Fabio Caraffini

Computational nutrition needs precise ingredient data, but current databases are incomplete, inconsistent, and built for human reference rather than automated reasoning. LLMs could help fill these gaps, but single-pass outputs are unreliable and can introduce silent errors into downstream computation. We present a quality-controlled LLM pipeline for ingredient data acquisition that combines robust statistical estimation, domain-specific invariant checks, and a web-fetch fallback. An illustrative Heap's Law fit to 233 recipes suggests that unique-ingredient growth is sub-linear and front-loaded: the projected ratio of unique ingredients to recipes falls from 1.74 at 100 recipes to 0.19 at 5,000. For each ingredient attribute, repeated LLM queries are treated as samples from a model-induced answer distribution, and we apply robust point estimators and normalised confidence scores across numerical, Boolean, multiple-choice, open categorical, and optional integer types. An invariant guard layer enforces nutritional and logical self-consistency within each ingredient record. Minor numeric inconsistencies are reconciled via a linear program that minimises worst-case percentage deviation while preserving semantic zeros, and major violations are escalated to web-evidence-grounded repair, then human review only if that fails. On a curated 30-ingredient reference set, the pipeline achieves 98.4% exact match on nutrient flags and cuts median absolute percentage error on nutrient ratios from 31.9% for the median-aggregated baseline to 10.1%, a reduction of 21.8 percentage points, at an API cost of about $1 per ingredient. This frames LLM-assisted database construction as a controlled data-engineering workflow that makes uncertainty operational rather than discarding it.

This repository contains the code used in the study "Deep Learning–Based Detection of Epilepsy Biomarkers from Intracranial EEG Using Time–Frequency Distributions" submitted to PLOS. The code implements methods for extracting and analyzing epilepsy biomarkers from EEG data. Instructions for reproducing the results are provided in the README file. The original development repository is available on GitHub:https://github.com/neural-bit/epilepsy_biomarkers

Fluorescence microscopy is widely employed for the analysis of living biological samples; however, the utility of the resulting recordings is frequently constrained by noise, temporal variability, and inconsistent visualisation of signals that oscillate over time. We present a unique computational framework that integrates information from multiple time-resolved frames into a single high-quality image, while preserving the underlying biological content of the original video. We evaluate the proposed method through an extensive number of configurations (n = 111) and on a challenging dataset comprising dynamic, heterogeneous, and morphologically complex 2D monolayers of cardiac cells. Results show that our framework, which consists of a combination of explainable techniques from different computer vision application fields, is capable of generating composite images that preserve and enhance the quality and information of individual microscopy frames, yielding 44% average increase in cell count compared to previous methods. The proposed pipeline is applicable to other imaging domains that require the fusion of multi-temporal image stacks into high-quality 2D images, thereby facilitating annotation and downstream segmentation.

Repository Overview This Zenodo repository contains the supplementary materials supporting the study: “Multi-Temporal Frames Projection for Dynamic Processes Fusion in Fluorescence Microscopy”. The files included are: framesFusion.py Source code implementing the proposed framework for multi-temporal frames fusion. A link to the live GitHub repository is also provided below for version tracking and updates. Example_HL-1_(fluorescence).aviSample fluorescence microscopy data used as input for the algorithm. This is one of the original fluorescence microscopy input files from the microscope. Example_HL-1_(fluorescence).mp4 This file is a compressed .mp4 version of the original .avi file, provided for easier preview and broader compatibility across modern devices and platforms. While more convenient for viewing, it retains less information than the .avi file, and we do not recommend using it for testing (in the context of our study), despite being an acceptable format for our framework. Example_HL-1_(bright-field).mp4Companion bright-field video provided for interpretative purposes only (to visualise cell morphology and topology). Note: This file is not an input for our algorithm. SP_vs_AP_t-test.pdfDetailed statistical comparison of Sum Projection (SP) versus Average Projection (AP) across all processing method combinations described in the study. How to Use Use the Python file (famesFusion.py) together with the fluorescence sample video (Example_HL-1_(fluorescence).avi) to reproduce the steps and results described in the article. Detailed "How To" instructions (for any OpenCV-compatible dataset) are available in the GitHub repository. If you use this code or any material in this repository, please cite the associated work (DOI: 10.48550/arXiv.2601.10392) alongside the repository in line with the CC BY 4.0 principles.

Detecting fake news is essential in natural language processing to verify news authenticity and prevent misinformation-driven social, political, and economic disruptions targeting specific groups. A major challenge in multimodal fake news detection is effectively integrating textual and visual modalities, as semantic gaps and contextual variations between images and text complicate alignment, interpretation, and the detection of subtle or blatant inconsistencies. To enhance accuracy in fake news detection, this article introduces an ensemble-based framework that integrates textual and visual data using ViLBERT’s two-stream architecture, incorporates VADER sentiment analysis to detect emotional language, and uses Image–Text Contextual Similarity to identify mismatches between visual and textual elements. These features are processed through the Bi-GRU classifier, Transformer-XL, DistilBERT, and XLNet, combined via a stacked ensemble method with soft voting, culminating in a T5 metaclassifier that predicts the outcome for robustness. Results on the Fakeddit and Weibo benchmarking datasets show that our method outperforms state-of-the-art models, achieving up to 96% and 94% accuracy in fake news detection, respectively. This study highlights the necessity for advanced multimodal fake news detection systems to address the increasing complexity of misinformation and offers a promising solution.

Fluorescence microscopy is widely employed for the analysis of living biological samples; however, the utility of the resulting recordings is frequently constrained by noise, temporal variability, and inconsistent visualisation of signals that oscillate over time. We present a unique computational framework that integrates information from multiple time-resolved frames into a single high-quality image, while preserving the underlying biological content of the original video. We evaluate the proposed method through an extensive number of configurations (n = 111) and on a challenging dataset comprising dynamic, heterogeneous, and morphologically complex 2D monolayers of cardiac cells. Results show that our framework, which consists of a combination of explainable techniques from different computer vision application fields, is capable of generating composite images that preserve and enhance the quality and information of individual microscopy frames, yielding 44% average increase in cell count compared to previous methods. The proposed pipeline is applicable to other imaging domains that require the fusion of multi-temporal image stacks into high-quality 2D images, thereby facilitating annotation and downstream segmentation.

Repository Overview This Zenodo repository contains the supplementary materials supporting the study: “Multi-Temporal Frames Projection for Dynamic Processes Fusion in Fluorescence Microscopy”. The files included are: framesFusion.py Source code implementing the proposed framework for multi-temporal frames fusion. A link to the live GitHub repository is also provided below for version tracking and updates. Example_HL-1_(fluorescence).aviSample fluorescence microscopy data used as input for the algorithm. This is one of the original fluorescence microscopy input files from the microscope. Example_HL-1_(fluorescence).mp4 This file is a compressed .mp4 version of the original .avi file, provided for easier preview and broader compatibility across modern devices and platforms. While more convenient for viewing, it retains less information than the .avi file, and we do not recommend using it for testing (in the context of our study), despite being an acceptable format for our framework. Example_HL-1_(bright-field).mp4Companion bright-field video provided for interpretative purposes only (to visualise cell morphology and topology). Note: This file is not an input for our algorithm. SP_vs_AP_t-test.pdfDetailed statistical comparison of Sum Projection (SP) versus Average Projection (AP) across all processing method combinations described in the study. How to Use Use the Python file (famesFusion.py) together with the fluorescence sample video (Example_HL-1_(fluorescence).avi) to reproduce the steps and results described in the article. Detailed "How To" instructions (for any OpenCV-compatible dataset) are available in the GitHub repository. If you use this code or any material in this repository, please cite the associated work (DOI: 10.48550/arXiv.2601.10392) alongside the repository in line with the CC BY 4.0 principles.

Computational nutrition needs precise ingredient data, but current databases are incomplete, inconsistent, and built for human reference rather than automated reasoning. LLMs could help fill these gaps, but single-pass outputs are unreliable and can introduce silent errors into downstream computation. We present a quality-controlled LLM pipeline for ingredient data acquisition that combines robust statistical estimation, domain-specific invariant checks, and a web-fetch fallback. An illustrative Heap's Law fit to 233 recipes suggests that unique-ingredient growth is sub-linear and front-loaded: the projected ratio of unique ingredients to recipes falls from 1.74 at 100 recipes to 0.19 at 5,000. For each ingredient attribute, repeated LLM queries are treated as samples from a model-induced answer distribution, and we apply robust point estimators and normalised confidence scores across numerical, Boolean, multiple-choice, open categorical, and optional integer types. An invariant guard layer enforces nutritional and logical self-consistency within each ingredient record. Minor numeric inconsistencies are reconciled via a linear program that minimises worst-case percentage deviation while preserving semantic zeros, and major violations are escalated to web-evidence-grounded repair, then human review only if that fails. On a curated 30-ingredient reference set, the pipeline achieves 98.4% exact match on nutrient flags and cuts median absolute percentage error on nutrient ratios from 31.9% for the median-aggregated baseline to 10.1%, a reduction of 21.8 percentage points, at an API cost of about $1 per ingredient. This frames LLM-assisted database construction as a controlled data-engineering workflow that makes uncertainty operational rather than discarding it.

Computational nutrition needs precise ingredient data, but current databases are incomplete, inconsistent, and built for human reference rather than automated reasoning. LLMs could help fill these gaps, but single-pass outputs are unreliable and can introduce silent errors into downstream computation. We present a quality-controlled LLM pipeline for ingredient data acquisition that combines robust statistical estimation, domain-specific invariant checks, and a web-fetch fallback. An illustrative Heap's Law fit to 233 recipes suggests that unique-ingredient growth is sub-linear and front-loaded: the projected ratio of unique ingredients to recipes falls from 1.74 at 100 recipes to 0.19 at 5,000. For each ingredient attribute, repeated LLM queries are treated as samples from a model-induced answer distribution, and we apply robust point estimators and normalised confidence scores across numerical, Boolean, multiple-choice, open categorical, and optional integer types. An invariant guard layer enforces nutritional and logical self-consistency within each ingredient record. Minor numeric inconsistencies are reconciled via a linear program that minimises worst-case percentage deviation while preserving semantic zeros, and major violations are escalated to web-evidence-grounded repair, then human review only if that fails. On a curated 30-ingredient reference set, the pipeline achieves 98.4% exact match on nutrient flags and cuts median absolute percentage error on nutrient ratios from 31.9% for the median-aggregated baseline to 10.1%, a reduction of 21.8 percentage points, at an API cost of about $1 per ingredient. This frames LLM-assisted database construction as a controlled data-engineering workflow that makes uncertainty operational rather than discarding it.

This repository contains the code used in the study "Deep Learning–Based Detection of Epilepsy Biomarkers from Intracranial EEG Using Time–Frequency Distributions" submitted to PLOS. The code implements methods for extracting and analyzing epilepsy biomarkers from EEG data. Instructions for reproducing the results are provided in the README file. The original development repository is available on GitHub:https://github.com/neural-bit/epilepsy_biomarkers

This repository contains the code used in the study "Deep Learning–Based Detection of Epilepsy Biomarkers from Intracranial EEG Using Time–Frequency Distributions" submitted to PLOS. The code implements methods for extracting and analyzing epilepsy biomarkers from EEG data. Instructions for reproducing the results are provided in the README file. The original development repository is available on GitHub:https://github.com/neural-bit/epilepsy_biomarkers

Detecting fake news is essential in natural language processing to verify news authenticity and prevent misinformation-driven social, political, and economic disruptions targeting specific groups. A major challenge in multimodal fake news detection is effectively integrating textual and visual modalities, as semantic gaps and contextual variations between images and text complicate alignment, interpretation, and the detection of subtle or blatant inconsistencies. To enhance accuracy in fake news detection, this article introduces an ensemble-based framework that integrates textual and visual data using ViLBERT’s two-stream architecture, incorporates VADER sentiment analysis to detect emotional language, and uses Image–Text Contextual Similarity to identify mismatches between visual and textual elements. These features are processed through the Bi-GRU classifier, Transformer-XL, DistilBERT, and XLNet, combined via a stacked ensemble method with soft voting, culminating in a T5 metaclassifier that predicts the outcome for robustness. Results on the Fakeddit and Weibo benchmarking datasets show that our method outperforms state-of-the-art models, achieving up to 96% and 94% accuracy in fake news detection, respectively. This study highlights the necessity for advanced multimodal fake news detection systems to address the increasing complexity of misinformation and offers a promising solution.

Fluorescence microscopy is widely employed for the analysis of living biological samples; however, the utility of the resulting recordings is frequently constrained by noise, temporal variability, and inconsistent visualisation of signals that oscillate over time. We present a unique computational framework that integrates information from multiple time-resolved frames into a single high-quality image, while preserving the underlying biological content of the original video. We evaluate the proposed method through an extensive number of configurations (n = 111) and on a challenging dataset comprising dynamic, heterogeneous, and morphologically complex 2D monolayers of cardiac cells. Results show that our framework, which consists of a combination of explainable techniques from different computer vision application fields, is capable of generating composite images that preserve and enhance the quality and information of individual microscopy frames, yielding 44% average increase in cell count compared to previous methods. The proposed pipeline is applicable to other imaging domains that require the fusion of multi-temporal image stacks into high-quality 2D images, thereby facilitating annotation and downstream segmentation.

Fluorescence microscopy is widely employed for the analysis of living biological samples; however, the utility of the resulting recordings is frequently constrained by noise, temporal variability, and inconsistent visualisation of signals that oscillate over time. We present a unique computational framework that integrates information from multiple time-resolved frames into a single high-quality image, while preserving the underlying biological content of the original video. We evaluate the proposed method through an extensive number of configurations (n = 111) and on a challenging dataset comprising dynamic, heterogeneous, and morphologically complex 2D monolayers of cardiac cells. Results show that our framework, which consists of a combination of explainable techniques from different computer vision application fields, is capable of generating composite images that preserve and enhance the quality and information of individual microscopy frames, yielding 44% average increase in cell count compared to previous methods. The proposed pipeline is applicable to other imaging domains that require the fusion of multi-temporal image stacks into high-quality 2D images, thereby facilitating annotation and downstream segmentation.

Repository Overview This Zenodo repository contains the supplementary materials supporting the study: “Multi-Temporal Frames Projection for Dynamic Processes Fusion in Fluorescence Microscopy”. The files included are: framesFusion.py Source code implementing the proposed framework for multi-temporal frames fusion. A link to the live GitHub repository is also provided below for version tracking and updates. Example_HL-1_(fluorescence).aviSample fluorescence microscopy data used as input for the algorithm. This is one of the original fluorescence microscopy input files from the microscope. Example_HL-1_(fluorescence).mp4 This file is a compressed .mp4 version of the original .avi file, provided for easier preview and broader compatibility across modern devices and platforms. While more convenient for viewing, it retains less information than the .avi file, and we do not recommend using it for testing (in the context of our study), despite being an acceptable format for our framework. Example_HL-1_(bright-field).mp4Companion bright-field video provided for interpretative purposes only (to visualise cell morphology and topology). Note: This file is not an input for our algorithm. SP_vs_AP_t-test.pdfDetailed statistical comparison of Sum Projection (SP) versus Average Projection (AP) across all processing method combinations described in the study. How to Use Use the Python file (famesFusion.py) together with the fluorescence sample video (Example_HL-1_(fluorescence).avi) to reproduce the steps and results described in the article. Detailed "How To" instructions (for any OpenCV-compatible dataset) are available in the GitHub repository. If you use this code or any material in this repository, please cite the associated work (DOI: 10.48550/arXiv.2601.10392) alongside the repository in line with the CC BY 4.0 principles.

Repository Overview This Zenodo repository contains the supplementary materials supporting the study: “Multi-Temporal Frames Projection for Dynamic Processes Fusion in Fluorescence Microscopy”. The files included are: framesFusion.py Source code implementing the proposed framework for multi-temporal frames fusion. A link to the live GitHub repository is also provided below for version tracking and updates. Example_HL-1_(fluorescence).aviSample fluorescence microscopy data used as input for the algorithm. This is one of the original fluorescence microscopy input files from the microscope. Example_HL-1_(fluorescence).mp4 This file is a compressed .mp4 version of the original .avi file, provided for easier preview and broader compatibility across modern devices and platforms. While more convenient for viewing, it retains less information than the .avi file, and we do not recommend using it for testing (in the context of our study), despite being an acceptable format for our framework. Example_HL-1_(bright-field).mp4Companion bright-field video provided for interpretative purposes only (to visualise cell morphology and topology). Note: This file is not an input for our algorithm. SP_vs_AP_t-test.pdfDetailed statistical comparison of Sum Projection (SP) versus Average Projection (AP) across all processing method combinations described in the study. How to Use Use the Python file (famesFusion.py) together with the fluorescence sample video (Example_HL-1_(fluorescence).avi) to reproduce the steps and results described in the article. Detailed "How To" instructions (for any OpenCV-compatible dataset) are available in the GitHub repository. If you use this code or any material in this repository, please cite the associated work (DOI: 10.48550/arXiv.2601.10392) alongside the repository in line with the CC BY 4.0 principles.

This paper presents a joint learning framework for fake news detection, introducing an Enhanced BERT model that integrates named entity recognition, relational feature classification, and Stance Detection through a unified multi-task approach. The model incorporates task-specific masking and hierarchical attention mechanisms to capture both fine-grained and high-level contextual relationships across headlines and body text. Cross-task consistency losses are applied to ensure coherence and alignment with external factual knowledge. We analyse the average distance from components to the centroid of a news sample to differentiate genuine information from falsehoods in large-scale text data effectively. Experiments on two FakeNewsNet datasets show that our framework outperforms state-of-the-art models, with accuracy improvements of 2.17% and 1.03%. These results indicate the potential for applications needing detailed text processing, like automatic summarisation and misinformation detection.

Cardiac cell network segmentation is uniquely challenging because cardiomyocytes, unlike other cell types, form morphologically complex multicellular structures, causing generalist models like Cellpose to oversegment and perform inaccurately. We use our unique live cell imaging dataset of self-organised HL-1 networks to propose and assess various algorithmic configurations based on combinations of the Cellpose model and the Segment Anything Model, equipped with multiple pre- and post-processing routines. Our results demonstrate the advantages of integrating equalisation-based pre-processing with median filtering, fine-tuning Cellpose, and incorporating our post-processing routine into the segmentation pipeline, achieving up to 85% accuracy, 96% recall, 91% DICE, and 88% precision, while mitigating oversegmentation.

ABSTRACT This study presents a novel approach to crop mapping using remotely sensed satellite images. It addresses the significant classification modelling challenges, including (1) the requirements for extensive labelled data and (2) the complex optimisation problem for selection of appropriate temporal windows in the absence of prior knowledge of cultivation calendars. We compare the lightweight Dynamic Time Warping (DTW) classification method with the heavily supervised Convolutional Neural Network ‐ Long Short‐Term Memory (CNN‐LSTM) using high‐resolution multispectral optical satellite imagery (3 m/pixel). Our approach integrates effective practical preprocessing steps, including data augmentation and a data‐driven optimisation strategy for the temporal window, even in the presence of numerous crop classes. Our findings demonstrate that DTW, despite its lower data demands, can match the performance of CNN‐LSTM through our effective preprocessing steps while significantly improving runtime. These results demonstrate that both CNN‐LSTM and DTW can achieve deployment‐level accuracy and underscore the potential of DTW as a viable alternative to more resource‐intensive models. The results also prove the effectiveness of temporal windowing for improving runtime and accuracy of a crop classification study, even with no prior knowledge of planting timeframes.

This review investigates the application of Explainable Artificial Intelligence (XAI) in biomedical informatics, encompassing domains such as medical imaging, genomics, and electronic health records. Through a systematic analysis of 43 peer-reviewed articles, we examine current trends, as well as the strengths and limitations of methodologies currently used in real-world healthcare settings. Our findings highlight a growing interest in XAI, particularly in medical imaging, yet reveal persistent challenges in clinical adoption, including issues of trust, interpretability, and integration into decision-making workflows. We identify critical gaps in existing approaches and underscore the need for more robust, human-centred, and intrinsically interpretable models, with only 44% of the papers studied proposing human-centred validations. Furthermore, we argue that fairness and accountability, which are key to the acceptance of AI in clinical practice, can be supported by the use of post hoc tools for identifying potential biases but ultimately require the implementation of complementary fairness-aware or causal approaches alongside evaluation frameworks that prioritise clinical relevance and user trust. This review provides a foundation for advancing XAI research on the development of more transparent, equitable, and clinically meaningful AI systems for use in healthcare.

This work contributes towards balancing the inclusivity and global applicability of natural language processing techniques by proposing the first 'name entity recognition' dataset for Kurdish Sorani, a low-resource and under-represented language, that consists of 64,563 annotated tokens. It also provides a tool for facilitating this task in this and many other languages and performs a thorough comparative analysis, including classic machine learning models and neural systems. The results obtained challenge established assumptions about the advantage of neural approaches within the context of NLP. Conventional methods, in particular CRF, obtain F1-scores of 0.825, outperforming the results of BiLSTM-based models (0.706) significantly. These findings indicate that simpler and more computationally efficient classical frameworks can outperform neural architectures in low-resource settings.

Cardiac cell network segmentation is uniquely challenging because cardiomyocytes, unlike other cell types, form morphologically complex multicellular structures, causing generalist models like Cellpose to oversegment and perform inaccurately. We use our unique live cell imaging dataset of self-organised HL-1 networks to propose and assess various algorithmic configurations based on combinations of the Cellpose model and the Segment Anything Model, equipped with multiple pre- and post-processing routines. Our results demonstrate the advantages of integrating equalisation-based pre-processing with median filtering, fine-tuning Cellpose, and incorporating our post-processing routine into the segmentation pipeline, achieving up to 85% accuracy, 96% recall, 91% DICE, and 88% precision, while mitigating oversegmentation.

Abstract Deep learning models have been instrumental in extracting critical indicators for breast cancer diagnosis - the prevalent malignancy among women worldwide - from baseline magnetic resonance imaging. However, many existing models do not fully leverage the rich spatial information available in the 3D structure of medical imaging data, potentially overlooking important contextual details. This develops an explainable deep learning framework for classifying breast cancer that leverages the complete 3D and provides classification results alongside visual explanations of the decision-making process. The preprocessing pipeline is fed with 3D sequences containing ‘tumour’ and ‘non-tumour’ regions. It includes a 3D Adaptive Unsharp Mask (AUM) filter to reduce noise and augment image class, followed by normalisation and data augmentation. Classification is then achieved by training an augmented ResNet150 model. Three explainable artificial intelligence (XAI) techniques, including Shapley Additive Explanations, 3D Gradient-Weighted Class Activation Mapping, and Contextual Importance and Utility, are employed to provide improved interpretability. The model demonstrates state-of-the-art performance over the QIN-BREAST dataset, achieving testing accuracies of 98.861% for ‘tumours’ and 99.447% for ‘non-tumours’, as well as over the Duke Breast Cancer Dataset, where it achieves 99.104% for ‘tumours’ and 99.753% for ‘non-tumours’, while offering enhanced interpretability through XAI methods.

The rapid advancement of large language models has significantly enhanced the quality of AI-generated text, making it increasingly difficult for detection systems to distinguish from human-written content. Existing detection methods, such as statistical, linguistic, machine learning, and deep learning approaches, often exhibit a decline in performance when applied to new or previously unseen large language models. Additionally, they tend to become outdated due to their static frameworks and inability to adapt to emerging patterns in generative text. To address this limitation, we introduce a novel dynamic fusion framework that integrates multi-scale feature fusion to capture diverse text patterns and employs continual learning with Elastic Weight Consolidation (EWC) to adapt to new models while mitigating catastrophic forgetting. This is the first attempt, to the best of our knowledge, to develop such a dynamic framework for AI-generated text detection. Evaluated on the TuringBench and DeepfakeTextDetect benchmark datasets, our framework achieves an average accuracy of 95.78% and 92.39%, outperforming the standard model by 5.88% and 7.98%, respectively, in distinguishing AI-generated from human-written text across various language generative model architectures. The continual learning ensures that the model remains adaptive and accurate over time, which is essential for practical applications in dynamic environments. This dynamic and adaptive approach paves the way for resilient AI-generated text detection systems capable of evolving alongside the rapidly advancing landscape of generative language technologies.

This work contributes towards balancing the inclusivity and global applicability of natural language processing techniques by proposing the first 'name entity recognition' dataset for Kurdish Sorani, a low-resource and under-represented language, that consists of 64,563 annotated tokens. It also provides a tool for facilitating this task in this and many other languages and performs a thorough comparative analysis, including classic machine learning models and neural systems. The results obtained challenge established assumptions about the advantage of neural approaches within the context of NLP. Conventional methods, in particular CRF, obtain F1-scores of 0.825, outperforming the results of BiLSTM-based models (0.706) significantly. These findings indicate that simpler and more computationally efficient classical frameworks can outperform neural architectures in low-resource settings.

Efficient train scheduling on single-track railways represents a significant challenge due to operational constraints. Multiple trains share the same track and an optimal schedule must ensure that trains traveling in opposite directions do not collide while minimising delays. This paper proposes a novel approach to address this problem by formulating train scheduling as a distributed constraint satisfaction problem and applying a multi-agent system to solve it. We propose a simple, yet efficient, system in which agents cooperate to schedule trains on a single-track railway. The result shows that our system is reliable and fast in comparison to other popular approaches.

The rapid advancement of large language models has significantly enhanced the quality of AI-generated text, making it increasingly difficult for detection systems to distinguish from human-written content. Existing detection methods, such as statistical, linguistic, machine learning, and deep learning approaches, often exhibit a decline in performance when applied to new or previously unseen large language models. Additionally, they tend to become outdated due to their static frameworks and inability to adapt to emerging patterns in generative text. To address this limitation, we introduce a novel dynamic fusion framework that integrates multi-scale feature fusion to capture diverse text patterns and employs continual learning with Elastic Weight Consolidation (EWC) to adapt to new models while mitigating catastrophic forgetting. This is the first attempt, to the best of our knowledge, to develop such a dynamic framework for AI-generated text detection. Evaluated on the TuringBench and DeepfakeTextDetect benchmark datasets, our framework achieves an average accuracy of 95.78% and 92.39%, outperforming the standard model by 5.88% and 7.98%, respectively, in distinguishing AI-generated from human-written text across various language generative model architectures. The continual learning ensures that the model remains adaptive and accurate over time, which is essential for practical applications in dynamic environments. This dynamic and adaptive approach paves the way for resilient AI-generated text detection systems capable of evolving alongside the rapidly advancing landscape of generative language technologies.

This paper presents a joint learning framework for fake news detection, introducing an Enhanced BERT model that integrates named entity recognition, relational feature classification, and Stance Detection through a unified multi-task approach. The model incorporates task-specific masking and hierarchical attention mechanisms to capture both fine-grained and high-level contextual relationships across headlines and body text. Cross-task consistency losses are applied to ensure coherence and alignment with external factual knowledge. We analyse the average distance from components to the centroid of a news sample to differentiate genuine information from falsehoods in large-scale text data effectively. Experiments on two FakeNewsNet datasets show that our framework outperforms state-of-the-art models, with accuracy improvements of 2.17% and 1.03%. These results indicate the potential for applications needing detailed text processing, like automatic summarisation and misinformation detection.

This review investigates the application of Explainable Artificial Intelligence (XAI) in biomedical informatics, encompassing domains such as medical imaging, genomics, and electronic health records. Through a systematic analysis of 43 peer-reviewed articles, we examine current trends, as well as the strengths and limitations of methodologies currently used in real-world healthcare settings. Our findings highlight a growing interest in XAI, particularly in medical imaging, yet reveal persistent challenges in clinical adoption, including issues of trust, interpretability, and integration into decision-making workflows. We identify critical gaps in existing approaches and underscore the need for more robust, human-centred, and intrinsically interpretable models, with only 44% of the papers studied proposing human-centred validations. Furthermore, we argue that fairness and accountability, which are key to the acceptance of AI in clinical practice, can be supported by the use of post hoc tools for identifying potential biases but ultimately require the implementation of complementary fairness-aware or causal approaches alongside evaluation frameworks that prioritise clinical relevance and user trust. This review provides a foundation for advancing XAI research on the development of more transparent, equitable, and clinically meaningful AI systems for use in healthcare.

ABSTRACT This study presents a novel approach to crop mapping using remotely sensed satellite images. It addresses the significant classification modelling challenges, including (1) the requirements for extensive labelled data and (2) the complex optimisation problem for selection of appropriate temporal windows in the absence of prior knowledge of cultivation calendars. We compare the lightweight Dynamic Time Warping (DTW) classification method with the heavily supervised Convolutional Neural Network ‐ Long Short‐Term Memory (CNN‐LSTM) using high‐resolution multispectral optical satellite imagery (3 m/pixel). Our approach integrates effective practical preprocessing steps, including data augmentation and a data‐driven optimisation strategy for the temporal window, even in the presence of numerous crop classes. Our findings demonstrate that DTW, despite its lower data demands, can match the performance of CNN‐LSTM through our effective preprocessing steps while significantly improving runtime. These results demonstrate that both CNN‐LSTM and DTW can achieve deployment‐level accuracy and underscore the potential of DTW as a viable alternative to more resource‐intensive models. The results also prove the effectiveness of temporal windowing for improving runtime and accuracy of a crop classification study, even with no prior knowledge of planting timeframes.

Abstract Deep learning models have been instrumental in extracting critical indicators for breast cancer diagnosis - the prevalent malignancy among women worldwide - from baseline magnetic resonance imaging. However, many existing models do not fully leverage the rich spatial information available in the 3D structure of medical imaging data, potentially overlooking important contextual details. This develops an explainable deep learning framework for classifying breast cancer that leverages the complete 3D and provides classification results alongside visual explanations of the decision-making process. The preprocessing pipeline is fed with 3D sequences containing ‘tumour’ and ‘non-tumour’ regions. It includes a 3D Adaptive Unsharp Mask (AUM) filter to reduce noise and augment image class, followed by normalisation and data augmentation. Classification is then achieved by training an augmented ResNet150 model. Three explainable artificial intelligence (XAI) techniques, including Shapley Additive Explanations, 3D Gradient-Weighted Class Activation Mapping, and Contextual Importance and Utility, are employed to provide improved interpretability. The model demonstrates state-of-the-art performance over the QIN-BREAST dataset, achieving testing accuracies of 98.861% for ‘tumours’ and 99.447% for ‘non-tumours’, as well as over the Duke Breast Cancer Dataset, where it achieves 99.104% for ‘tumours’ and 99.753% for ‘non-tumours’, while offering enhanced interpretability through XAI methods.

Efficient train scheduling on single-track railways represents a significant challenge due to operational constraints. Multiple trains share the same track and an optimal schedule must ensure that trains traveling in opposite directions do not collide while minimising delays. This paper proposes a novel approach to address this problem by formulating train scheduling as a distributed constraint satisfaction problem and applying a multi-agent system to solve it. We propose a simple, yet efficient, system in which agents cooperate to schedule trains on a single-track railway. The result shows that our system is reliable and fast in comparison to other popular approaches.