Abstract

Abstract
Segmentation of the left ventricle (LV) in echocardiography is an important task for the quantification of volume and mass in heart disease. Continuing advances in echocardiography have extended imaging capabilities into the 3D domain, subsequently overcoming the geometric assumptions associated with conventional 2D acquisitions. Nevertheless, the analysis of 3D echocardiography (3DE) poses several challenges associated with limited spatial resolution, poor contrast-to-noise ratio, complex noise characteristics, and image anisotropy. To develop automated methods for 3DE analysis, a sufficiently large, labeled dataset is typically required. However, ground truth segmentations have historically been difficult to obtain due to the high inter-observer variability associated with manual analysis. We address this lack of expert consensus by registering labels derived from higher-resolution subject-specific cardiac magnetic resonance (CMR) images, producing 536 annotated 3DE images from 143 human subjects (10 of which were excluded). This heterogeneous population consists of healthy controls and patients with cardiac disease, across a range of demographics. To demonstrate the utility of such a dataset, a state-of-the-art, self-configuring deep learning network for semantic segmentation was employed for automated 3DE analysis. Using the proposed dataset for training, the network produced measurement biases of -9 +/- 16 ml, -1 +/- 10 ml, -2 +/- 5 %, and 5 +/- 23 g, for end-diastolic volume, end-systolic volume, ejection fraction, and mass, respectively, outperforming an expert human observer in terms of accuracy as well as scan-rescan reproducibility. As part of the Cardiac Atlas Project, we present here a large, publicly available 3DE dataset with ground truth labels that leverage the higher resolution and contrast of CMR, to provide a new benchmark for automated 3DE analysis. Such an approach not only reduces the effect of observer-specific bias present in manual 3DE annotations, but also enables the development of analysis techniques which exhibit better agreement with CMR compared to conventional methods. This represents an important step for enabling more efficient and accurate diagnostic and prognostic information to be obtained from echocardiography.

Abstract
Low-cost, small-sized portable laboratories, or take-home laboratories, have been increasing in popularity worldwide. One example of such a successful Arduino-based kit is the Temperature Control Laboratory (TCLab), originally proposed by [1]. This kit has been used in Portugal for control engineering education since 2018. This paper proposes a TCLab demo session, reflecting the use of this kit in Portugal across different educational contexts. © 2023 IEEE.

Abstract
Artificial Intelligence has been introduced in many applications, namely in artificial vision-based systems with object detection tasks. This paper presents an object localization system with a motivation to use it in autonomous mobile robots at robotics competitions. The system aims to allow robots to accomplish their tasks more efficiently. Object detection is performed using a camera and artificial intelligence based on the YOLOv4 Tiny detection model. An algorithm was developed that uses the data from the system to estimate the parameters of location, distance, and orientation based on the pinhole camera model and trigonometric modelling. It can be used in smart identification procedures of objects. Practical tests and results are presented, constantly locating the objects and with errors between 0.16 and 3.8 cm, concluding that the object localization system is adequate for autonomous mobile robots. © 2023 IEEE.

Abstract
Machine learning algorithms offer the capability to analyze large volumes of real-time data, providing transport authorities with valuable insights into traffic conditions, congestion hotspots, and incident detection from diverse data sources. However, these algorithms face challenges related to data quality and reliability. We conducted a comparative analysis of machine-learning models that can be used to identify and filter transportation content from social media or other sources that can provide small and concise text. The filtrated result can then feed models and/or tools used to improve and automate traffic control, operational management, and tactical management decision-making. We consider factors such as run time, generalization capacity, and performance metrics as criteria to assess their suitability for different decision levels. The analysis is supported by a dataset consisting of Twitter content. The predictions from three groups of algorithms are evaluated: traditional machine learning algorithms (Support Vector Machines, Logistic Regression, and Random Forest), a fine-tuned Google BERT model, and Google BERT models without training (BERT-base and BERT-large). The tests are performed using New York, London, and Melbourne data. The findings of this research aim to assist decision-makers in making informed choices when selecting the most appropriate method to filtrate information subsequently used for models that contribute to different traffic management tasks.

Abstract