This scoping review examines current evidence supporting multimodal artificial intelligence, continuous monitoring, and digital twin concepts in spine care. Our primary aims were to (1) characterize the state of digital twin development in spine care, (2) identify key technological and conceptual gaps, and (3) evaluate translational barriers to clinical implementation. A scoping review was conducted following PRISMA-ScR guidelines. PubMed/MEDLINE, Scopus, and Web of Science were searched for studies published between January 2010 and March 2025. Findings were synthesized qualitatively. Twenty-six studies met inclusion criteria. Existing spine prediction models demonstrate modest discrimination and are predominantly static. Imaging-based AI shows weak associations with pain and disability. Wearable sensor monitoring is feasible but lacks consistent evidence for improved outcomes. Spine-specific digital twins remain conceptual, with no prospective validation demonstrating improved decision-making. Multimodal AI-enabled digital twins represent a compelling conceptual framework for personalized spine care, but current evidence does not support clinical superiority or readiness for implementation. Progress will require prospective validation, standardized data integration, and regulatory clarity.
Background Central visual field loss (CFL) is the most common irreversible visual impairment in aging and is associated with higher fall risk and concerns about falling. This study explored the links between CFL severity, functional balance, and walking-related attentional processing implicated in reduced gait performance. Methods In Study 1, 29 individuals with CFL and 29 age-matched controls completed the Timed Up and Go (TUG) test. In Study 2, 10 CFL participants and 10 controls performed the TUG while acceleration data were collected from head and trunk IMUs. For both studies, we assessed visual impairment severity (contrast sensitivity) and participants’ attentional processing during walking (Gait-Specific Attentional Profile, G-SAP). Results Both groups showed positive correlations between TUG duration and G-SAP subscales. G-SAP scores were lower in CFL participants with worse contrast sensitivity indicating reduced cognitive processing during walking. Worse contrast sensitivity was also associated with greater head and trunk acceleration and acceleration variability during walking, suggesting reduced gait stability. Higher rumination and conscious movement processing scores also correlated with improved segmental control in CFL. Significance Increased cognitive processing of gait is associated with impaired functional balance. This association appears to be reversed in CFL, with severe visual deficit diverting cognitive resources from movement control. This altered strategy may prioritise the acquisition and processing of visuospatial information in CFL. The observed postural instability with increasing CFL severity and a lack of excessive cognitive involvement in movement control suggest heightened gait-specific attention could be leveraged for balance and gait training in CFL.
In humans and other animals, social robots can serve as effective tutors for learning new skills. Young oscines learn their song by imitating conspecific adults. In a previous study, we demonstrated that a robotic bird can be as effective as a live tutor in training a young zebra finch (Taeniopygia guttata) to imitate a song model. Here, we take this further by investigating the role of behavioural contingency in developmental song learning and in shaping the birds' engagement with the robot. Two groups of young male finches were exposed to a robotic tutor under contingent (CON) or non-contingent (NCON) conditions. In the CON group, the robot produced a call in response to a call emitted by the bird. When the bird perched nearby, the robot oriented towards it and broadcast a song. While song imitation was slightly better in the CON group, the difference was not statistically significant. However, birds in the CON group spent more time near the robot and interacted with it more frequently compared to NCON-birds. These findings highlight the importance of behavioural contingency in social robotics and offer novel insights into the use of robotic agents in studies with non-human animals.
Audio description (AD) narrates visual elements in video for blind and low-vision audiences. Recent work has shown that giving novice describers an AI-generated draft to start from helps produce higher-quality AD and lowers the barrier to entry. What remains an open question is how draft quality shapes the editing process. We investigate this through GenAD, an AD generation pipeline that incorporates accessibility guidelines and contextual video information, and RefineAD, an editing interface for human revisions. Human-AI contributions are measured across text, timing, and delivery. In a within-subjects study, we compared authoring from scratch against editing AI drafts of varying quality. GenAD drafts cut completion time by more than half and significantly reduced cognitive load. In contrast, baseline drafts generated from simple, unguided prompts offered only modest benefits, pointing to a minimum quality threshold for effectiveness. Qualitative findings suggest this threshold is content-dependent; as visual complexity increases, so does the quality needed from AI drafts. We propose this as a design principle: effective AI assistance should clear a quality threshold suited to the target content, rather than simply be present.
