BACKGROUND:Musculoskeletal ultrasound (US) is a noninvasive tool for joint assessment in persons with hemophilia. Early detection of joint bleeding using a remote US system operated by patients or caregivers and reviewed by Comprehensive Care Centers could improve personalized management. A computer-aided diagnosis (CAD) system for automatic detection of joint effusion may support clinicians in prioritizing interventions. OBJECTIVES:This study aimed to validate a novel CAD system using a deep-learning algorithm to identify joint capsule distension in musculoskeletal US images. METHODS:Longitudinal scans of the subquadricipital recess of the knee were collected from people with hemophilia and varying degrees of arthropathy and labeled by an expert. The multitask learning algorithm was trained to detect the recess and classify images as distended or not. RESULTS:A total of 8634 images (2267 scans) were acquired from 158 adult persons with hemophilia (mean age 44.7 ± 18.6 years) and 66 age-matched healthy controls. After selecting longitudinal subquadricipital recess images, 814 images were used, of which 711 for training and 103 for testing, ensuring a patient-based split. The model achieved a classification accuracy of 89.2% and a balanced accuracy of 93.9% compared with expert annotations. No significant differences were observed in classification performance between male and female healthy controls, supporting its broader applicability. CONCLUSION:The CAD system for automatic detection of joint capsule distension is feasible and reliable. It represents an important step toward telemedicine in hemophilia, enabling early recognition of joint bleeding and supporting personalized, timely therapeutic interventions to prevent further joint damage.

This paper presents an end-to-end pipeline for transforming indoor floor plans into accessible tactile maps enriched with audio annotations. The approach integrates SIM (Semantic Interior Mapology), which allows authors to trace floor plans and assign semantic labels to relevant spatial features (e.g., rooms, corridors, restrooms, and stairs), with CamIO, a camera-based web application that delivers spoken feedback when users point to interactive regions (hotspots) on a tactile graphic. A pilot study with two blind participants and a follow-up study with five additional participants showed that users could successfully complete room-finding and route-tracing tasks and appreciated the system's support for spatial understanding. Usability evaluation yielded a mean SUS score of 77, indicating good usability, while NASA-TLX results showed a relatively low perceived workload, supporting the effectiveness of the proposed solution. Results also highlighted areas for improvement, including lower latency and clearer interaction around doors and small hotspots.

Interacting with video games is challenging for people with upper-limb impairments, especially when multiple hand-based inputs are required in rapid succession. Human cooperation, where another person assists the player, has been proposed as a solution, but it is limited by copilot availability and co-location. An alternative is partial automation, where the player is assisted by a software agent. We present a study with 13 participants with upper-limb impairments, investigating how they collaborate with a copilot in both human cooperation and partial automation. The experiment is supported by GamePals, a configurable framework we developed to enable both human cooperation and partial automation in existing third-party video games.

For blind or low vision individuals, tactile graphics (TGs) provide essential spatial information but are restricted in the amount of data they can convey - especially semantic information, which is represented using the small amount of braille abbreviations that fit on the TG. Using computer vision, TGs can be enhanced by adding audio labels, in which audio descriptions are triggered when the user touches elements on the TG; these audio descriptions can contain unlimited amounts of semantic information and are accessible even to those who don't read braille. Unfortunately, existing audio label systems are closed systems with severe limitations, some of which are costly and/or tied to specific hardware platforms. We address this problem by creating CamIO-Web, an open-source version of our CamIO (short for "Camera Input-Output") audio label system, which runs in the browser of virtually any computer or mobile device. The system includes facilities that allow users to create their own TGs with any audio labels ( audio recordings or Text-to-Speech), and the open-source code base makes it extensible.

Independent navigation remains a significant challenge for blind and low vision individuals, especially in unfamiliar environments. In this paper, we introduce the Parsimonious Instructions design principle, which aims to enhance navigation safety while minimizing the number of instructions delivered to the user. We demonstrate the application of this principle through NavGraph, a navigation application adopting a modular architecture comprising four components: localization, routing, guidance, and user interface. NavGraph is designed to provide effective, non-intrusive navigation assistance by optimizing route computation and instruction delivery. We evaluated NavGraph in a user study with 10 blind participants, comparing it to a baseline solution. Results show that NavGraph significantly reduces the number of instructions and improves clarity and safety, without compromising navigation time. These findings support the potential of the Parsimonious Instructions design principle in assistive navigation technologies.

Timely diagnosis of joint bleeds is crucial for preventing long-term damage in patients with hemophilia. However, access to specialized care and the operator-dependent nature of ultrasound imaging pose significant challenges for remote monitoring. We present GAJA (Guided self-Acquisition of Joint ultrAsound images), a mobile system that interactively guides patients during the acquisition of joint ultrasound images without requiring real-time supervision. The version of GAJA presented in this paper extends support beyond the knee to include elbow and ankle joints and integrates with the CADET platform, enabling clinicians to remotely assess the acquired images. In this demo, we showcase GAJA’s real-time guidance interaction and its integration with remote clinical workflows.

This paper presents MIAS-SAM, a novel approach for the segmentation of anomalous regions in medical images. MIAS-SAM uses a patch-based memory bank to store relevant image features, which are extracted from normal data using the SAM encoder. At inference time, the embedding patches extracted from the SAM encoder are compared with those in the memory bank to obtain the anomaly map. Finally, MIAS-SAM computes the center of gravity of the anomaly map to prompt the SAM decoder, obtaining an accurate segmentation from the previously extracted features. Differently from prior works, MIAS-SAM does not require to define a threshold value to obtain the segmentation from the anomaly map. Experimental results conducted on three publicly available datasets, each with a different imaging modality (Brain MRI, Liver CT, and Retina OCT) show accurate anomaly segmentation capabilities measured using DICE score. The code is available at: https://github.com/warpcut/MIAS-SAM

In this paper we explore the use of Augmented Reality as a means to provide more widespread and equitable access to art venues and artworks, in particular for people with low vision. We investigate how people with low vision frame, access and interact with artworks in Augmented Reality, using a mobile app specifically designed for accessible and inclusive museum visits. Through a user study with 10 participants in real museum settings, we explore the specific challenges related to the accessibility of different artwork types: a medium size painting, a large tapestry, a statue, and a historical keyboard instrument. Results show that participants were able to access all artworks through an Augmented Reality mobile app running on the user’s own device. The system is also perceived as useful and usable. Additionally, we uncover human and environment factors that influence the way users access different artwork types, in particular considering Augmented Reality interactions.

For individuals who are blind or have low vision, tactile maps provide essential spatial information but are limited in the amount of data they can convey. Digitally augmented tactile maps enhance these capabilities with audio feedback, thereby combining the tactile feedback provided by the map with an audio description of the touched elements. In this context, we explore an embodied interaction paradigm to augment tactile maps with conversational interaction based on Large Language Models, thus enabling users to obtain answers to arbitrary questions regarding the map. We analyze the types of questions the users are interested in asking, engineer the Large Language Model’s prompt to provide reliable answers, and study the resulting system with a set of 10 participants, evaluating how the users interact with the system, its usability, and user experience.

This paper describes the development of PRACTICE, a distributed healthcare technological platform that supports various research initiatives by the University of Milan and the Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Fondazione IRCCS Ca’ Granda, Ospedale Maggiore Policlinico. PRACTICE includes three main components: a mobile app that patients can use to self-acquire ultrasound images at home, a computer-aided diagnosis web application that supports the practitioners through a set of machine learning models, and a set of web tools for image annotation, a prerequisite for training the machine learning models. Although PRACTICE was designed in the specific context of supporting the detection of joint recess blood effusions in hemophilic patients, this paper describes the main design and implementation challenges that apply to other applications of a research-oriented health platform.

For individuals who are blind or have low vision, tactile maps provide essential spatial information but are limited in the amount of data they can convey. Digitally augmented tactile maps enhance these capabilities with audio feedback, thereby combining the tactile feedback provided by the map with an audio description of the touched elements. In this context, we explore an embodied interaction paradigm to augment tactile maps with conversational interaction based on Large Language Models, thus enabling users to obtain answers to arbitrary questions regarding the map. We analyze the type of questions the users are interested in asking, engineer the Large Language Model's prompt to provide reliable answers, and study the resulting system with a set of 10 participants, evaluating how the users interact with the system, its usability, and user experience.

We have devised a novel "Point-and-Tap" interface that enables people who are blind or visually impaired (BVI) to easily acquire multiple levels of information about tactile graphics and 3D models. The interface uses an iPhone's depth and color cameras to track the user's hands while they interact with a model. When the user points to a feature of interest on the model with their index finger, the system reads aloud basic information about that feature. For additional information, the user lifts their index finger and taps the feature again. This process can be repeated multiple times to access additional levels of information. For instance, tapping once on a region in a tactile map could trigger the name of the region, with subsequent taps eliciting the population, area, climate, etc. No audio labels are triggered unless the user makes a pointing gesture, which allows the user to explore the model freely with one or both hands. Multiple taps can be used to skip through information levels quickly, with each tap interrupting the current utterance. This allows users to reach the desired level of information more quickly than listening to all levels in sequence. Experiments with six BVI participants demonstrate that the approach is practical, easy to learn and effective.

For people with upper extremity motor impairments, interaction with mobile devices is challenging because it relies on the use of the touchscreen. Existing assistive solutions replace inaccessible touchscreen interactions with sequences of simpler and accessible ones. However, the resulting sequence takes longer to perform than the original interaction, and therefore it is unsuitable for mobile video games. In this paper, we expand our prior work on accessible interaction substitutions for video games with a new interaction modality: using facial gestures. Our approach allows users to play existing mobile video games using custom facial gestures. The gestures are defined by each user according to their own needs, and the system is trained with a small number of face gesture samples collected from the user. The recorded gestures are then mapped to the touchscreen interactions required to play a target game. Each interaction corresponds to a single face gesture, making this approach suitable for the interaction with video games. We describe the facial gesture recognition pipeline, motivating the implementation choices through preliminary experiments conducted on example videos of face gestures collected by one user without impairments. Preliminary results show that an accurate classification of facial gestures (97%) is possible even with as few as 5 samples of the user.

This paper presents LoRIS (Localized Reconstruction-by-Inpainting with Single-mask), a novel weakly-supervised anomaly detection technique designed to identify knee joint recess distension in musculoskeletal ultrasound images, which are noisy and unbalanced (as distended cases are rarer). In this context, supervised techniques require a high number of annotated images of both classes (distended and nondistended). On the other hand, we show that existing unsupervised anomaly detection techniques, which can be trained with images from a single class, are ineffective and often unable to correctly localize the anomaly. To overcome these issues, LoRIS is trained with nondistended images only and uses the recess bounding box as location prior to guide the reconstruction. Experimental results show that LoRIS outperforms state-of-the-art unsupervised anomaly detection techniques. When compared to a state-of-the-art fully supervised solution, LoRIS presents similar performance but has two key advantages: during training it requires images from a single class only, and it also outputs the recess segmentation, without the need for segmentation annotations.

INTRODUCTION:The improved quality of care and increased drug availability have shifted the goal of treating people with hemophilia from life-threatening bleeding prevention to joint health preservation and quality of life amelioration. Many tools are now available to the clinician in order to optimize the management of hemophilic arthropathy. AREAS COVERED:This paper reviews the pivotal role of ultrasound evaluation in early detection of joint bleeding and differential diagnosis of joint pain, with a focus on the feasibility of a long-term monitoring of joint health through the use of artificial intelligence and telemedicine. The literature search methodology included using keywords to search in PubMed and Google Scholar, and articles used were screened by the coauthors of this review. EXPERT OPINION:Joint ultrasound is a practical point-of-care tool with many advantages, including immediate correlation between imaging and clinical presentation, and dynamic evaluation of multiple joints. The potential of telemedicine care, coupled with a point-of-care detection device assisted by artificial intelligence, holds promises for even earlier diagnosis and treatment of joint bleeding. A multidisciplinary approach including early intervention by physical medicine and rehabilitation (PMR) physicians and physiotherapists is crucial to ensure the best possible quality of life for the patient.

We are pleased to present the first Special Issue on the International Web for All Conference (W4A) series featured in the ACM Transactions on Accessible Computing (TACCESS) journal. This volume presents seven articles that are extended versions of the conference papers presented at the 18th International Web for All Conference, which was held online on April 19–20, 2021. Authors of several top papers from the conference submitted manuscripts for consideration, which then underwent a full journal review process. The guest editors for this issue are Victoria Yaneva (National Board of Medical Examiners, USA; University of Wolverhampton, UK) and Dragan Ahmetovic (University of Milan, Italy). The guest editors thank the authors for their excellent submissions, and they also thank all of the journal reviewers who contributed their time and expertise to this process. The first article, titled “AccessComics2: Understanding the User Experience of an Accessible Comic Book Reader for Blind People with Textual Sound Effects,” proposes an accessible digital comic-book reader for people with visual impairments. Surveys and interviews with participants who are blind or have low vision revealed preference for the inclusion of brief scene descriptions and sound effects. These components were integrated into the system and further evaluated, showing that the presence of scene descriptions was useful for concentration and understanding, while the sound effects made the book reading experience more immersive and realistic. The second article, “The Transparency of Automatic Web Accessibility Evaluation Tools: Design Criteria, State of the Art, and User Perception,” presents a comprehensive survey of the instruments available for automated website accessibility evaluation, the metrics they adopt, and how these are presented to the user. Through a survey with 138 users of evaluation tools and a study with 18 accessibility and web design experts, the authors identify a number of design criteria aimed to support the transparency of the reported results and their interpretability by end-users. The third article, “The Accessibility of Data Visualizations on the Web for Screen Reader Users: Practices and Experiences during COVID-19,” explores the level of accessibility of web-based data visualizations by screen reader users. To this end, the authors conduct an accessibility audit of 87 data visualizations by 3 expert auditors, a follow-up survey with 127 screen reader users, and an observational study with 12 participants interacting with accessible web visualizations. A final discussion proposes recommendations for designing more accessible data visualizations. The fourth article, “WordMelodies: Supporting the Acquisition of Literacy Skills by Children with Visual Impairment through a Mobile App,” presents a mobile app designed to support inclusive teaching of literacy skills for primary school students. The app includes over 80 different exercise types in Italian and English with adjustable difficulty levels. Evaluation with primary

People with hemophilia require frequent diagnoses of joint bleeding. This is currently achieved with visits to specialized centers. One possibility is to have a point-of-care acquisition of the ultrasound joint image by the patients themselves, followed by a remote evaluation by the practitioner. However, the acquisition of US images is operator-dependent, so it is unclear to what extent patients can acquire images that are suitable for remote diagnosis. In this paper, we present GAJA (Guided Acquisition of Joint ultrAsound), an application designed to guide the patient in collecting US images of their own joints, which are then transmitted to a medical practitioner. GAJA uses a collaborative interaction approach, in which an expert practitioner collects a reference US image of a specific scan during an in-person clinical visit. Anatomical markers for the target joint are automatically extracted and then used as a reference to guide the patient in properly positioning the US probe.

Navigation assistance services for people with visual impairment pose the challenge of providing accurate guidance through non-visual navigation instructions. This paper proposes two techniques to guide the user during navigation assistance, and in particular during turns and straight paths. Both techniques adopt a combination of speech and non-speech audio with the aim of providing continuous, accurate, and unobtrusive guidance. The two solutions differ in the sonification technique used to generate the non-speech audio. The techniques were evaluated in a real-world environment with nine participants having severe visual impairment. Experimental results show that one of the proposed techniques is significantly more effective than a baseline approach adopted in existing solutions, in terms of navigation accuracy during turns and straight paths. Thus, it is a practical and effective solution to the problem of non-visual navigation assistance for people with visual impairment.

Joint bleeding is a common condition for people with hemophilia and, if untreated, can result in hemophilic arthropathy. Ultrasound imaging has recently emerged as an effective tool to diagnose joint recess distension caused by joint bleeding. However, no computer-aided diagnosis tool exists to support the practitioner in the diagnosis process. This paper addresses the problem of automatically detecting the recess and assessing whether it is distended in knee ultrasound images collected in patients with hemophilia. After framing the problem, we propose two different approaches: the first one adopts a one-stage object detection algorithm, while the second one is a multi-task approach with a classification and a detection branch. The experimental evaluation, conducted with $483$ annotated images, shows that the solution based on object detection alone has a balanced accuracy score of $0.74$ with a mean IoU value of $0.66$, while the multi-task approach has a higher balanced accuracy value ($0.78$) at the cost of a slightly lower mean IoU value.

People with blindness or severe low vision access mobile devices using screen readers. However, noisy environments can impair screen reader intelligibility. During mobility, this could disorient or even endanger the user. To address this issue, we propose three screen reader speech compensation techniques based on environmental noise: speech rate slowing, adaptive volume increase, and adaptive equalization. Through a study with 12 participants in three simulated noise scenarios, we evaluate screen reader intelligibility and the perceived distraction from the soundscape, with and without compensations. Four of the proposed compensations, in particular those that pair speech rate reduction with volume or equalization adaptation, significantly improve screen reader's speech intelligibility in all the considered scenarios, and the compensations do not have a significant impact on the distraction from the soundscape.