SALC-Net: A lightweight contour-preserving segmentation network for yak body segmentation in complex grazing environments.

Across Canada, cannabis use is highly prevalent among young adults. Although cannabis use is generally associated with low health risk in the general population, specific consumption behaviors (eg, frequency, mode of administration), individual factors (eg, age, genetic vulnerability), and societal factors (eg, normalization of cannabis use) increase the likelihood of negative health outcomes [,]. One subgroup of particular concern is youth experiencing a first episode of psychosis (FEP), who show a significantly higher prevalence of cannabis use disorder (CUD; 42%‐53%) [,] compared to the general population (about 3%) [,]. Continued cannabis use while receiving early intervention services (EIS) for psychosis is associated with increased symptom severity, higher relapse rates, more frequent hospitalizations, and reduced functional outcomes [,,]. Although abstinence is often recommended as the ideal treatment target, many young users are ambivalent or unwilling to quit cannabis completely, limiting their engagement and treatment outcomes []. Notably, few face-to-face interventions aimed at abstinence or reduction have had significant effects on abstinence or reduction rates in young people with psychotic disorders and CUD [,]. Harm-reduction approaches, which aim to mitigate substance use–related harms without requiring abstinence, offer a pragmatic and person-centered alternative that is progressively being recommended by public health authorities, clinicians, and people with lived experience [-]. The Lower-Risk Cannabis Use Guidelines (LRCUG), recently adapted to address psychosis-related risks (LRCUG-PSYCH), provide a foundation for evidence-based harm reduction but have yet to be operationalized into scalable interventions [,]. Digital health interventions may help address this gap by offering flexible and accessible support that can be integrated into users’ daily lives [-]. However, findings from recent reviews suggest mixed effectiveness of digital interventions for cannabis use among youth, partly due to heterogeneity in intervention design, targeted behaviors, and user engagement [,]. Recent literature emphasizes engagement as a critical determinant of successful digital mental health interventions. Engagement (defined as the extent to which users interact with the intervention) has also been conceptualized as a multidimensional construct influenced not only by the intervention itself but also by the context of use, the mechanisms of action behind the digital tool, and the targeted behavior that work together [-]. Common indicators include uptake (ie, downloading and using the intervention at least once), sustained use (ie, continued activity over time), and adherence or completion (ie, following the intervention as intended) []. Understanding and clarifying how users interact with these tools and what drives sustained use is essential, while taking into account that these metrics may capture related but distinct aspects of intervention use and therefore require cautious interpretation []. Some studies also suggest that individual-level factors such as gender, education, and digital literacy may influence how users interact with and benefit from digital tools [,]. In particular, baseline social support might be relevant to engagement, as individuals with higher interpersonal (relational and peer) support may have more resources, incentives, and/or stability to initiate and sustain their participation in digital interventions []. Conversely, supportive features embedded within digital interventions (eg, peer interaction, built-in buddy features, e-coaching, etc) have also been described as strategies associated with greater use and engagement [].