All traces

Sentinel-2 and CART-based operational AGB mapping in Pinus brutia produced total estimates that differed by more than 700,000 Mg across the same 13,687 ha area depending solely on which allometric reference was used, making reference selection a dominant source of uncertainty.

Technostress mediated the relationship, such that when openness to change and positive AI attitudes were low, higher technostress was associated with reduced innovative work behavior among the same hospital staff.

In educational use, ChatGPT can generate wrong information and reflect training-data biases that may augment existing biases, raising privacy issues.

Multi-omics biomarker approaches face persistent challenges in data heterogeneity, reproducibility, and clinical validation across diverse patient populations.

Oncology AI systems can reproduce or amplify existing disparities across patient populations, and efforts to enforce fairness definitions often conflict with overall predictive performance.

When used for prostate MRI reporting, LLMs can hallucinate measurements, flip negations, misstate laterality, and overstate cancer likelihood, creating patient-safety and accountability risks especially if reports are copied outside clinical governance.

Perceived risk around AIGC reduced fashion designers' feelings of autonomy, competence and relatedness, undermining psychological conditions for adoption.

Most models lacked external validation, only one was low risk of bias, publication bias was detected, and performance dropped in HIV-positive populations, leaving models not ready for routine clinical implementation.

AI systems applied to healthcare decision-making, medical diagnosis, and other domains can lead to unfair outcomes that perpetuate existing inequalities and reinforce harmful stereotypes, including generative biases in synthetic media.

Artificial intelligence approaches for CT-based prediction of hematoma expansion and adverse outcomes after spontaneous ICH are limited by small cohorts, overfitting, dataset heterogeneity, insufficient external validation, poor interpretability and lack of workflow integration.

Despite belief AI will revolutionise nursing education, actual implementation remains conservative at augmentation level with none achieving transformative redefinition.

ML applications in clinical microbiology face typical challenges including class imbalance, limited generalization, robustness issues, and need for model interpretation and explainability to achieve robust performance under real-world variability.

Models that were near-perfect during internal cross-validation diverged significantly on the separate external cohort, showing reduced out-of-distribution generalization for cross-species histological classification.

ML, DL and RL-based cybersecurity solutions are susceptible to adversarial attacks, and ChatGPT-like tools can be manipulated to threaten data integrity, confidentiality and availability.

The LLM-assisted workflow did not consistently outperform ICD-based retrieval, with no statistically significant AUC difference in Cohort 1 and lower AUC than ICD codes for heart failure in that cohort.