Abstract
Low-cost smartphone-compatible portable ophthalmoscopes can capture visuals of the patient's retina to screen several ophthalmological diseases like glaucoma. The images captured have lower quality and resolution than standard retinography devices but enough for glaucoma screening. Small videos are captured to improve the chance of inspecting the eye properly; however, those videos may not always have enough quality for screening glaucoma, and the patient needs to repeat the inspection later. In this paper, a method for automatic assessment of the quality of videos captured using the D-Eye lens is proposed and evaluated with a personal dataset with 539 videos. Based on two methods developed for retina localization on the images/frames, the Circle Hough Transform method with a precision of 78,12% and the YOLOv7 method with a precision of 99,78%, the quality assessment method automatically decides on the quality of the video by measuring the number of frames of good-quality in each video, according to the chosen threshold. © 2024 Elsevier B.V.. All rights reserved.

Abstract
Recent advancements in digital technology have significantly impacted healthcare, with the rise of chatbots as a promising avenue for healthcare services. These chatbots aim to provide prevention, diagnosis, and treatment services, thereby reducing the workload on medical professionals. Despite this trend, limited research has explored the variables influencing user experiences in the design of healthcare chatbots. While the impact of visual representation within chatbot systems is recognized, existing studies have primarily focused on efficiency and accuracy, neglecting graphical interfaces and non-verbal visual communication tools. This research aims to delve into user experience aspects of symptom checker chatbots, including identity design, interface layout, and visual communication mechanisms. Data was collected through a comprehensive questionnaire involving three distinct chatbots (Healthily, Mediktor and Adele - a self-developed solution) and underwent meticulous analysis, yielding valuable insights to aid the decision process when designing effective chatbots for symptom checking.

Abstract
As we transition from centralized machine learning to distributed learning, new practices can significantly enhance intelligent Internet of Things (IoT) systems. This article introduces the concept of Opportunistic Distributed Learning (ODL), a general framework that enables any node in a network to initiates learning tasks by leveraging local, unused distributed resources collaboratively. ODL, facilitated by edge intelligence, promotes collective responsibility, pervasive and flexible distributed learning, allowing participating nodes to freely move, group, and regroup based on their conditions and benefits. The article discusses key research challenges of ODL in intelligent IoT systems, presents the ODL framework, proposes a reputation-based node selection scheme, and highlights the benefits and future research directions of the ODL system.

Abstract

Abstract
TADA is a unique toolkit designed to foster the use and implementation of approximate distributed agreement primitives. Developed in Java, TADA provides ready-to-use implementations of several approximate agreement algorithms, as well as the tools to enable programmers/researchers to easily implement further protocols: A template that enables new protocol implementations to be created by simply changing specific functions; and high-level abstractions for communication and concurrency control. As an example, the toolkit includes a ready-to-use implementation for clock synchronisation between distributed processes. Further use cases can include sensor input stabilisation and distributed machine learning, or other instances of distributed agreement where network synchrony cannot be assumed, byzantine fault tolerance may be required and a bounded divergence in decision values can be tolerated.

Abstract
The deep-sea environment still presents many challenges for systematic, comprehensive data acquisition. The current generation of SMART cables incorporates low-power sensors in long-range telecommunication cables to improve knowledge of ocean variables, aid in earthquake and tsunami warnings, and enhance coastal protection. The K2D Project seeks to expand SMART cables' capabilities by increasing the diversity of sensors along deep water cables, integrating active devices, and leveraging mobile platforms like deep-water AUVs, thereby improving spatial coverage and advancing ocean monitoring technology. This paper discusses a demonstration of these capabilities, focusing on the description of the main building blocks developed along the project, with results from a sea deployment in September 2023.