In the last decades, the remarkable impact achieved by Artificial Intelligence (AI) in business and industry has not been mirrored in critical real-world applications. The industrial diffusion of AI in healthcare is facing some resistance due to the lack of uniform legal frameworks and general scepticism among society and medical personnel. This paper proposes a multidisciplinary approach to fill the gap between the theoretical AI-based framework and real clinical practice, tailored to the problem of Posterior Urethral Valves (PUVs) diagnosis in paediatric patients. The multidisciplinary core of the work allows tackling the problem not only under the technical lens, but also from a clinical and industrial perspective: through the adoption of classifier composition mechanisms, this study presents the lessons learned in developing a reliable PUV classifier, as well as in its empirical assessment against real-world data and within a structured diagnostic process. The main contribution of this study is the design of a clinical decision support system for medical experts, which evaluates the behaviour of the model clinically and validates the extracted rules using explainability techniques on real-world data. AI classifiers leveraging vertical training of specialised models were adopted, achieving an overall accuracy of 70 %.

The availability of clean and safe drinking water is essential for public health and sustainable development. This study uses a robust machine learning-based methodology to predict water potability using genuine water quality parameters. The methodology adopted consists of intelligent feature selection along with comparative evaluation of multiple classifiers. As a novelty, this study employs a hybrid resampling technique (Synthetic Minority Over-sampling Technique combined with Edited Nearest Neighbors (SMOTEENN)) that integrates Synthetic Minority Over-sampling Technique (SMOTE) and Edited Nearest Neighbors (ENN) techniques that improve class balance based on noise reduction and oversampling. The experimental results on the public water quality dataset show strong and well-rounded performance across many performance metrics, with most models achieving high scores based on general predictive quality. Further, visualizations such as Receiver Operating Characteristic (ROC) curves and feature importance plots support interpretability and offer insights regarding model behaviour. The main contribution of this paper is in finding cost-effective and scalable solutions for smart water quality monitoring and decision support in public health and environmental safety systems: the proposed approach performs better with the respect to the current state of the art.

Software quality is a critical factor for the overall success and acceptability of software products. To evaluate software quality effectively, standardized models consider both major characteristics and sub-features of the software. As these sub-features often conflict with one another, a crisp and exact approach is neither feasible nor effective. The novelty of this study is to evaluate the quality of Microsoft Word using a hierarchical three-level model based on a Fuzzy Inference System (FIS) aligned with the International Organization for Standardization (ISO)/International Electrotechnical Commission (IEC) 25010 standard. Expert judgments were employed to determine weights for the relative importance of each quality attribute to enhance the realism and accuracy of the assessment. The research establishes that software quality is a hierarchical and dynamic concept, whereby the quality of components across various phases of development directly impacts the quality of the final product and stresses the role played by formal quality evaluation models in guiding the growth and selection of effective and user-oriented software solutions.

The growing interest in artificial intelligence applications in real clinical practice has made the development of personalised medicine possible. Digital Twins support diagnosis and treatment by providing an overall view of patients’ health status. The definition of a patient’s digital model requires the integration of different sources of information that contribute to a holistic view of the subject. In this work, we propose a preliminary step toward the definition of paediatric digital twins, providing a tool for Region of Interest identification on X-ray images. In detail, the proposed tool, STELLA (Segmentation Tool for Enhanced Localisation and Labelling of diagnostic Areas), is adopted to automatically detect the bladder and urethra regions on the images obtained from the cystourethrography exam. STELLA pipeline is based on Segment Anything Model (SAM) for the segmentation task and Resnet-18 for masks classification: SAM is leveraged for automatic masks generation and ResNet18 is trained on labelled masks for Regions of Interest classification. This is framed in a larger context, whose aim is to support posterior urethral valves diagnosis.

Recently, the field of anesthesiology has increasingly recognized the need for personalized medicine, aiming to tailor drug administration based on the unique physiological and clinical profiles of individual patients. This approach is particularly crucial in the administration of anaesthetic drugs, where inter-individual variability in response can significantly affect both the efficacy and safety of treatment. The main focus of this paper is on the relationship between the healthcare domain and innovative Machine Learning technologies. The specific case study analysed concerns the implementation of a predictive Bayesian Network (BN) model for the inductive administration of Propofol, an anaesthetic drug. The available data is taken from the PhysioNet platform and it relates to a study conducted on nine healthy volunteers who underwent drug administration for approximately three hours, measuring their vital parameters. The results indicate that the choice of the Bayesian Network (BN) formalism is highly suitable for the analysed case study. In conclusion, it is hypothesized that an analysis focused on patient-specific characteristics, such as gender, age, and medical history, could significantly improve the model’s accuracy.

In the medical domain, the early detection and monitoring of specific diseases require both accuracy and interpretability to support clinical decisions. Human digital twin systems are increasingly used in this context, but their adoption often requires data-intensive process for the learning tasks and a high degree of explainability to ensure clinical reliability. To address these challenges, we propose a pipeline that prioritises lightweight and explainability, to extract actionable knowledge from patient data in terms of rules. The approach follows a traditional Machine Learning methodology, including data pre-processing, followed by the construction and validation of a classification model. A rule extraction phase is then introduced to make the classifier’s decision process interpretable. The reliability of the pipeline was evaluated by extracting decision rules for the detection of kidney damage in patients with Congenital Solitary Functioning Kidney. Through the analysis of patient data and the use of a Random Forest classifier, key clinical parameters (e.g., creatinine levels, Holter monitor measurements, and kidney volume) were identified as fundamental to support accurate and reliable diagnoses in clinical practice.

Modern public and private healthcare structures are facing the problem of improving the quality of patient health without increasing costs. Smart healthcare is currently transforming the traditional medical practices, resulting in a more efficient, convenient and personalized healthcare. In this paper, a solution for a fair usage of economic resources is proposed: the FIDCARE approach. Based on a flexible software architecture, with the capability to be extended by external “plugins”, the FIDCARE platform conjugates both the needs. IoT technologies and AI algorithms are at the basis of the entire platform to enable a proper level of flexibility. The paper presents the approach with the case study of oncological therapy.

The goal of this research is to create an intelligent system to assess the manufacturing system’s level of risk for wheat four products. Five Fuzzy Inference Systems (FISs) are arranged in two layers of the model to assess the risk associated with a system that produces wheat four products. There are four FISs with three criteria (Occurrence, Severity, and Detectability) in the model’s first layer. The final input for the manufacturing system will be determined from every physical, chemical, biological, and environmental failure. The suggested model, which is based on Mamdani FISs, ranks the manufacturing systems for wheat four products according to their performance. A four-step approach (i.e., eliciting hazard information for experts, fuzzification, inference, and defuzzification) brings to an evaluation of the final risk level in a real-world wheat four manufacturing system to 22.5%, which shows a fair situation, and it represents a manufacturing system with a high-risk level.

In healthcare, Digital Twins (DTs) promise to personalise treatment plans, simulate surgeries, and forecast individual responses to particular therapies. By adopting Machine Learning methodologies, it is possible to figure out some insights hidden among the features for enhancing medical diagnosis. Our contribution leverages the role of intraoperative ultrasound in liver surgery in building a Bayesian Network (BN) model for enabling the early localisation of hepatic cancer. Under this premise, we aim to determine how a possible diagnosis error could be affected by factors such as age, gender, and before-surgery treatment. The mean to this objective is the construction of a BN model by using both an explicit top-down approach and parameter learning approaches. This is the first step toward the DT definition of a patient affected by hepatic cancer in charge of continuously monitoring the health status.

Human Digital Twin is an emerging technology that could revolutionize the current healthcare system by enabling the delivery of Personalized Health Services through the use of tools such as Artificial intelligence. However, the considerable complexity of the structure of the human body, brought about by continuous molecular and physiological changes, makes it extremely difficult to process medical data extracted by Artificial intelligence techniques. The latter requires a large amount of data for reliable performance, which is often difficult to obtain due to limited quality and availability. In this paper, we propose a methodology to generate Artificial medical data. In detail, we focus on generating Artificial voice signals. The analysis of voice recordings is fundamental to diagnose specific pneumo-articulatory apparatus diseases, such as dysphonia. The generative neural network employed is based on the WaveNet model, due to its autoregressive sampling, which enables generating recordings of variable length. We propose a setup which enables to generate Artificial samples of required sex and pathology to balance and augment the dataset using only one generative network. The quality of the generative network is assessed by balancing the training dataset by generated data and training a convolutional classifier, which is tested on a dataset which was not introduced to the generative network during training. We achieved reasonable improvements in classification accuracy, particularly for the under-represented sex in terms of accuracy, arguing that this approach is worthy of future research.

Recently, Large Language Models (LLMs) have seen some impressive leaps, achieving the ability to accomplish several tasks, from text completion to powerful chatbots. The great variety of available LLMs and the fast pace of technological innovations in this field, is making LLM assessment a hard task to accomplish: understanding not only what such a kind of systems generate but also which is the quality of their results is of a paramount importance. Generally, the quality of a synthetically generated object could refer to the reliability of the content, to the lexical variety or coherence of the text. Regarding the quality of text generation, an aspect that up to now has not been adequately discussed is concerning the readability of textual artefacts. This work focuses on the latter aspect, proposing a set of experiments aiming to better understanding and evaluating the degree of readability of texts automatically generated by an LLM. The analysis is performed through an empirical study based on: considering a subset of five pre-trained LLMs; considering a pool of English text generation tasks, with increasing difficulty, assigned to each of the models; and, computing a set of the most popular readability indexes available from the computational linguistics literature. Readability indexes will be computed for each model to provide a first perspective of the readability of textual contents artificially generated can vary among different models and under different requirements of the users. The results obtained by evaluating and comparing different models provide interesting insights, especially into the responsible use of these tools by both beginners and not overly experienced practitioners.

Human-Machine Interfaces (HMIs) are getting more and more important in a hyper-connected society. Traditional HMIs are built considering cognitive features while emotional ones are often neglected, bringing sometimes such interfaces to misuse. As a part of a long run research, oriented to the definition of an HMI engineering approach, this paper concretely proposes a method to build an emotional-aware explicit model of the user starting from the behaviour of the human with a virtual agent. The paper also proposes an instance of this model inference process in voice assistants in an automatic depression context, which can constitute the core phase to realize a Human Digital Twin of a patient. The case study generated a model composed of Fluid Stochastic Petri Net sub-models, achieved after the data analysis by a Support Vector Machine.

In a world whose attention to environmental and health problems is very high, the issue of properly managing nuclear waste is of a primary importance. Information and Communication Technologies have the due to support the definition of the next-generation plants for temporary storage of such wasting materials. This paper investigates on the adoption of one of the most cutting-edge techniques in computer science and engineering, i.e. Digital Twins, with the combination of other modern methods and technologies as Internet of Things, model-based and data-driven approaches. The result is the definition of a methodology able to support the construction of risk-aware facilities for storing nuclear waste.

Currently, research in Artificial Intelligence, both in Machine Learning and Deep Learning, paves the way for promising innovations in several areas. In healthcare, especially, where large amounts of quantitative and qualitative data are transferred to support studies and early diagnosis and monitoring of any diseases, potential security and privacy issues cannot be underestimated. Federated learning is an approach where privacy issues related to sensitive data management can be significantly reduced, due to the possibility to train algorithms without exchanging data. The main idea behind this approach is that learning models can be trained in a distributed way, where multiple devices or servers with decentralized data samples can provide their contributions without having to exchange their local data. Recent studies provided evidence that prototypes trained by adopting Federated Learning strategies are able to achieve reliable performance, thus by generating robust models without sharing data and, consequently, limiting the impact on security and privacy. This work propose a literature overview of Federated Learning approaches and systems, focusing on its application for healthcare. The main challenges, implications, issues and potentials of this approach in the healthcare are outlined.

Edge Analytics and Artificial Intelligence are important features of the current smart connected living community. In a society where people, homes, cities, and workplaces are simultaneously connected through various devices, primarily through mobile devices, a considerable amount of data is exchanged, and the processing and storage of these data are laborious and difficult tasks. Edge Analytics allows the collection and analysis of such data on mobile devices, such as smartphones and tablets, without involving any cloud-centred architecture that cannot guarantee real-time responsiveness. Meanwhile, Artificial Intelligence techniques can constitute a valid instrument to process data, limiting the computation time, and optimising decisional processes and predictions in several sectors, such as healthcare. Within this field, in this article, an approach able to evaluate the voice quality condition is proposed. A fully automatic algorithm, based on Deep Learning, classifies a voice as healthy or pathological by analysing spectrogram images extracted by means of the recording of vowel /a/, in compliance with the traditional medical protocol. A light Convolutional Neural Network is embedded in a mobile health application in order to provide an instrument capable of assessing voice disorders in a fast, easy, and portable way. Thus, a straightforward mobile device becomes a screening tool useful for the early diagnosis, monitoring, and treatment of voice disorders. The proposed approach has been tested on a broad set of voice samples, not limited to the most common voice diseases but including all the pathologies present in three different databases achieving F1-scores, over the testing set, equal to 80%, 90%, and 73%. Although the proposed network consists of a reduced number of layers, the results are very competitive compared to those of other “cutting edge” approaches constructed using more complex neural networks, and compared to the classic deep neural networks, for example, VGG-16 and ResNet-50.

The growing number of people suffering from depression makes it increasingly necessary to find new approaches able to support medical experts in its diagnosis. The early detection of depressive symptoms is crucial in limiting the co-occurrence of associated behavioural disorders such as psycho-motor retardation symptoms and social withdrawal. Therefore, automatic detection systems represent promising solutions not only for supporting the early diagnosis of the disease but also for monitoring patient’s health status, thus improving both the quality of the care process and life quality of patients. At the light of these considerations, this paper proposes an automatic system exploiting a machine learning algorithm, to distinguish among depressed and healthy subjects through the analysis of selected acoustic features extracted from spontaneous speech narratives produced by healthy and depressed subjects. The proposed system achieves a classification accuracy of about 85%, proving to be a promising solution for supporting the diagnosis of depression in real-time in a reliable, fast, inexpensive and non-intrusive ways.

Federated Learning is a valuable instrument for building AI-based systems that preserve the privacy and security of sensitive data, based on the main concept of shifting no more the data to the edges but moving computations to data, avoiding the collection, sharing, and use of such data by third parties. More robust federated learning systems should be able of preventing malicious inference over both data exchanged during training and the final trained model while ensuring the resulting model also has acceptable predictive accuracy. This study proposes a preliminary analysis to investigate and evaluate the effectiveness and efficiency of a federated approach to ensure valid classification accuracy and data security. A real case study from the ANDROIDS project, concerning the application of machine learning-based systems for supporting mental-health disorders detection, was considered. Large amounts of sensitive patient information are collected, which must be obfuscated or anonymized to provide a preliminary level of protection. Unfortunately, the real bottleneck lies in the difficulty of extracting all sensitive data for anonymization, due to a lot of data to handle as well as the considerable effort required. We propose a Natural Language Processing approach for sensitive knowledge detection and classification, performed by adopting a federated approach. Accuracy decay and latency introduced by applying a decentralized learning approach compared to the same task and data performed in a centralized way were evaluated. Preliminary results proved that effectiveness can be reached by a correct tuning of the federated algorithm and by choosing the right number of participants to the federation.

Natural Language Processing (NLP) is being recently explored also to its application in supporting malicious activities and objects detection. Furthermore, NLP and Deep Learning have become targets of malicious attacks too. Very recent researches evidenced that adversarial attacks are able to affect also NLP tasks, in addition to the more popular adversarial attacks on deep learning systems for image processing tasks. More precisely, while small perturbations applied to the data set adopted for training typical NLP tasks (e.g., Part-of-Speech Tagging, Named Entity Recognition, etc..) could be easily recognized, models poisoning, performed by the means of altered data models, typically provided in the transfer learning phase to a deep neural networks (e.g., poisoning attacks by word embeddings), are harder to be detected. In this work, we preliminary explore the effectiveness of a poisoned word embeddings attack aimed at a deep neural network trained to accomplish a Named Entity Recognition (NER) task. By adopting the NER case study, we aimed to analyze the severity of such a kind of attack to accuracy in recognizing the right classes for the given entities. Finally, this study represents a preliminary step to assess the impact and the vulnerabilities of some NLP systems we adopt in our research activities, and further investigating some potential mitigation strategies, in order to make these systems more resilient to data and models poisoning attacks.

Recent years have seen the widespread adoption of Artificial Intelligence techniques in several domains, including healthcare, justice, assisted driving and Natural Language Processing (NLP) based applications (e.g., the Fake News detection). Those mentioned are just a few examples of some domains that are particularly critical and sensitive to the reliability of the adopted machine learning systems. Therefore, several Artificial Intelligence approaches were adopted as support to realize easy and reliable solutions aimed at improving the early diagnosis, personalized treatment, remote patient monitoring and better decision-making with a consequent reduction of healthcare costs. Recent studies have shown that these techniques are venerable to attacks by adversaries at phases of artificial intelligence. Poisoned data set are the most common attack to the reliability of Artificial Intelligence approaches. Noise, for example, can have a significant impact on the overall performance of a machine learning model. This study discusses the strength of impact of noise on classification algorithms. In detail, the reliability of several machine learning techniques to distinguish correctly pathological and healthy voices by analysing poisoning data was evaluated. Voice samples selected by available database, widely used in research sector, the Saarbruecken Voice Database, were processed and analysed to evaluate the resilience and classification accuracy of these techniques. All analyses are evaluated in terms of accuracy, specificity, sensitivity, F1-score and ROC area.

Atherosclerosis is a multifactorial disease that affects a significant number of people during their lifetime. It is a pathology clinically silent for years that develops with a gradual thickening of the vessel walls and a consecutive formation of plaque. This is the cause of several dangerous conditions such as ischemic stroke, the most common cause of stroke in middle-aged people. To avoid and reduce these events a continuos and meticulous monitoring of patients with any carotid diseases is necessary. This paper presents an objective measure of the progression of a carotid patology. An index capable of distinguishing between the initial state of thickening of the carotid arterial walls and the successive presence of more serious plaque has been defined. The presence of thickening or plaque is estimated by evaluating Heart Rate Variability. This is a non-invasive approach, able to estimate characteristic parameters in an easy and efficient way, constituting an accurate and optimum instrument for a real-time continuous monitoring. Locally Weighted Learning has been used to automatically find a relationship between these parameters and the presence of a disorder, tested on an available dataset.