Search

This systematic literature review employs the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology to investigate recent applications of explainable AI (XAI) over the past three years. From an initial pool of 664 articles identified through the Web of Science database, 512 peer-reviewed journal articles met the inclusion criteria—namely, being recent, high-quality XAI application articles published in English—and were analyzed in detail. Both qualitative and quantitative statistical techniques were used to analyze the identified articles: qualitatively by summarizing the characteristics of the included studies based on predefined codes, and quantitatively through statistical analysis of the data. These articles were categorized according to their application domains, techniques, and evaluation methods. Health-related applications were particularly prevalent, with a strong focus on cancer diagnosis, COVID-19 management, and medical imaging. Other significant areas of application included environmental and agricultural management, industrial optimization, cybersecurity, finance, transportation, and entertainment. Additionally, emerging applications in law, education, and social care highlight XAI’s expanding impact. The review reveals a predominant use of local explanation methods, particularly SHAP and LIME, with SHAP being favored for its stability and mathematical guarantees. However, a critical gap in the evaluation of XAI results is identified, as most studies rely on anecdotal evidence or expert opinion rather than robust quantitative metrics. This underscores the urgent need for standardized evaluation frameworks to ensure the reliability and effectiveness of XAI applications. Future research should focus on developing comprehensive evaluation standards and improving the interpretability and stability of explanations. These advancements are essential for addressing the diverse demands of various application domains while ensuring trust and transparency in AI systems.

Background/Objectives: Many older adults living in long-term care (LTC) environments have varying numbers of retained natural teeth. The objective of this study was to assess the disease findings based on tooth type and estimate the role of tooth types in occlusion. Methods: We conducted clinical oral examinations of 276 LTC residents. The disease findings were analyzed for each tooth type and to determine their associations with dementia. Results: In total, 67.8% of the participants had molar teeth. Premolars/canines were often present as residual roots and had caries. Deepened periodontal pockets and higher plaque index (PI) values for molars had odds ratios of 2.5 (95% confidence interval [CI] of 1.59–3.91) and 1.61 (95% CI of 1.29–2.02), respectively. Participants with dementia were more likely to have incisors and premolars/canines in the form of root remnants and higher PI scores for all tooth types (p ≤ 0.01), as well as more deepened periodontal pockets in all teeth (p = 0.029), than those without dementia. The median number of remaining molars per participant was 3 out of a maximum of 12 (95% CI 3.4–4.0); thus, they often lacked occlusal contact. Conclusions: The LTC residents’ molars were more prone to periodontal problems, whereas their premolars/canines often had caries or were present in the form of root remnants, especially in the participants with dementia. People’s teeth should be treated in an easy-to-maintain way before they move into LTC to enable daily oral hygiene measures and maintain oral health.

Background: Measurement of students' self-regulation skills is an active topic in education research, as effective assessment helps devising support interventions to foster academic achievement. Measures based on event tracing usually require large amounts of data (e.g., MOOCs and large courses), while aptitude measures are often qualitative and need careful interpretation. Precise and interpretable evaluation of self-regulation skills in a normal K-12 classroom thus poses a challenge. Objectives: The present study proposes and explores a learning analytics method of combining aptitude and event measures to evaluate student's self-regulation skills. Methods: An explorative learning analytics study was conducted in a junior high school mathematics class (N = 20 students), using a three-lesson intervention with digital learning materials. Students first assessed their self-regulation skills with a self-report questionnaire, after which trace logs and observations of student behaviour were collected. Learning sessions were extracted from trace logs, clustered, and linked to learning strategies. Students were clustered by the self-report results and learning behaviour profiles. Session clusters, student behaviour clusters and assignment grades were also tested for association. Results and Conclusions: The detected session and student behaviour types were linked to learning tactics and strategies found in prior studies. Additionally, association was found between self-reported self-regulation skills and the student behaviour obtained from trace logs. Implications: The results demonstrate the feasibility of concurrently using aptitude and event measures on a classroom scale, providing teachers with a tool to evaluate and support self-regulated learning. Combined with further measures like predictive learning analytics, teachers can obtain an early and highly interpretable picture of at-risk students in their classes.

Abstract Following a request from the European Commission, EFSA was asked to deliver a scientific opinion on the assessment of the application of renewal of Limosilactobacillus fermentum NCIMB 30169 as a technological feed additive (functional group: silage additives) for all animal species. The applicant has provided evidence that the additive currently on the market complies with the existing terms of the authorisation. The EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) concluded that the additive remains safe for all animal species, consumers, and the environment. Regarding user safety, the additive should be considered a skin and respiratory sensitiser. No conclusions can be drawn on the eye irritancy potential of the additive. There is no need for assessing the efficacy of the additive in the context of the renewal of the authorisation.

Abstract Following a request from the European Commission, EFSA was asked to deliver a scientific opinion on the assessment of the application for renewal of Lentilactobacillus buchneri ATCC PTA‐6138 as a technological additive (functional group: silage additives) for all animal species. The applicant has provided evidence that the additive currently on the market complies with the existing terms of the authorisation. The EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) concluded that the additive remains safe for all animal species, consumers and the environment. Regarding user safety, the additive should be considered as a skin and respiratory sensitiser. No conclusions could be drawn on the eye irritancy potential of the additive. There is no need for assessing the efficacy of the additive in the context of the renewal of the authorisation.

Abstract Following a request from the European Commission, EFSA was asked to deliver a scientific opinion on the assessment of the application for renewal of Lentilactobacillus buchneri ATCC PTA‐2494 as a technological additive (functional group: silage additives) for all animal species. The applicant has provided evidence that the additive currently on the market complies with the existing terms of the authorisation. The EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) concluded that the additive remains safe for all animal species, consumers and the environment. Regarding user safety, the Panel considers that any exposure through skin and respiratory tract is considered a risk. The Panel cannot conclude on the eye irritation potential of the additive due to the lack of data. There is no need for assessing the efficacy of the additive in the context of the renewal of the authorisation.

Abstract Following a request from the European Commission, EFSA was asked to deliver a scientific opinion on the safety and efficacy of 6‐phytase (Quantum® Blue) as a zootechnical feed additive for fin fish. The additive is authorised for use in poultry and pigs. The additive is available in solid and liquid forms, and the 6‐phytase contained in the product is produced by fermentation with a genetically modified strain of Trichoderma reesei. The Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) concluded that the genetic modification of the production strain does not give rise to safety concerns; viable cells of the production strain and its DNA were not detected in the final products. The FEEDAP Panel concluded that, based on the data available, the additive tested is safe for fin fish at the highest recommended level of 2500 phytase activity unit (FTU)/kg complete feed. The Panel concluded that Quantum® Blue is not an irritant to skin and eyes nor a skin sensitiser. Owing to the proteinaceous nature of the active substance, 6‐phytase (Quantum® Blue) is considered a respiratory sensitiser. The use of Quantum® Blue as a feed additive is considered safe for the environment. The additive is considered to be efficacious as a zootechnical additive for salmonids and ornamental fish at 500 FTU/kg complete feed and other fin fish at 2500 FTU/kg complete feed.

Angiosarcoma is a rare soft tissue sarcoma originating from endothelial cells. Given that current treatments for advanced disease have shown limited efficacy, alternative therapies need to be identified. In rare diseases, patient-derived cell models are crucial for screening anti-tumour activity. In this study, cell line models were characterised in 2D and 3D cultures. The cell lines’ growth, migration and invasion capabilities were explored, confirming them as useful tools for preclinical angiosarcoma studies. By screening a drug library, we identified potentially effective compounds: 8-amino adenosine impacted cell growth and inhibited migration and invasion at considerably low concentrations as a single agent. No synergistic effect was detected when combining with paclitaxel, gemcitabine or doxorubicin. These results suggest that this compound could be a potentially useful drug in the treatment of AGS.

IntroductionThis was a secondary analysis aiming to assess whether hydrophilic or hydrophobic statins have a differential effect on urinary retention (UR) and lower urinary tract symptoms (LUTS) in men following a prostate biopsy (PBx), who were at risk for prostate cancer development.MethodsThis was a population-based cohort study with data incorporated from the Institute for Clinical and Evaluative Sciences database to identify all Ontarian men aged 66 and above with a history of a single negative PBx between 1994 and 2016, with no drug prescription history of any of several putative chemo-preventative medications (statins, proton pump inhibitors, five-alpha-reductase inhibitors, and alpha-blockers). Multivariable Cox regression models with time-dependent covariates were used to assess the association of hydrophilic and hydrophobic statins with UR and LUTS within 30 days of a PBx. All models were adjusted for other known putative chemopreventive medications, age, rurality, pharmacologically treated diabetes, comorbidity score, and study inclusion year.ResultsOverall, 21 512 men were included, with a median followup time of 9.4 years (interquartile range [IQR] 5.4-13.4 years). Hydrophobic and hydrophilic statins were initiated by 30.7% and 19.6% of men, respectively, after the first negative PBx. UR and LUTS were experienced by 2.2% and 10% of men, respectively. Cox models demonstrated hydrophilic statins were associated with a lower risk of UR (hazard ratio [HR] 0.56, 95% confidence interval [CI] 0.38-0.83, p=0.0038) and LUTS (HR 0.86, 95% CI 0.76-0.98, p=0.022), while no such association was shown for hydrophobic statins.ConclusionsInitiation of hydrophilic statins in men older than 66 appears to be inversely associated with the risk of UR and LUTS within 30 days of a PBx.

NASA's Global Ecosystem Dynamics Investigation (GEDI) is collecting spaceborne full waveform lidar data with a primary science goal of producing accurate estimates of forest aboveground biomass density (AGBD). This paper presents the development of the models used to create GEDI's footprint-level (~25 m) AGBD (GEDI04_A) product, including a description of the datasets used and the procedure for final model selection. The data used to fit our models are from a compilation of globally distributed spatially and temporally coincident field and airborne lidar datasets, whereby we simulated GEDI-like waveforms from airborne lidar to build a calibration database. We used this database to expand the geographic extent of past waveform lidar studies, and divided the globe into four broad strata by Plant Functional Type (PFT) and six geographic regions. GEDI's waveform-to-biomass models take the form of parametric Ordinary Least Squares (OLS) models with simulated Relative Height (RH) metrics as predictor variables. From an exhaustive set of candidate models, we selected the best input predictor variables, and data transformations for each geographic stratum in the GEDI domain to produce a set of comprehensive predictive footprint-level models. We found that model selection frequently favored combinations of RH metrics at the 98th, 90th, 50th, and 10th height above ground-level percentiles (RH98, RH90, RH50, and RH10, respectively), but that inclusion of lower RH metrics (e.g. RH10) did not markedly improve model performance. Second, forced inclusion of RH98 in all models was important and did not degrade model performance, and the best performing models were parsimonious, typically having only 1-3 predictors. Third, stratification by geographic domain (PFT, geographic region) improved model performance in comparison to global models without stratification. Fourth, for the vast majority of strata, the best performing models were fit using square root transformation of field AGBD and/or height metrics. There was considerable variability in model performance across geographic strata, and areas with sparse training data and/or high AGBD values had the poorest performance. These models are used to produce global predictions of AGBD, but will be improved in the future as more and better training data become available.

IntroductionDespite the association between cannabis use and higher prevalence of suicidal ideation and attempt, the effect of cannabis legalization and regulation in Canada on intentional self-harm has not been determined. MethodsWe used an interrupted time series of population-based rates of emergency department (ED) visits and hospitalizations for intentional self-harm per 100 000 in Ontario and Alberta from January/April 2010 to February 2020. Aggregate monthly counts of ED visits and hospitalizations for intentional self-harm (ICD-10 codes X60–X84, R45.8) were obtained from the National Ambulatory Care Reporting System and Discharge Abstract Database, respectively. ResultsThe legalization and regulation of cannabis in Canada was not significantly associated with a change in rates of ED visits for intentional self-harm in Ontario (level = 0.58, 95% CI: −1.14 to 2.31; trend = −0.17, 95% CI: −0.35 to 0.01) or Alberta (level = −0.06, 95% CI: −2.25 to 2.12; trend = −0.07, 95% CI: −0.27 to 0.13). Hospitalizations for intentional self-harm also remained unchanged in Ontario (level = −0.14, 95% CI: −0.48 to 0.20; trend = 0.01, 95% CI: −0.03 to 0.04) and Alberta (level = −0.41, 95% CI: −1.03 to 0.21; trend = −0.03, 95% CI: −0.08 to 0.03). ConclusionLegalization and regulation of cannabis in Canada has not increased rates of ED visits or hospitalizations for intentional self-harm in Ontario and Alberta. Individual-level analyses that account for demographic characteristics and include other provinces and territories are needed.

IntroductionMalgré l’association connue entre consommation de cannabis et prévalence accrue d’idées suicidaires et de tentatives de suicide, on n’a pas encore mesuré l’effet de la légalisation et de la réglementation du cannabis au Canada sur les lésions auto-infligées. MéthodologieNous avons utilisé des séries chronologiques interrompues des taux de consultation aux urgences et d’hospitalisation pour lésions auto-infligées pour 100 000 habitants en Ontario et en Alberta entre janvier ou avril 2010 et février 2020. Les effectifs mensuels globaux de consultations aux urgences et d’hospitalisations pour lésions auto-infligées (codes X60-X84 et R45.8 selon la CIM-10) ont été obtenus grâce respectivement au Système national d’information sur les soins ambulatoires et à la Base de données sur les congés des patients. RésultatsLa légalisation et la réglementation du cannabis au Canada n’ont pas été associées de manière statistiquement significative à une modification des taux de consultation aux urgences pour lésions auto-infligées en Ontario (niveau = 0,58 avec IC à 95 % : −1,14 à 2,31; tendance = −0,17 avec IC à 95 % : −0,35 à 0,01) ou en Alberta (niveau = −0,06 avec IC à 95 % : −2,25 à 2,12; tendance = −0,07 avec IC à 95 % : −0,27 à 0,13). Les hospitalisations pour lésions auto-infligées sont également demeurées inchangées en Ontario (niveau = −0,14 avec IC à 95 % : −0,48 à 0,20; tendance = 0,01 avec IC à 95 % : −0,03 à 0,04) et en Alberta (niveau = −0,41 avec IC à 95 % : −1,03 à 0,21; tendance = −0,03 avec IC à 95 % : −0,08 à 0,03). ConclusionLa légalisation et la réglementation du cannabis au Canada n’ont pas entraîné d’augmentation des taux de consultation aux urgences ou d’hospitalisation pour lésions auto-infligées en Ontario ou en Alberta. Des analyses à l’échelle individuelle tenant compte des caractéristiques individuelles et incluant d’autres provinces et territoires sont nécessaires.

Technology integration promises better quality in education. This integration is challenging to accomplish, especially for teachers in a developing country like Nigeria where the demand for education remains a struggle in the face of dwindling resources. The technological pedagogical content knowledge (TPACK) framework promotes designing strategies suitable for the teachers' needs. Therefore, in order to determine the readiness of the Nigerian teachers for technology integration, this study examines the Nigerian teacher educators' (N=136) TPACK and the relationship among the constructs using self-completion survey and partial least square techniques. The results reveal that among the seven knowledge constructs, the teachers' technological, pedagogical, and technological pedagogical knowledge are the most significant predictors of their TPACK. The theoretical and practical implications of the result are discussed thereafter.

Abstract Background Falls in long-term care are common. The aim of our study was to explore how medication use is associated with incidence of falls, related consequences, and all-cause mortality among long-term care residents. Methods Five hundred thirty two long-term care residents aged 65 years or older participated in this longitudinal cohort study in 2018–2021. Data on medication use were retrieved from medical records. Polypharmacy was defined as use of 5–10 medications and excessive polypharmacy as use of > 10 medications. The numbers of falls, injuries, fractures, and hospitalizations were collected from medical records over 12 months following baseline assessment. Participants were followed for three years for mortality. All analysis were adjusted for age, sex, Charlson Comorbidity Index, Clinical dementia rating, and mobility. Results A total of 606 falls occurred during the follow-up. Falls increased significantly with the number of medications used. Fall rate was 0.84/person-years (pyrs) (95% CI 0.56 to 1.13) for the non-polypharmacy group, 1.13/pyrs (95% CI 1.01 to 1.26) for the polypharmacy group, and 1.84/pyrs (95% CI 1.60 to 2.09) for the excessive polypharmacy group. Incidence rate ratio for falls was 1.73 (95% CI 1.44 to 2.10) for opioids, 1.48 (95% CI 1.23 to 1.78) for anticholinergic medication, 0.93 (95% CI 0.70 to 1.25) for psychotropics, and 0.91 (95% CI 0.77 to 1.08) for Alzheimer medication. The three-year follow-up showed significant differences in mortality between the groups, the lowest survival rate (25%) being in the excessive polypharmacy group. Conclusion Polypharmacy, opioid and anticholinergic medication use predicted incidence of falls in long-term care. The use of more than 10 medications predicted all-cause mortality. Special attention should be paid to both number and type of medications when prescribing in long-term care.

Remotely sensed data are frequently used for predicting and mapping ecosystem characteristics, and spatially explicit wall-to-wall information is sometimes proposed as the best possible source of information for decision-making. However, wall-to-wall information typically relies on model-based prediction, and several features of model-based prediction should be understood before extensively relying on this type of information. One such feature is that model-based predictors can be considered both unbiased and biased at the same time, which has important implications in several areas of application. In this discussion paper, we first describe the conventional model-unbiasedness paradigm that underpins most prediction techniques using remotely sensed (or other) auxiliary data. From this point of view, model-based predictors are typically unbiased. Secondly, we show that for specific domains, identified based on their true values, the same model-based predictors can be considered biased, and sometimes severely so.We suggest distinguishing between conventional model-bias, defined in the statistical literature as the difference between the expected value of a predictor and the expected value of the quantity being predicted, and design-bias of model-based estimators, defined as the difference between the expected value of a model-based estimator and the true value of the quantity being predicted. We show that model-based estimators (or predictors) are typically design-biased, and that there is a trend in the design-bias from overestimating small true values to underestimating large true values. Further, we give examples of applications where this is important to acknowledge and to potentially make adjustments to correct for the design-bias trend. We argue that relying entirely on conventional model-unbiasedness may lead to mistakes in several areas of application that use predictions from remotely sensed data.