Conference (BBCC) 2024.
- Development of a telecommunication-based model for studying brain network functionality under neurodegenerative disease. Progress of the activity: 50%. Finally, in collaboration with the Ca Foscari University of Venice, we have started a new activity to explore a model based on telecommunication to study brain network functionality. The brain network is modeled as a telecommunication channel, with messages sent and received between brain areas, and the impact of disease is analyzed in terms of damage to the network. This research branch has already led to a conference article at the 15th IFIP Wireless and Mobile Networking Conference (WMNC), Venice, and a journal submission is in preparation.
Year 2025:
Journals
- Patrizia Ribino, Claudia Di Napoli, Giovanni Paragliola, Davide Chicco, and Francesca Gasparini Multivariate longitudinal clustering reveals neuropsychological factors as dementia predictors in an Alzheimer’s disease progression study. BIODataMining (Under 2° round of revision) 2025. Status: Under 2nd round of revision.
- Patrizia Ribino, Claudia Di Napoli, Giovanni Paragliola, Maria Mannone, LongitProgression: a Python tool for studying factors of disease progression through multivariate longitudinal clustering, SoftwareX. Status: Under revision.
- Chicco, D., Fabris, A., & Jurman, G. (2025). The Venus score for the assessment of the quality and trustworthiness of biomedical datasets. BioData Mining, 18(1), 1.
- Oneto, L., & Chicco, D. (2025). Eight quick tips for biologically and medically informed machine learning. PLOS Computational Biology, 21(1), e1012711.
- Mannone, M., Marwan, N., Fazio, P., & Ribino, P. (2025). Limbic and cerebellar effects in Alzheimer-Perusini’s disease: A physics-inspired approach. Biomedical Signal Processing and Control, 103, 107355.
Conferences
- Patrizia Ribino, Giovanni Paragliola, Claudia Di Napoli, Luca Serino, Davide Chicco and Francesca Gasparini, Longitudinal analysis of disease progression in the elderly: an approach to mitigate the burden of frailty, functional and cognitive decline. Workshop on
Scaling Up Care for Older Adults - Scale-IT-up @Biostec 2025. Status: Accepted
- Maria Mannone, Patrizia Ribino, Aurora Saibene, Peppino Fazio, Sofia Fazio, Francesca Gasparini, Marco Gherardi and Norbert Marwan, Computing the Time-Dependent Krankheit-Operator in Epilepsy from ECoG: a Case Study. 7th International Conference on Advances in Signal Processing and Artificial Intelligence (ASPAI' 2025). Status: Submitted
Dissemination:
Year 2024:
Journal
- Cerono, G., Melaiu, O., & Chicco, D. (2024). Clinical feature ranking based on ensemble machine learning reveals top survival factors for glioblastoma multiforme. Journal of Healthcare Informatics Research, 8(1), 1-18.
- Cerono, G., & Chicco, D. (2024). Ensemble machine learning reveals key features for diabetes duration from electronic health records. PeerJ Computer Science, 10, e1896.
- Mollura, M., Chicco, D., Paglialonga, A., & Barbieri, R. (2024). Identifying prognostic factors for survival in intensive care unit patients with SIRS or sepsis by machine learning analysis on electronic health records. PLOS Digital Health, 3(3), e0000459.
- Bonnici, V., & Chicco, D. (2024). Seven quick tips for gene-focused computational pangenomic analysis. BioData Mining, 17(1), 28.
- Cisotto, G., & Chicco, D. (2024). Ten quick tips for clinical electroencephalographic (EEG) data acquisition and signal processing. PeerJ Computer Science, 10, e2256.
- Chicco, D., Karaiskou, A. I., & De Vos, M. (2024). Ten quick tips for electrocardiogram (ECG) signal processing. PeerJ Computer Science, 10, e2295.
- Liu, W., He, H., & Chicco, D. (2024). Gene signatures for cancer research: A 25-year retrospective and future avenues. PLOS Computational Biology, 20(10), e1012512.
- Mannone, M., Fazio, P., Kurths, J. et al. A brain-network operator for modeling disease: a first data-based application for Parkinson’s disease. Eur. Phys. J. Spec. Top. (2024).
- Mannone, M., Fazio, P., Ribino, P., & Marwan, N. (2024). On disease and healing: a theoretical sketch. Frontiers in Applied Mathematics and Statistics, 10, 1468556.
- M. Mannone, P. Fazio, N. Marwan, P. Ribino. Limbic and cerebellar effects in Alzheimer-Perusini's disease: a physics-inspired approach. Biomedical signal processing and control. Status: Under review
Conferences
- P. Fazio, M. Mannone, B. Marchiori, P. Riello, N. Marwan and M. Voznak. A Biological Fading Channel Model for Neural Communication Networks: a Preliminary Sketch. 2024 15th IFIP Wireless and Mobile Networking Conference (WMNC), Venice, 2024, pp. 75-79.
- Antonelli Laura, Di Napoli Claudia, Maddalena Lucia, Paragliola Giovanni, Ribino, Patrizia, Serino Luca, Granata, Ilaria. Clustering-based stratification of mild cognitive impairment: Insights from blood transcriptomic data, F1000Research 2024,15,1473 (poster) 2024, https://doi.org/10.7490/f1000research.1120051.1.
- M. Mannone, P. Fazio, N. Marwan, P. Ribino. A Dialogue between Formalization and Data: Neurological Disease as the Action of a Mathematical Operator. NeuroMI 2024, Milano, October 2024, (poster).
- Ribino, P., Di Napoli, C., Paragliola, G., & Serino, L. (2024). Hyper-Parameter Optimization through Reinforcement Learning for Survival Prediction of Patients with Heart Failure. Procedia Computer Science, 239, 1754-1761.
- Grossi, A., Gasparini, F., Saibene, A. (2024). On the Exploitation of CEEMDAN for PPG Synthetic Data Generation. In: Bochicchio, M., Siciliano, P., Monteriù, A., Bettelli, A., De Fano, D. (eds) Ambient Assisted Living. ForItAAL 2023. Lecture Notes in Bioengineering. Springer, Cham.
- Grossi, A., Gasparini, F. (2024). SER_AMPEL: A Multi-source Dataset for Speech Emotion Recognition of Italian Older Adults. In: Bochicchio, M., Siciliano, P., Monteriù, A., Bettelli, A., De Fano, D. (eds) Ambient Assisted Living. ForItAAL 2023. Lecture Notes in Bioengineering. Springer, Cham.
- Paragliola G., Ribino P. and Mannone M. (2024). A Federated K-Means-Based Approach in eHealth Domains with Heterogeneous Data Distributions. In Proceedings of the 16th International Joint Conference on Computational Intelligence - Volume 1: NCTA; ISBN 978-989-758-721-4, SciTePress, pages 550-559.
- M. Mannone, P. Fazio, N. Marwan. An Operator Acting on the Brain Network and Provoking Disease: A Conceptual Model and a First Data-Based Application, Dynamics Days of Europe. Bremen, July 29 - August 2, 2024. https://dynamicsdays.eu/bremen2024/Program/bookofabstracts.pdf.
- M. Mannone, P. Fazio, N. Marwan, P. Ribino. Brain-Network Mathematical Modeling for Neurodegenerative Disease. COMPSTAT 24. Gießen, Germany August 27 – 30, 2024. http://www.compstat2024.org/docs/ COMPSTAT2024_BoA.pdf?20240730003320.
- M. Mannone, P. Fazio, N. Marwan, P. Ribino, $K$-operator as a predictor for Alzheimer-Perusini’s disease, HCist - International Conference on Health and Social Care Information Systems and Technologies. 13 - 15 November 2024. Madeira, Portugal.
- Ribino, P., Paragliola, G., Di Napoli, C., Mannone, M., Chicco, D., & Gasparini, F. (2024). Clustering of longitudinal Clinical Dementia Rating data to identify predictors of Alzheimer's disease progression. Procedia Computer Science, 251, 326-333.
- Patrizia Ribino, Claudia Di Napoli, Giovanni Paragliola, Luca Serino, Davide Chicco and Francesca Gasparini, Longitudinal clustering on electronic mental health records reveals meaningful groups of disease trajectories, 19th conference on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB 24) September 4 - 6 Benevento 2024, Italy.
- Fazio, S., Ribino, P., Gasparini, F., Marwan, N., Fazio, P., Gherardi, M., & Mannone, M. (2024). A physics-based view of brain-network alteration in neurological disease. In Proceedings of the 3rd AIxIA Workshop on Artificial Intelligence For Healthcare (HC@AIxIA 2024) co-located with the 23rd International Conference of the Italian Association for Artificial Intelligence (AIxIA 2024), Bolzano, Italy, 27-28 November 2024, CEUR-WS.org, Vol. 3880, pp. 169-181
- Ribino, P., Mannone, M., Di Napoli, C., Paragliola, G., Chicco, D., & Gasparini, F. (2024). Analyzing trajectories of clinical markers in patients with sepsis through multivariate longitudinal clustering. In Proceedings of the 3rd AIxIA Workshop on Artificial Intelligence For Healthcare (HC@AIxIA 2024) co-located with the 23rd International Conference of the Italian Association for Artificial Intelligence (AIxIA 2024),Bolzano, Italy, 27-28 November 2024, CEUR-WS.org, Vol. 3880, pp. 247-256.
Dissemination:
- June 13th, first Italian Workshop on computational methods fro mental health and well being, Department of Informatics, Systems and Communication of the University of Milano-Bicocca.
- Francesca Gasparini, “AI personalized models based on subjective data collected from older people: case studies in the field affective computing and remote monitoring” at ADVANCES IN ARTIFICIAL INTELLIGENCE, Lake Como School of Advanced Studies, September 23-27, 2024, Villa del Grumello (Como, ITALY).
Year 2023:
Journal
- Gasparini, F., Grossi, A., Giltri, M., Nishinari, K., & Bandini, S. (2023). Behavior and task classification using wearable sensor data: A study across different ages. Sensors, 23(6), 3225.
- Chicco, D., Haupt, R., Garaventa, A., Uva, P., Luksch, R., & Cangelosi, D. (2023). Computational intelligence analysis of high-risk neuroblastoma patient health records reveals time to maximum response as one of the most relevant factors for outcome prediction. European Journal of Cancer, 193, 113291.
- Cerono, G., Melaiu, O., & Chicco, D. (2024). Clinical feature ranking based on ensemble machine learning reveals top survival factors for glioblastoma multiforme. Journal of Healthcare Informatics Research, 8(1), 1-18.
- Chicco, D., Spolaor, S., & Nobile, M. S. (2023). Ten quick tips for fuzzy logic modeling of biomedical systems. PLoS Computational Biology, 19(12), e1011700.
Conferences
- Grossi, A., Fratti, G., & Gasparini, F. (2023). A computational framework for speech emotion recognition in case of multisource data, Proceedings of the 4th Italian Workshop on Artificial Intelligence for an Ageing Society co-located with 22nd International Conference of the Italian Association for Artificial Intelligence, Rome, Italy, Vol-3623, pp.113-126, CEUR
- Ribino, P., Di Napoli, C., Paragliola, G., Serino, L., Gasparini, F., & Chicco, D. (2023). Exploratory analysis of longitudinal data of patients with dementia through unsupervised techniques. Proceedings of the 4th Italian Workshop on Artificial Intelligence for an Ageing Society co-located with 22nd International Conference of the Italian Association for Artificial Intelligence, Rome, Italy, Vol-3623, pp. 67-87, CEUR.
Dissemination:
- 4th Italian Workshop on Artificial Intelligence for an Ageing Society co-located with 22nd International Conference of the Italian Association for Artificial Intelligence, Rome, Italy