Causal evidence from invasive recordings reveals distributed acoustic feature encoding underlying categorical voice perception.

Neurological injuries and conditions commonly result in substantial loss of upper limb function, which greatly limits an individual’s ability to perform activities of daily living (ADLs). Across diverse neurological populations, impaired hand function consistently ranks as a primary rehabilitation concern. For instance, individuals with spinal cord injury (SCI) consistently report regaining hand and arm function as the top recovery priority []. Similarly, for individuals who have experienced a stroke, data suggest upper-limb strength is a major predictor of quality of life []. Therefore, restoring hand function is a primary focus of rehabilitation efforts, as it directly impacts the ability to regain independence and quality of life. Regular and accurate monitoring of hand function is essential for promoting recovery []. It enables clinical assessment and progress tracking, thereby allowing therapists to adjust treatment plans, and ensure therapy addresses real-world functional needs []. However, monitoring this function outside of the clinic remains a significant challenge [, ]. Conventional outpatient therapy is typically limited to face-to-face assessments and patient self-reports, which often fail to capture the complexity and variability of real-life hand use at home [–]. Patient self-reporting also relies on an individual’s ability to recall activities performed since their last therapy session and is subject to reporting biases, such as cognitive deficits and social desirability [–]. These limitations prevent clinicians in outpatient settings from gaining a comprehensive understanding of how patients perform ADLs in their everyday environments, thus hindering the ability to tailor therapy plans on the basis of real-world functional performance. Wearable technology offers an avenue for measuring hand function outside of the clinic. In particular, head-mounted egocentric cameras can provide rich, contextual insights into patients’ hand function during ADLs [–]. However, clinicians have shown reluctance to adopting new technologies in their practices, citing concerns about time constraints and uncertainty regarding clinical value [, –]—factors identified by the Technology Acceptance Model [] as primary barriers to adoption []. Digital tools, such as clinical decision support systems (CDSS), have demonstrated success in implementing evidence-based guidelines at the point of care and improving patient outcomes in various healthcare domains [], including rehabilitation [, ], chronic disease management [], and prescription and medication safety []. However, this success hinges on whether they present information that is useful, interpretable, and actionable for clinicians [–]. Therefore, understanding how clinicians perceive and value specific types of information from wearable devices is crucial for developing systems that address time and value concerns that have previously hindered the adoption of new technologies in clinical practice.