Segmentation of the lungs in Computed Tomography (CT) is very challenging due to changes in lung shape, size, and parenchyma pattern, as well as differences in imaging acquisition protocols. As a consequence, these models may not be robust and may decrease their performance when deployed in a clinical setting. The Continual Learning paradigm holds great promise since learning models continually acquire incoming knowledge, having the ability to adapt to changing environments. In this work, experience replay with random sampling of past data was implemented, using the original CT images and the corresponding ground-truths. Data from four different institutions were used to develop the experiments, and the models were evaluated on a cross-cohort dataset. Using raw data, the goal was to study how the datasets and their imaging patterns were related and what impact the training datasets have on one another. The catastrophic forgetting effect diminished for almost all datasets. For two of the in-domain test datasets there was forward and backward transfer, results that could be linked to a possible similarity between them. A mean DSC of 0.94 was obtained across all datasets. The results showed how the similarity or disparity between data from different institutions can influence the performance of learning models.

Lung cancer remains the leading cause of cancer related deaths globally, responsible for approximately 1.8 million deaths each year. A key contributor to this high mortality rate is the late-stage diagnosis of the disease, underscoring the urgent need for effective early detection strategies. Low-dose computed tomography (CT) has shown great value in early screening, particularly when paired with clinical information. Clinical data, while valuable, lacks spatial and morphological insights essential for comprehensive evaluation. Combining both modalities offers a more holistic approach for lung cancer classification. This study presents AI-based methods for lung cancer classification using unimodal approaches - structured clinical data and chest CT imaging - alongside a novel multimodal deep learning framework that integrates both data types to classify lung nodules as malignant or benign. For the clinical modality, machine learning models including logistic regression, random forests, LightGBM, XGBoost, and multilayer perceptrons were evaluated with extensive hyperparameter tuning. In the imaging modality, ResNet18 and ResNet34 convolutional neural networks were used, with and without data augmentation. The study explored both intermediate and late fusion strategies to combine modality-specific representations. Results show that multimodal models consistently outperformed their unimodal counterparts, achieving a best-case area under the ROC curve (AUC) of 0.9138, with an accuracy of 0.8424 and an F1-score of 0.8422. These findings highlight the complementary strengths of imaging and clinical data and support the growing potential of multimodal deep learning in improving diagnostic accuracy in lung cancer classification.

The recognition and evaluation of emotional states have important applications in the medical domain. Emotion recognition has been an active area of research in recent years, with a significant focus on data sources such as images or text, but also physiological signals like the electrocardiogram or electroencephalogram. However, traditional data collection and labelling methods often rely on exaggeratedly acted emotions or intense real emotions triggered by strong stimuli. These fail to mimic the conditions of most real world scenarios, where highly-intense emotional manifestations are not the most common. This work addresses these issues by evaluating the suitability of using physiological signals for emotion classification. We conducted a data collection protocol to acquire realistic emotional data in an isolated and undisturbed setup, with exposure to weak stimuli and standard activities, to capture low-intensity emotional manifestations. We present an initial exploratory analysis of physiological data, followed by the development of an emotion recognition training strategy using Machine Learning algorithms. Our approach achieved an optimal balanced accuracy of 47.97% and an area under the receiver operating characteristic curve of 72.09% for a multi-class classification problem with four emotion classes. Although the results may seem modest at a first glance, it is important to consider the inherent difficulty of distinguishing between subtle, low-intensity emotions, as well as the relevance of the problem to healthcare applications. Therefore, this work supports the development of more holistic, patient-centred healthcare solutions with emotion recognition.

MQTT is a widely used application-layer protocol for Internet-of-Things (IoT) and Industrial IoT networks. However, its centralized architecture, which relies on a broker that mediates all data transmissions, can lead to traffic congestion and reduced network capability, particularly in networks with a large number of devices, such as edge networks. This paper proposes an optimized SDN-based real-time MQTT architecture for edge networks, called Multicast Real-Time MQTT (MRT-MQTT) that leverages multicast routing mechanisms to efficiently distribute data while supporting timeliness requirements and minimizing network usage. Extensive emulation experiments show the effectiveness of the proposed approach and its superiority over Direct Multicast-MQTT (DM-MQTT) and Standard MQTT (STD-MQTT). We achieved a reduction in transmission delay of 29% and 23% for QoS=0 and QoS=1, respectively, when compared to DM-MQTT and 55% and 43% compared to STD-MQTT. In this last case, MRT-MQTT also reduced network usage by 58.0% and 45.0% for QoS=0 and QoS=1, respectively.

Embedded systems (ES) are wide-spread in our world and responsible for many critical systems.More recently, machine learning (ML) tools have become a well-established solution for data-intensive tasks, but their application in embedded systems is still gaining traction and their real-time performance is often unclear.We provide a (non-extensive) review of the ML tools that may be suited for deployment in ES, from which we selected two representative tools -the wellestablished Python-based Scikit-Learn, and the interoperabilityoriented ONNX Runtime -to compare their response time.Using archetypal datasets and four pre-trained ML models, we measure the prediction time for each sample, for each model, in Scikit-Learn and ONNX Runtime in a standard desktop (to compare performance of the tools in the same platform), and for ONNX Runtime in a representative ES, a Raspberry Pi v4 (to compare performance of the same tool across platforms).We report that ONNX considerably improves over Scikit-Learn, and experiences a negligible performance degradation when ported to the RPi.