Objectives: To develop and validate a framework combining deep learning-based detection with geometric severity assessment of open gingival embrasures (OGEs) from intraoral photographs. Methods: A total of 3995 OGEs from 653 intraoral photographs collected across three orthodontic centres were annotated at the pixel level for maxillary tooth crowns and anterior OGEs. Model development used five-fold cross-validation on images from Centres 1 and 2, with patient-level separation to prevent data leakage. Five cross-validation models per architecture were evaluated on an independent test set from Centre 3. Performance was evaluated at the pixel (DSC, IoU), instance (precision, recall, F1-score), and grading level (accuracy, AUC). Results: Mask R-CNN achieved superior OGE segmentation performance (DSC 0.812 f 0.111 vs. 0.597 f 0.135 for YOLOv8), with improved boundary accuracy, while both models demonstrated high tooth crown segmentation accuracy (DSC > 0.95). Among successfully detected and matched OGEs, the grading framework demonstrated strong discriminative performance, with area under the curve values of 0.939 f 0.009 for YOLOv8 and 0.933 f 0.008 for Mask R-CNN, and comparable classification accuracy between models (0.86 f 0.01 for YOLOv8 and 0.86 f 0.03 for Mask R-CNN). As no Grade III OGEs were present in the external test set, grading performance was evaluated only between Grades I and II. Conclusions: This study presents a framework combining deep learning-based OGE detection with interpretable geometric severity grading. Grading performance was evaluated only between Grades I and II in the external test set; Grade III requires further validation.
Comfort foods are commonly consumed for emotional consolation, and cravings for such foods are driven by psychological and physiological processes. Habitual dietary patterns may further influence these cravings through differences in sensory exposure and familiarity. As plant-based diets become more widely adopted worldwide, emerging evidence suggests that they may be associated with differences in sensory perception, food preference, and craving patterns. However, whether such changes extend specifically to comfort food cravings remains unclear. This study examined comfort food cravings among women following plant-based and omnivorous dietary patterns using a questionnaire study. Participants listed their three most frequently craved comfort foods over the past three months and evaluated the associated sensory and emotional attributes using the checkall-that-apply (CATA) method. In addition, attitudes toward comfort foods were assessed. A total of 522 women took part in the study (195 plant-based dieters, including 80 vegans; 327 omnivores). Across dietary groups, the most frequently craved comfort foods were comparable in terms of energy content. However, foods reported by some plant-based groups had a higher estimated protein contribution than those reported by omnivores (p < 0.05). Sensory characteristics also differed across dietary groups, with plant-based dieters more frequently selecting 'vegetable' and 'earthy' smells, whereas omnivores more frequently selecting 'sweet' taste and 'meaty' smell (all p < 0.05). The most frequently reported emotions associated with comfort foods were consistently 'satisfied' and 'happy' across all groups. These findings provide new insights into how comfort food cravings and associated sensory characteristics may vary across dietary patterns among women.
Background In Aotearoa New Zealand (NZ) there have been multiple changes to the childhood pneumococcal vaccine schedule since 2008 when PCV7 was first introduced. Most recently PCV13 was reintroduced in late 2022 after 5 years of PCV10. In the present study, we investigate the shifts in invasive pneumococcal disease (IPD) incidence rates attributed to PCV-specific serotypes. Methods We report PCV-specific trends in IPD, PCV-specific vaccine impact (VI) by age group, and report the timing for expected reductions in PCV13-specific (non-PCV7 serotypes) IPD for children under 5 and adults 65 or over. Findings IPD incidence rates in NZ are at historically high levels for adults 65 or over, while rates for children under 2 and 2–4 years old have dropped significantly since the re-introduction of PCV13. Māori and Pacific peoples are experiencing the highest rates of IPD since 2009. The percent of IPD cases attributed to PCV13-specific serotypes has decreased across all age groups, while the percent of cases attributed to PCV20-specific serotypes has increased for most age groups. The VI of PCV13 was 79% (95% CI 59–90%) for children under 2 years of age. Similarly, the timing of a 50% decrease in risk for PCV13-specific (non-PCV7 serotypes) IPD was very rapid for children under 2 (within 1 year) and delayed for adults. Interpretation Though overall incidence rates of IPD remains very high, this is driven primarily by non-PCV serotypes. The VI for all PCVs remains high, especially for younger cohorts. Delayed indirect benefits are continuing to be experienced by older populations. Funding The NZ Ministry of Health funds IPD surveillance in NZ.
Addiction is a chronic, relapsing disorder comprising three stages: binge, withdrawal, and preoccupation. Preclinical evidence suggests that the habenula, a small epithalamic region, is involved across these stages and may contribute to long-term neuroadaptations in reward, stress, and control systems. Human research on the habenula in addiction is growing but remains fragmented. This scoping review synthesizes human evidence on the habenula in addiction and assesses convergence with preclinical models. Following PRISMA-ScR guidelines, databases and grey literature were searched in November 2025. Studies were included if they examined habenular measures in relation to addiction in humans, and findings were synthesized by measure category: volume, structural connectivity, functional connectivity, activity, and genetic measures. Within each category, findings were interpreted by addiction stage where applicable. Nineteen studies were included. Across categories, studies reported reduced fronto–habenula white-matter integrity, elevated habenula–striatal resting-state functional connectivity, and withdrawal-related habenula–frontal connectivity. By contrast, volumetric studies did not find consistent evidence linking habenula volume to addiction status. Activity findings included increased habenula blood flow during withdrawal, as well as task-based fMRI evidence of reduced differentiation between positive and negative feedback and heightened reward responses. Genetic findings most consistently implicated nicotine-related habenula signaling. Human findings converged most clearly with preclinical models for withdrawal and nicotine mechanisms, but no domain showed direct one-to-one human–preclinical alignment. Several domains were either one-sided, based on non-equivalent measures, or absent in both literatures. Future studies using comparable designs are needed to clarify how the habenula contributes to addiction.
Biodiversity loss poses profound strategic and operational challenges for firms, yet management research offers limited evidence on which organisational mechanisms effectively reduce corporate biodiversity impacts. This study introduces the Biodiversity Strategy Configuration Framework (BSCF) to enhance the understanding of firm-level biodiversity strategy. Drawing on dynamic capabilities, institutional theory, and collaborative governance, this study examines how biodiversity strategy (hereinafter, biodiversity impact reduction) depends on the alignment between nature-based governance initiative portfolios and external collaborative resources. Using S&P 1500 firms from 2005 to 2024, this study tests the predictions through logit and probit models, interaction effects across four governance portfolios, latent class analysis, marginal effects, and a difference-in-differences design. Results show that nature-based initiatives significantly increase the likelihood that firms will adopt a biodiversity strategy. The study documents that firms with more substantial capability portfolios (both in CSR sustainability and environmental restoration initiatives) implement more proactive and impactful biodiversity strategies. Overall, the results suggest that they support broader sustainability objectives, particularly those outlined in the Sustainable Development Goals (SDG 13: Climate Action, SDG 14: Life Below Water, and SDG 15: Life on Land).
