Class Activities Research Articles

Strategi penerapan toleransi beragama di TK PGRI 88 Semarang

AbstractKeywords:Implementation Strategies;Religious Tolerance;Early Childhood; PGRI 88 Semarang Kindergarten is one of the educational institutions that accepts early childhood with different religious backgrounds. For the strategy of implementing religious tolerance, the learning is designed by the religion teacher. This research aims to find out how the strategy of implementing religious tolerance in PGRI 88 Semarang Kindergarten. The research method used in this research is qualitative by collecting data through interviews, observation and documentation. The subject of this research is early childhood aged 5-6 years old who are in the school environment. The results of the study show that PGRI 88 Semarang Kindergarten applies a holistic integrative tolerance implementation strategy, namely by integrating tolerance attitudes in every child's activity both in class and outside the classroom, in order to increase religious tolerance in early childhood. Through the holistic integrative implementation strategy, it is hoped that schools will be able to create a harmonious environment and respect religious differences between children. This can be seen from the increase in children's understanding of religious diversity, the development of tolerance attitudes, and the ability to communicate well with friends with different religious backgrounds.

Resilience of Participatory Communication Adolescent Community Communities in the Prevention of Child Marriage in Sukabumi

This research has an urgency that participatory communication resilience can be formed in communities with young ages, namely adolescents who are members of adolescent communities. This research focuses on how resilience is formed in adolescent communities to prevent child marriage in Sukabumi Regency to improve the ability of adolescents and advocate for vulnerable peers trapped in child marriage. This research uses qualitative research methods with a case study approach to explore the experiences of adolescent members of 5 regional children's forums in Sukabumi as research subjects who actively participate in the prevention of child marriage through the Yes I Do program. The results showed that participatory communication of child communities in preventing child marriage in formal and informal peers showed the collaboration of child communities with diverse cross-communities (heteroglacia). In addition to heteroglacia, Wabinar Activities are open dialogue and sharing experiences and knowledge with peers (dialogue) Movie watching and business class activities are a process to unite some of the differences that exist into an understanding about the impact of child marriage (polypony). Youth Camp and Watch Together activities filled by the children's community are a form of fun participatory communication (carnival). The children's community builds meaning, persistence, self-confidence, patience, independence or what is referred to as peer resilience with peer education activities. Resilience in adolescents who are vulnerable to being trapped in child marriage is an important period and affects the basis of adulthood, individual changes and life development.

An Experiential Journey: A Year of a Professor Using AI in the Classroom and Research

This paper explores the integration of artificial intelligence (AI) in education and research, highlighting its potential to transform traditional teaching methods and enhance academic inquiry. The use of AI in education is examined, focusing on its ability to streamline administrative tasks, provide personalized instruction, and support teacher professional development. The benefits and challenges of AI adoption in the classroom are discussed, including its impact on teacher burnout and the potential disruption of teacher-student relationships. Additionally, the paper explores the applications of AI in scientific research, emphasizing its role in data analysis, literature review, predictive modeling, and collaborative research. The experiences of a professor who integrated AI into their classroom are shared, and the decision-making process, adjustments made, and the importance of balancing AI-driven instruction with interactive in-class activities are discussed. The paper highlights the need for ongoing evaluation, adaptation, and human-AI collaboration to ensure AI's responsible and effective integration in education and research.

Estimating infant age from skull X-ray images using deep learning

This study constructed deep learning models using plain skull radiograph images to predict the accurate postnatal age of infants under 12 months. Utilizing the results of the trained deep learning models, it aimed to evaluate the feasibility of employing major changes visible in skull X-ray images for assessing postnatal cranial development through gradient-weighted class activation mapping. We developed DenseNet-121 and EfficientNet-v2-M convolutional neural network models to analyze 4933 skull X-ray images collected from 1343 infants. Notably, allowing for a ± 1 month error margin, DenseNet-121 reached a maximum corrected accuracy of 79.4% for anteroposterior (AP) views (average: 78.0 ± 1.5%) and 84.2% for lateral views (average: 81.1 ± 2.9%). EfficientNet-v2-M reached a maximum corrected accuracy 79.1% for AP views (average: 77.0 ± 2.3%) and 87.3% for lateral views (average: 85.1 ± 2.5%). Saliency maps identified critical discriminative areas in skull radiographs, including the coronal, sagittal, and metopic sutures in AP skull X-ray images, and the lambdoid suture and cortical bone density in lateral images, marking them as indicators for evaluating cranial development. These findings highlight the precision of deep learning in estimating infant age through non-invasive methods, offering the progress for clinical diagnostics and developmental assessment tools.

Determination of concrete compressive strength from surface images with the integration of CNN and SVR methods

In this study, a new method has been developed using Convolutional Neural Networks (CNN) and Support Vector Regression (SVR) integration to determine the compressive strength of concrete, which is the most common structural material used in the construction sector. The developed CNN model (MPaConvNet) extracts important features from concrete surface images. These extracted features are then used to predict the compressive strength of concrete using the SVR algorithm. The dataset used in the study consists of 1453 concrete surface images obtained from 203 core samples collected from 115 real buildings. The proposed method predicted the compressive strength of concrete with the values of 99.59%, 0.0209, 0.6836, 0.4673, and 1.0000 according to the R2, MAPE, RMSE, MSE, and a-20 metrics, respectively. Additionally, in the study, the regions focused on by the proposed CNN architecture in concrete surface images were detected using the Gradient-weighted Class Activation Mapping algorithm (Grad-CAM).

Faitatala as a Teaching Tool in Science: Contextualizing the Teaching of Science in Samoa

International literature states that the emphasis of science education should be on inquiry learning, where students are encouraged to construct meaningful understanding of scientific concepts through the use of contextualised instructional strategies. In this paper, the author reports on the findings about the use of faitatala as a contextualised instructional tool in the teaching and learning of science in Samoa. Faitatala in this paper specifically highlights two fundamental processes that are relevant to science education in Samoa. The first explores the processes of faitatala (including active listening and speaking) while the second emphasizes the presentation of information. This paper evolved from recent studies in Year 12 chemistry classrooms which formulated the foundation for this study. Learning is an active participation in class activities, active listening and reflective practices supported by constructive and contextualised instructional strategies (Suaalii & Bhattacharya, 2007; Suaalii, 2013) and positive learning environments (Suaalii, 2021). Therefore, this study explored the implementation of a science course (HTE580) at the Faculty of Education, National University of Samoa. Through interviews and talanoa sessions, the research participants shared their perspectives about the use of the faitatala method in the teaching and learning of science. The findings identified constructive and reflective perceptions about the faitatala method as a teaching tool for science.

Artificial intelligence for predicting shockable rhythm during cardiopulmonary resuscitation: In-hospital setting

Aim of the studyThis study aimed to develop an artificial intelligence (AI) model capable of predicting shockable rhythms from electrocardiograms (ECGs) with compression artifacts using real-world data from emergency department (ED) settings. Additionally, we aimed to explore the black box nature of AI models, providing explainability. MethodsThis study is retrospective, observational study using a prospectively collected database. Adult patients who presented to the ED with cardiac arrest or experienced cardiac arrest in the ED between September 2021 and February 2024 were included. ECGs with a compression artifact of 5 s before every rhythm check were used for analysis. The AI model was designed based on convolutional neural networks. The ECG data were assigned into training, validation, and testing sets on a per-patient basis to ensure that ECGs from the same patient did not appear in multiple sets. Gradient-weighted class activation mapping was employed to demonstrate AI explainability. ResultsA total of 1,889 ECGs with compression artifacts from 172 patients were used. The area under the receiver operating characteristic curve (AUROC) for shockable rhythm prediction was 0.8672 (95% confidence interval [CI]: 0.8161–0.9122). The AUROCs for manual and mechanical compression were 0.8771 (95% CI: 0.8054–0.9408) and 0.8466 (95% CI: 0.7630–0.9138), respectively. ConclusionThis study was the first to accurately predict shockable rhythms during compression using an AI model trained with actual patient ECGs recorded during resuscitation. Furthermore, we demonstrated the explainability of the AI. This model can minimize interruption of cardiopulmonary resuscitation and potentially lead to improved outcomes.

Exploring Spectrogram-Based Audio Classification for Parkinson's Disease: A Study on Speech Classification and Qualitative Reliability Verification.

Patients suffering from Parkinson's disease suffer from voice impairment. In this study, we introduce models to classify normal and Parkinson's patients using their speech. We used an AST (audio spectrogram transformer), a transformer-based speech classification model that has recently outperformed CNN-based models in many fields, and a CNN-based PSLA (pretraining, sampling, labeling, and aggregation), a high-performance model in the existing speech classification field, for the study. This study compares and analyzes the models from both quantitative and qualitative perspectives. First, qualitatively, PSLA outperformed AST by more than 4% in accuracy, and the AUC was also higher, with 94.16% for AST and 97.43% for PSLA. Furthermore, we qualitatively evaluated the ability of the models to capture the acoustic features of Parkinson's through various CAM (class activation map)-based XAI (eXplainable AI) models such as GradCAM and EigenCAM. Based on PSLA, we found that the model focuses well on the muffled frequency band of Parkinson's speech, and the heatmap analysis of false positives and false negatives shows that the speech features are also visually represented when the model actually makes incorrect predictions. The contribution of this paper is that we not only found a suitable model for diagnosing Parkinson's through speech using two different types of models but also validated the predictions of the model in practice.

Predicting Rail Corrugation Based on Convolutional Neural Networks Using Vehicle's Acceleration Measurements.

This paper presents a deep learning approach for predicting rail corrugation based on on-board rolling-stock vertical acceleration and forward velocity measurements using One-Dimensional Convolutional Neural Networks (CNN-1D). The model's performance is examined in a 1:10 scale railway system at two different forward velocities. During both the training and test stages, the CNN-1D produced results with mean absolute percentage errors of less than 5% for both forward velocities, confirming its ability to reproduce the corrugation profile based on real-time acceleration and forward velocity measurements. Moreover, by using a Gradient-weighted Class Activation Mapping (Grad-CAM) technique, it is shown that the CNN-1D can distinguish various regions, including the transition from damaged to undamaged regions and one-sided or two-sided corrugated regions, while predicting corrugation. In summary, the results of this study reveal the potential of data-driven techniques such as CNN-1D in predicting rails' corrugation using online data from the dynamics of the rolling-stock, which can lead to more reliable and efficient maintenance and repair of railways.

Applying Of Jigsaw Model To Enhance Students’ Reading Comprehension

The purpose of the research was to know the Jigsaw learning model can enhance on VII grade students’ cognitive learning outcomes in reading comprehension at SMPN 1 Praya Barat. It was participant classroom action research. It was done in two cycles. Each cycle consisted of planning, action, observation, and reflection stages. Subject of the research was 32 students. Instruments used were students' test, students' and teacher’s activities used observation sheets, at the end of each cycle. The research findings show that implementation of Jigsaw learning model can improve students' cognitive competence. Percentage of students' classical achievement in cognitive competence improves time by time. In pre-cycle, the percentage is 46% (not enough), in cycle I is 50% (not enough) and in cycle II is 92% (very good). Meanwhile, percentage of students' Activity also improves. In pre-cycle, the percentage is 52% (enough), in cycle I is 64% (good) and in cycle II is 80% (very good). From the finding, it proves that the implementation of Jigsaw learning model can improve students' cognitive competence and Activity in class VII of SMPN 1 Praya Barat

Four Transformer-Based Deep Learning Classifiers Embedded with an Attention U-Net-Based Lung Segmenter and Layer-Wise Relevance Propagation-Based Heatmaps for COVID-19 X-ray Scans.

Background: Diagnosing lung diseases accurately is crucial for proper treatment. Convolutional neural networks (CNNs) have advanced medical image processing, but challenges remain in their accurate explainability and reliability. This study combines U-Net with attention and Vision Transformers (ViTs) to enhance lung disease segmentation and classification. We hypothesize that Attention U-Net will enhance segmentation accuracy and that ViTs will improve classification performance. The explainability methodologies will shed light on model decision-making processes, aiding in clinical acceptance. Methodology: A comparative approach was used to evaluate deep learning models for segmenting and classifying lung illnesses using chest X-rays. The Attention U-Net model is used for segmentation, and architectures consisting of four CNNs and four ViTs were investigated for classification. Methods like Gradient-weighted Class Activation Mapping plus plus (Grad-CAM++) and Layer-wise Relevance Propagation (LRP) provide explainability by identifying crucial areas influencing model decisions. Results: The results support the conclusion that ViTs are outstanding in identifying lung disorders. Attention U-Net obtained a Dice Coefficient of 98.54% and a Jaccard Index of 97.12%. ViTs outperformed CNNs in classification tasks by 9.26%, reaching an accuracy of 98.52% with MobileViT. An 8.3% increase in accuracy was seen while moving from raw data classification to segmented image classification. Techniques like Grad-CAM++ and LRP provided insights into the decision-making processes of the models. Conclusions: This study highlights the benefits of integrating Attention U-Net and ViTs for analyzing lung diseases, demonstrating their importance in clinical settings. Emphasizing explainability clarifies deep learning processes, enhancing confidence in AI solutions and perhaps enhancing clinical acceptance for improved healthcare results.

Towards Health Promoting Schools: Pilot Training and Evaluation of the Updated Urbani School Health Kit (USHK) among School Teachers in Manila.

Committed to enhancing healthy living, learning, and working conditions, a health-promoting school is a potent influencer of behaviors and habits, reaching families and communities effectively. In the Western Pacific region, the Urbani School Health Kit (USHK) is one of the tools developed to integrate health promotion in schools; however, it needs to be updated to adapt to the evolving health challenges of lifestyle-related diseases. Hence, this study aimed to conduct a pilot training and evaluation of the updated USHK among school teachers in Manila. The USHK was updated with new materials on (1) health-enhancing physical activity; (2) healthy nutrition; and (3) family and community engagement. A two-day training-workshop with 30 school teachers was then conducted to facilitate the integration of the updated USHK in their class activities. We used a multi method evaluation design to assess the implementation of the USHK. Particularly, quantitative data were obtained from the participants' feedback on the toolkit and their knowledge of health-related practices. One month later, field visits were conducted to assess the participants' abilities in utilizing the toolkit into their classes or school activities. Qualitative interviews and classroom observations were also collected post-implementation to determine potential facilitators and barriers to program delivery, and suggestions for improvement. Descriptive statistics were used to summarize participant feedback, while Wilcoxon signed rank test was utilized to determine changes in participant knowledge pre- and post-training. Qualitative data were synthesized through content analysis. Participants provided high satisfaction ratings for the training they received, as well as high scores for the updated USHK, in terms of its appropriateness and acceptability. Significant improvements in participants' overall health promotion knowledge were also noted (Z = -4.456, p <0.001), particularly involving the domains of nutrition (Z = -2.972, p = 0.003), physical activity (Z = -3.564, p <0.001), and family/community engagement (Z = -2.531, p = 0.011). Meanwhile, participants also suggested further improvements in the toolkit to enhance its utilization in the local context. Administrative support was a crucial facilitator for implementing the USHK, while resource limitations were identified as significant barriers. The updated USHK, which provides a more comprehensive health promotion approach for schools, is potentially feasible for implementation in educational institutions in Manila. The toolkit can be utilized by teachers and school nurses to integrate health promotion activities into the school environment and classroom activities. To facilitate its wider uptake and implementation in other schools, government support and resource availability are crucial.

The Role of Local Learning Media in Civics Subjects for Class V Elementary School Students

This research aims to examine the role of local learning media in Civics Education (PKn) subjects for fifth grade students in elementary schools. This research uses a qualitative descriptive method with data collection techniques through observation, interviews, questionnaires and documentation. The research subjects consisted of 30 fifth grade students and one Civics teacher. The research results show that the use of local learning media, such as folklore and traditional games, increases learning motivation, understanding of the material, and student participation in the Civics learning process. As many as 85% of students felt more motivated, and 78% of students understood the material more easily when local learning media was used. In addition, students become more active in class discussions and activities. However, there are several obstacles, such as difficulties in developing teaching materials that are in accordance with local wisdom and limited time and resources. To overcome this obstacle, it is recommended that there be training for teachers and the development of adequate resources. Overall, this research confirms the importance of local learning media in improving the quality of Civics learning and provides practical recommendations for its integration in the curriculum.

From image-level to pixel-level labeling: A weakly-supervised learning method for identifying aquaculture ponds using iterative anti-adversarial attacks guided by aquaculture features

Aquaculture mapping is essential for monitoring and managing aquaculture resources. However, accurately geotargeting individual aquaculture ponds from medium-resolution remote sensing imagery remains challenging, and convolutional deep learning methods for identifying aquaculture ponds require labor-intensive pixel-level annotations. This paper presents a novel weakly-supervised learning method to derive pixel-level labels from image-level annotations for aquaculture ponds. Our approach uses iterative anti-adversarial attacks to refine localization results from multi-scale class activation maps (CAMs). The improved method integrates two regularization methods guided by aquaculture features to form a joint loss function for adversarial samples: discriminative water region suppression and non-aquaculture water class suppression. We also propose an aquaculture feature termed CFNDWI to constrain the localization results and generate high-quality pseudo-labels. As a result, the pseudo-labels are used to train semantic segmentation networks for accurately identifying aquaculture ponds. We evaluated the performance of our method using commonly-used backbones on 10 m Sentinel-2 imagery. Our method achieves Intersection over Union (IoU) values of 0.618–0.655 for pseudo-label generation, and IoU values of 0.664–0.708 for semantic segmentation, outperforming state-of-the-art weakly-supervised methods and public datasets. The effectiveness of each module of our method was also testified through ablation experiments. Our method leverages knowledge-driven aquaculture features to guide the data-driven adversarial learning process, addressing the lack of high-quality aquaculture datasets for model training. The code for implementing our method will be accessible at https://github.com/designer1024/WSLM-AQ.

Deep transfer learning for detection of breast arterial calcifications on mammograms: a comparative study

IntroductionBreast arterial calcifications (BAC) are common incidental findings on routine mammograms, which have been suggested as a sex-specific biomarker of cardiovascular disease (CVD) risk. Previous work showed the efficacy of a pretrained convolutional network (CNN), VCG16, for automatic BAC detection. In this study, we further tested the method by a comparative analysis with other ten CNNs.Material and methodsFour-view standard mammography exams from 1,493 women were included in this retrospective study and labeled as BAC or non-BAC by experts. The comparative study was conducted using eleven pretrained convolutional networks (CNNs) with varying depths from five architectures including Xception, VGG, ResNetV2, MobileNet, and DenseNet, fine-tuned for the binary BAC classification task. Performance evaluation involved area under the receiver operating characteristics curve (AUC-ROC) analysis, F1-score (harmonic mean of precision and recall), and generalized gradient-weighted class activation mapping (Grad-CAM++) for visual explanations.ResultsThe dataset exhibited a BAC prevalence of 194/1,493 women (13.0%) and 581/5,972 images (9.7%). Among the retrained models, VGG, MobileNet, and DenseNet demonstrated the most promising results, achieving AUC-ROCs > 0.70 in both training and independent testing subsets. In terms of testing F1-score, VGG16 ranked first, higher than MobileNet (0.51) and VGG19 (0.46). Qualitative analysis showed that the Grad-CAM++ heatmaps generated by VGG16 consistently outperformed those produced by others, offering a finer-grained and discriminative localization of calcified regions within images.ConclusionDeep transfer learning showed promise in automated BAC detection on mammograms, where relatively shallow networks demonstrated superior performances requiring shorter training times and reduced resources.Relevance statementDeep transfer learning is a promising approach to enhance reporting BAC on mammograms and facilitate developing efficient tools for cardiovascular risk stratification in women, leveraging large-scale mammographic screening programs.Key points• We tested different pretrained convolutional networks (CNNs) for BAC detection on mammograms.• VGG and MobileNet demonstrated promising performances, outperforming their deeper, more complex counterparts.• Visual explanations using Grad-CAM++ highlighted VGG16’s superior performance in localizing BAC.Graphical

Cooking Class Activities: Fostering Independence and Building Self-Confidence in Early Childhood

Learning innovation is very important because it can provide fun and meaningful learning experiences for young children. Cooking class activities in ECE are one application of Project Learning, which is an implementation of the Independent Curriculum that involves children actively in the learning process. This qualitative research aims to discover how ECE teachers design cooking class activities, how students are involved during the activity, and what aspects of development young children achieve through this activity. The research was conducted at the Open Flow Learning Centre, Badung, Bali, using observation, interviews, and documentation methods for 4 ECE teachers and twelve students (three boys and nine girls). Through observations, it was found that cooking class activities at this institution are unique because they involve students from collecting ingredients to handling leftover ingredients that are not used. It was also observed that children had independence and self-confidence when carrying out activities. Considering the many benefits children get from cooking class activities, it is hoped that ECE teachers in other institutions can implement this activity to maximize children's involvement and development.

Student’s Perception On The Use Flipped Classroom In Learning English at The Second Grade Of SMPN 2 X Koto

This research aimed to investigate the students’ perceptions of the flipped classroom approach in learning English at SMPN 2 X Koto Tanah Datar. The flipped classroom model, which reversed traditional in-class activities and homework, introduced new content at home through videos or readings, and dedicated class time to interactive, hands-on activities. This method sought to enhance student engagement, promote active learning, and provide opportunities for personalized instruction and immediate feedback. The study employed a descriptive quantitative method, focusing on second-grade students. Data were collected using a questionnaire that was distributed to 64 students, who were selected through purposive sampling. The questionnaire aimed to gauge students' perceptions of various aspects of the flipped classroom, including self-directed learning, collaboration, responsibility, learning performance, flexibility, and student satisfaction. The population for this study consisted of all second-grade students at SMPN 2 X Koto for the academic year 2021/2022, totaling 95 students across three classes. The sample was selected using purposive sampling, focusing on classes VIII.1 and VIII.2, to ensure participants had the specific attributes necessary for the research. Based on the findings and analysis, it can be concluded that students at SMPN 2 X Koto Tanah Datar have a highly favorable perception of the Flipped Classroom approach in learning English. The study, which involved 95 second-grade students, revealed positive responses across various aspects such as autonomy in learning, collaboration, responsibility, learning performance, flexibility, individualized experiences, interactions, and overall satisfaction. With an average score of 81.41%, categorized as "Very Good" by Arikunto's criteria, the Flipped Classroom methodology is well-received and deemed beneficial for English education at SMPN 2 X Koto Tanah Datar.

Needs Analysis and Design of a Master’s Level Academic Reading Course in English

This article reports an English language needs analysis of Colombian Social Science master’s students. Information from faculty interviews, course syllabi, and student surveys shows that students need English to access and update their disciplinary knowledge through research articles and book chapters to successfully participate in class activities in Spanish, for which low English proficiency and lack of graduate reading skills can be an obstacle. These findings inspired the creation of Reading Research Articles in the Social Sciences course and confirmed the importance of source variety, triangulation, and addressing needs when identified. Methodological and practical contributions are discussed, emphasizing syllabi as key information sources not considered in other needs analyses.

Patient-ventilator asynchrony classification in mechanically ventilated patients: Model-based or machine learning method?

Background and objectivePatient-ventilator asynchrony (PVA) is associated with poor clinical outcomes and remains under-monitored. Automated PVA detection would enable complete monitoring standard observational methods do not allow. While model-based and machine learning PVA approaches exist, they have variable performance and can miss specific PVA events. This study compares a model and rule-based algorithm with a machine learning PVA method by retrospectively validating both methods using an independent patient cohort. MethodsHysteresis loop analysis (HLA) which is a rule-based method (RBM) and a tri-input convolutional neural network (TCNN) machine learning model are used to classify 7 different types of PVA, including: 1) flow asynchrony; 2) reverse triggering; 3) premature cycling; 4) double triggering; 5) delayed cycling; 6) ineffective efforts; and 7) auto triggering. Class activation mapping (CAM) heatmaps visualise sections of respiratory waveforms the TCNN model uses for decision making, improving result interpretability. Both PVA classification methods were used to classify incidence in an independent retrospective clinical cohort of 11 mechanically ventilated patients for validation and performance comparison. ResultsSelf-validation with the training dataset shows overall better HLA performance (accuracy, sensitivity, specificity: 97.5 %, 96.6 %, 98.1 %) compared to the TCNN model (accuracy, sensitivity, specificity: 89.5 %, 98.3 %, 83.9 %). In this study, the TCNN model demonstrates higher sensitivity in detecting PVA, but HLA was better at identifying non-PVA breathing cycles due to its rule-based nature. While the overall AI identified by both classification methods are very similar, the intra-patient distribution of each PVA type varies between HLA and TCNN. ConclusionThe collective findings underscore the efficacy of both HLA and TCNN in PVA detection, indicating the potential for real-time continuous monitoring of PVA. While ML methods such as TCNN demonstrate good PVA identification performance, it is essential to ensure optimal model architecture and diversity in training data before widespread uptake as standard care. Moving forward, further validation and adoption of RBM methods, such as HLA, offers an effective approach to PVA detection while providing clear distinction into the underlying patterns of PVA, better aligning with clinical needs for transparency, explicability, adaptability and reliability of these emerging tools for clinical care.

Assessing fidelity in XAI post-hoc techniques: A comparative study with ground truth explanations datasets

The evaluation of the fidelity of eXplainable Artificial Intelligence (XAI) methods to their underlying models is a challenging task, primarily due to the absence of a ground truth for explanations. However, assessing fidelity is a necessary step for ensuring a correct XAI methodology. In this study, we conduct a fair and objective comparison of the current state-of-the-art XAI methods by introducing three novel image datasets with reliable ground truth for explanations. The primary objective of this comparison is to identify methods with low fidelity and eliminate them from further research, thereby promoting the development of more trustworthy and effective XAI techniques. Our results demonstrate that XAI methods based on the direct gradient calculation and the backpropagation of output information to input yield higher accuracy and reliability compared to methods relying on perturbation based or Class Activation Maps (CAM). However, these methods tend to generate more noisy saliency maps. These findings have significant implications for the advancement of XAI methods, enabling the elimination of erroneous explanations and fostering the development of more robust and reliable XAI.