Explainable Artificial Intelligence Research Articles

Autonomous intrusion detection for IoT: a decentralized and privacy preserving approach

The Internet of Things (IoT) has been increasingly adopted in domains such as smart infrastructure, healthcare, supply chain, transportation, and many others. However, the constrained computational resources of these devices make conventional security approaches against security threats not applicable. This limitation emphasizes the need to explore new approaches, specifically tailored to these kinds of devices. To increase protection against cyberattacks in IoT devices, Intrusion Detection Systems (IDSs) are considered an effective approach. Machine Learning (ML) techniques can be combined with Federated Learning to enhance the privacy and scalability of these systems. Many IDSs have been introduced, but there is a research gap concerning Host Intrusion Detection System (HIDS), which is the primary focus of our current work. Additionally, existing research predominantly focuses on the application of ML techniques and their evaluation, with limited attention to real-world implementation. We propose a lightweight HIDS that relies on the analysis of system call traces to detect malicious activities. The proposed HIDS achieved an accuracy rate of approximately 98%. Finally, by using eXplainable Artificial Intelligence methods, we sought to provide explanations for these these results.

Navigating ethical challenges of explainable ai in autonomous systems

The rapid integration of autonomous systems, such as vehicles, drones, and robots, into various sectors brings forth significant ethical challenges concerning their decision-making processes. This paper examines the role of Explainable AI (XAI) in addressing these challenges, particularly regarding accountability in the event of accidents and the necessity of human oversight in automated environments. We discuss the critical ethical implications of transparency, emphasizing how XAI can bridge the gap between complex algorithmic decision-making and public understanding, thereby fostering trust in these technologies. The paper also outlines the current regulatory frameworks for AI safety, analysing their effectiveness in promoting responsible innovation. Furthermore, we investigate the consequences of opaque algorithms, especially in life-critical applications where the stakes are exceptionally high. Through an analysis of case studies, we showcase how organizations have successfully implemented XAI to enhance safety measures and uphold ethical responsibility in their autonomous systems. Ultimately, this study advocates for the integration of XAI as a vital component in developing responsible autonomous technologies, ensuring accountability, and safeguarding public trust in an era increasingly defined by automation.

A SYSTEMATIC LITERATURE REVIEW OF PREDICTIVE MODELS AND ANALYTICS IN AI-DRIVEN CREDIT SCORING

This systematic review examines the transformative role of AI-driven models in credit scoring, highlighting their advances over traditional statistical methods in terms of predictive accuracy, adaptability, and inclusivity. By synthesizing findings from 70 studies, this review demonstrates that machine learning techniques, particularly ensemble models such as random forests and gradient boosting, effectively capture complex, non-linear relationships in borrower data, providing more accurate risk assessments across diverse demographics. Deep learning models, especially convolutional and recurrent neural networks, extend credit scoring capabilities to unstructured and alternative data sources, supporting financial inclusion by enabling assessments of individuals without traditional credit histories. Hybrid models that integrate logistic regression with neural networks offer an optimal balance between interpretability and predictive power, addressing regulatory demands for transparency while maintaining robust accuracy. Ensemble techniques like stacking and blending enhance model adaptability, allowing credit scoring systems to integrate multiple perspectives and improve prediction accuracy in varied borrower contexts. Despite these advancements, challenges remain in the form of ethical concerns and the need for model interpretability, particularly with complex deep-learning architectures. The review underscores the importance of developing fairness-aware and explainable AI frameworks to ensure that as AI-driven credit scoring evolves, it remains both transparent and equitable. These insights suggest that with careful attention to ethics and transparency, AI has the potential to create a more inclusive and resilient credit scoring landscape, accommodating the needs of an increasingly diverse global population.

XAI-Based Accurate Anomaly Detector That Is Robust Against Black-Box Evasion Attacks for the Smart Grid

In the realm of smart grids, machine learning (ML) detectors—both binary (or supervised) and anomaly (or unsupervised)—have proven effective in detecting electricity theft (ET). However, binary detectors are designed for specific attacks, making their performance unpredictable against new attacks. Anomaly detectors, conversely, are trained on benign data and identify deviations from benign patterns as anomalies, but their performance is highly sensitive to the selected threshold values. Additionally, ML detectors are vulnerable to evasion attacks, where attackers make minimal changes to malicious samples to evade detection. To address these limitations, we introduce a hybrid anomaly detector that combines a Deep Auto-Encoder (DAE) with a One-Class Support Vector Machine (OCSVM). This detector not only enhances classification performance but also mitigates the threshold sensitivity of the DAE. Furthermore, we evaluate the vulnerability of this detector to benchmark evasion attacks. Lastly, we propose an accurate and robust cluster-based DAE+OCSVM ET anomaly detector, trained using Explainable Artificial Intelligence (XAI) explanations generated by the Shapley Additive Explanations (SHAP) method on consumption readings. Our experimental results demonstrate that the proposed XAI-based detector achieves superior classification performance and exhibits enhanced robustness against various evasion attacks, including gradient-based and optimization-based methods, under a black-box threat model.

Guiding computationally intensive theory development with explainable artificial intelligence: The case of shapley additive explanations

This study advances the field of Computationally Intensive Theory Development (CTD) by examining the capabilities of Explainable Artificial Intelligence (XAI), in particular SHapley Additive exPlanations (SHAP), for theory development, while providing guidelines for this process. We evaluate SHAP’s methodological abilities and develop a structured approach for using SHAP to harness insights from black-box predictive models. For this purpose, we leverage a dual-methodological approach. First, to assess SHAP’s capabilities in uncovering patterns that shape a phenomenon, we conduct a Monte-Carlo simulation study. Second, to illustrate and guide the theory development process with SHAP for CTD, we apply SHAP in a use-case using real-world data. Based on these analyses, we propose a stepwise uniform and replicable approach giving guidance that can benefit rigorous theory development and increase the traceability of the theorizing process. With our structured approach, we contribute to the use of XAI approaches in research and, by uncovering patterns in black-box prediction models, add to the ongoing search for next-generation theorizing methods in the field of Information Systems (IS).

An ensemble machine learning framework with explainable artificial intelligence for predicting haemoglobin anaemia considering haematological markers

Anaemia is a disorder marked by low blood levels of haemoglobin (HGB), affecting people of all ages and ethnicities and is a major global public health concern. Anaemia must be diagnosed as soon as possible to enable prompt treatment and intervention, which can reduce complications and enhance patient outcomes. With the ability to improve diagnostic precision and expedite patient care procedures, machine learning (ML) has become a potent instrument in the healthcare industry. Hence, we examine the use of ML approaches to predict haemoglobin-like anaemia in this research article. Based on a heterogeneous dataset of blood markars, we investigate the performance of many machine learning techniques such as Logistic Regression, CatBoost, XgBoost Decision Trees, KNN and others. The algorithms are further ensembled using a customized stacking approach. The ML models' judgments are interpreted using explainable artificial intelligence (XAI) methods. The xgboost and the stacking classifier obtained an accuracy, precision and recall of 99% each. Our research shows how ML models can help with the early diagnosis and treatment of anaemia, which will ultimately lead to better patient outcomes and healthcare results. Overall, the research shows how ML emphasizes the value of interdisciplinary cooperation in solving challenging medical problems.

Revolutionizing Radiology With Artificial Intelligence.

Artificial intelligence (AI) is rapidly transforming the field of radiology, offering significant advancements in diagnostic accuracy, workflow efficiency, and patient care. This article explores AI's impact on various subfields of radiology, emphasizing its potential to improve clinical practices and enhance patient outcomes. AI-driven technologies such as machine learning, deep learning, and natural language processing (NLP) are playing a pivotal role in automating routine tasks, aiding in early disease detection, and supporting clinical decision-making, allowing radiologists to focus on more complex diagnostic challenges. Key applications of AI in radiology include improving image analysis through computer-aided diagnosis (CAD) systems, which enhance the detection of abnormalities in imaging, such as tumors. AI tools have demonstrated high accuracy in analyzing medical images, integrating data from multiple imaging modalities such as CT, MRI, and PET to provide comprehensive diagnostic insights. These advancements facilitate personalized treatment planning and complement radiologists' workflows. However, for AI to be fully integrated into radiology workflows, several challenges must be addressed, including ensuring transparency in how AI algorithms work, protecting patient data, and avoiding biases that could affect diverse populations. Developing explainable AI systems that can clearly show how decisions are made is crucial, as is ensuring AI tools can seamlessly fit into existing radiology systems. Collaboration between radiologists, AI developers, and policymakers, alongside strong ethical guidelines and regulatory oversight, will be key to ensuring AI is implemented safely and effectively in clinical practice. Overall, AI holds tremendous promise in revolutionizing radiology. Through its ability to automate complex tasks, enhance diagnostic capabilities, and streamline workflows, AI has the potential to significantly improve the quality and efficiency of radiology practices. Continued research, development, and collaboration will be crucial in unlocking AI's full potential and addressing the challenges that accompany its adoption.

Forecasting Patient Early Readmission from Irish Hospital Discharge Records Using Conventional Machine Learning Models.

Predicting patient readmission is an important task for healthcare risk management, as it can help prevent adverse events, reduce costs, and improve patient outcomes. In this paper, we compare various conventional machine learning models and deep learning models on a multimodal dataset of electronic discharge records from an Irish acute hospital. We evaluate the effectiveness of several widely used machine learning models that leverage patient demographics, historical hospitalization records, and clinical diagnosis codes to forecast future clinical risks. Our work focuses on addressing two key challenges in the medical fields, data imbalance and the variety of data types, in order to boost the performance of machine learning algorithms. Furthermore, we also employ SHapley Additive Explanations (SHAP) value visualization to interpret the model predictions and identify both the key data features and disease codes associated with readmission risks, identifying a specific set of diagnosis codes that are significant predictors of readmission within 30 days. Through extensive benchmarking and the application of a variety of feature engineering techniques, we successfully improved the area under the curve (AUROC) score from 0.628 to 0.7 across our models on the test dataset. We also revealed that specific diagnoses, including cancer, COPD, and certain social factors, are significant predictors of 30-day readmission risk. Conversely, bacterial carrier status appeared to have minimal impact due to lower case frequencies. Our study demonstrates how we effectively utilize routinely collected hospital data to forecast patient readmission through the use of conventional machine learning while applying explainable AI techniques to explore the correlation between data features and patient readmission rate.

Explainable AI for automated respiratory misalignment detection in PET/CT imaging

Purpose.Positron emission tomography (PET) image quality can be affected by artifacts emanating from PET, computed tomography (CT), or artifacts due to misalignment between PET and CT images. Automated detection of misalignment artifacts can be helpful both in data curation and in facilitating clinical workflow. This study aimed to develop an explainable machine learning approach to detect misalignment artifacts in PET/CT imaging.Approach.This study included 1216 PET/CT images. All images were visualized and images with respiratory misalignment artifact (RMA) detected. Using previously trained models, four organs including the lungs, liver, spleen, and heart were delineated on PET and CT images separately. Data were randomly split into cross-validation (80%) and test set (20%), then two segmentations performed on PET and CT images were compared and the comparison metrics used as predictors for a random forest framework in a 10-fold scheme on cross-validation data. The trained models were tested on 20% test set data. The model's performance was calculated in terms of specificity, sensitivity, F1-Score and area under the curve (AUC).Main results.Sensitivity, specificity, and AUC of 0.82, 0.85, and 0.91 were achieved in ten-fold data split. F1_score, sensitivity, specificity, and AUC of 84.5 vs 82.3, 83.9 vs 83.8, 87.7 vs 83.5, and 93.2 vs 90.1 were achieved for cross-validation vs test set, respectively. The liver and lung were the most important organs selected after feature selection.Significance.We developed an automated pipeline to segment four organs from PET and CT images separately and used the match between these segmentations to decide about the presence of misalignment artifact. This methodology may follow the same logic as a reader detecting misalignment through comparing the contours of organs on PET and CT images. The proposed method can be used to clean large datasets or integrated into a clinical scanner to indicate artifactual cases.

Low conduction velocity is a key factor for localisation of reentrant drivers for atrial fibrillation

Abstract Background Atrial Fibrillation (AF) is a prevalent cardiac arrhythmia globally, associated with heightened risks of stroke, dementia, and heart failure. Catheter ablation, the sole curative therapy for AF, yields suboptimal success rates for persistent AF cases. Understanding the mechanisms and dynamics of AF remains challenging, hindering treatment advancements. Novel ablation strategies aimed at targeting substrates harbouring AF reentrant drivers (RD), such as low-voltage ablation and rotor-based ablation, have yet to demonstrate superiority over the gold standard ablation strategy – pulmonary vein isolation. Purpose The purpose of the study is to integrate in-silico 3D left atrial (LA) simulations and explainable artificial intelligence (AI) to pinpoint key factors for RD localisation in persistent AF, aiming to improve the mechanistic understanding of AF and help develop a more effective ablation strategy. Methods 3D LA models were derived from magnetic resonance images of 41 patients and combined with atrial cell electrophysiology models. Persistent AF was initiated in each patient-specific LA model with fast pacing, and five functional features (action potential duration (APD), conduction velocity (CV), wavelength, APD spatial gradient, CV spatial gradient, and wavelength spatial gradient) and three structural features (fibrotic density, fibrotic entropy, and image intensity ratio) were evaluated. A random forest AI model was trained to classify phase singularities corresponding to the RDs (PS-RD) in five distinct tissue classes (total LA tissue, healthy tissue, fibrotic border zone, dense fibrotic tissue, and PS-RD tissue) based on the functional and structural features. Shapley additive explanations (SHAP) analysis was then used to find the most influential feature in the AI model’s decision process. Results Simulations of the patient-specific LA models revealed that RDs were typically localised in regions with the lowest atrial CV, which often coincided with patches of dense fibrotic tissue. In a five-fold cross-validation, the random forest AI model achieved high accuracy in classifying PS-RD regions: a mean ROC AUC of 0.99 ± 0.010, a mean F1 score of 0.86 ± 0.034, a mean recall of 0.95 ± 0.058, and a mean precision score of 0.79 ± 0.085. The further SHAP analysis revealed that low CV was the key determinant in classifying the PS-RD regions. Conclusion Our study combined AF computational modelling and explainable AI to identify the key factors in RD localisation. Both the computational and AI models concluded that low CV was the most important determinant in localising RDs. These findings provide mechanistic evidence for the efficacy of ablation strategies that target low CV atrial areas for persistent AF patients.

An explainable AI for trustworthy detection of atrial fibrillation on reduced lead ECGs in mobile applications

Abstract Introduction Artificial intelligence (AI), particularly deep learning (DL), has demonstrated high performance in various diagnostic problems, including the detection of paroxysmal atrial fibrillation (AF) from electrocardiograms (ECGs). In mobile applications, these approaches can help to detect AF at an early stage and thus prevent possible secondary diseases. However, due to their black-box nature, DL approaches lack explainability and interpretability. Furthermore, 12-lead ECGs are the clinical standard, but mobile devices usually only provide 1–3 leads, which are similar but not identical to Einthoven leads and differ in morphology. Purpose We apply a recently introduced explainable ECG analysis architecture (xECGArch) to single-lead ECGs to verify its performance in mobile applications. xECGArch can learn rhythmic and morphological characteristics using two parallel convolutional neural networks of different dimensions (Fig. 1). When combined with methods from explainable AI, the classification can be traced back to rhythmic and morphological characteristics. Methods We used a subset of 9 854 ECGs (n(AF) = n(non-AF) = 4 927) from 4 public databases (Chapman-Shaoxing, CPSC2018, Georgia, and PTB-XL). Noise was removed using a cascade of high-pass filtering at 0.3 Hz and discrete wavelet transform. For each lead, we trained xECGArch on single-lead ECGs of 10 s length to classify AF vs. non-AF using 90% of the ECGs with 5-fold cross-validation and tested it on the remaining randomly selected ECGs (n(AF) = 507, n(non-AF) = 479). Consistent data splitting was ensured for all leads. Model explanations were generated using deep Taylor decomposition. Results Both models achieved minimum 91.1% accuracy for each ECG lead (Fig. 1). On lead I, the morphology model outperformed the rhythm model with 92.8% vs. 91.5% accuracy, while the rhythm model slightly outperformed the morphology model on leads II and III with 94.2%–95.0% accuracy vs. 91.1%–94.0%. Both models achieved maximum accuracy of 95.0% (rhythm) and 94.0% (morphology) on lead II. By combining both models, xECGArch reached 93.7%–95.1% accuracy. The model explanations confirm that the rhythm model primarily considers QRS complexes as relevant, while the morphology model focuses on fibrillatory waves (Fig. 1). Conclusions Both rhythm and morphology models achieved reliable results across all leads, competing with the state-of-the-art for automatic AF detection from single-lead ECGs. The combination of both models in xECGArch further increased the accuracy. xECGArch is based on the medical reading of ECGs by considering rhythm and morphology characteristics and is therefore interpretable by design. The model explanations are consistent with the diagnostic criteria for AF (Fibrillatory waves and absolute arrhythmia) across leads. With high accuracy on different single-lead ECGs and improved interpretability, xECGArch provides a trustworthy method for ECG analysis in mobile applications.

Automated System for Credibility Diagnosis of Forged Facial Images Using Ensemble Deep Learning and Explainable AI

Automated System for Credibility Diagnosis of Forged Facial Images Using Ensemble Deep Learning and Explainable AI

Leveraging Explainable Artificial Intelligence (XAI) for Expert Interpretability in Predicting Rapid Kidney Enlargement Risks in Autosomal Dominant Polycystic Kidney Disease (ADPKD)

Autosomal dominant polycystic kidney disease (ADPKD) is the predominant hereditary factor leading to end-stage renal disease (ESRD) worldwide, affecting individuals across all races with a prevalence of 1 in 400 to 1 in 1000. The disease presents significant challenges in management, particularly with limited options for slowing cyst progression, as well as the use of tolvaptan being restricted to high-risk patients due to potential liver injury. However, determining high-risk status typically requires magnetic resonance imaging (MRI) to calculate total kidney volume (TKV), a time-consuming process demanding specialized expertise. Motivated by these challenges, this study proposes alternative methods for high-risk categorization that do not rely on TKV data. Utilizing historical patient data, we aim to predict rapid kidney enlargement in ADPKD patients to support clinical decision-making. We applied seven machine learning algorithms—Random Forest, Logistic Regression, Support Vector Machine (SVM), Light Gradient Boosting Machine (LightGBM), Gradient Boosting Tree, XGBoost, and Deep Neural Network (DNN)—to data from the Polycystic Kidney Disease Outcomes Consortium (PKDOC) database. The XGBoost model, combined with the Synthetic Minority Oversampling Technique (SMOTE), yielded the best performance. We also leveraged explainable artificial intelligence (XAI) techniques, specifically Local Interpretable Model-Agnostic Explanations (LIME) and Shapley Additive Explanations (SHAP), to visualize and clarify the model’s predictions. Furthermore, we generated text summaries to enhance interpretability. To evaluate the effectiveness of our approach, we proposed new metrics to assess explainability and conducted a survey with 27 doctors to compare models with and without XAI techniques. The results indicated that incorporating XAI and textual summaries significantly improved expert explainability and increased confidence in the model’s ability to support treatment decisions for ADPKD patients.

A novel risk-stratification model that predicts long-term clinical outcomes in patients with recurrent pericarditis

Abstract Background Recurrent Pericarditis (RP) is condition associated with significant debility and morbidity. Previous studies have identified variables such as elevated C-Reactive protein, corticosteroid usage, and presence of delayed hyper-enhancement (DHE) on cardiac magnetic resonance imaging are associated with recurrence. To date however, there is no established model for risk stratifying outcomes in these patients, therefore we attempted to develop a model that could predict long-term outcomes in patients with RP and stratify them into groups based on risk. Methods A total of 365 consecutive patients with RP from 2012-2019 were retrospectively examined. Our primary outcome was clinical remission (CR) defined as symptom resolution in addition to complete cessation of all anti-inflammatory therapy. We employed 5 survival models: CoxPH, RSF, SSVM, GBSA and XGBoost with Cox loss function to determine the likelihood of achieving CR within 5 years and to stratify patients into specific risk groups. We divided the dataset into training (70%) and test (30%) sets. Model training and optimization of the training set occurred via 5-fold cross validation. Model evaluation and selection was based on C-index. Finally, we employed an explainable artificial intelligence approach, SHapley Additive exPlanations to generate both individual and global explanation of the models decision. Results Amongst the cohort, the mean age was 46 ± 15 years, 205 (56%) were female, and the main etiology was idiopathic in 223 (61%). CR was achieved in 118 (32%) patients. The final model included steroid dependency, total number of recurrences, pericardial DHE, age, etiology, gender, ejection fraction and heart rate as its most important parameters. We observed similar CR rates across three risk score ranges and accordingly stratified patients into low (n=62), intermediate (n=154), and high-risk (n=149) groups. Scores of 3 or lower were categorized as low risk, with CR rates exceeding 90%. Scores of 4 to 7 were associated with CR rates of 50%. Scores of 8 or higher were deemed high-risk, with CR rates notably below 10%. Overall, the model predicted the outcome with C-index of 0.800 on the test set and exhibited a significant ability in stratification of patients into low-, intermediate-, and high-risk groups (log-rank test; P <0.0001). Conclusion We created a novel risk prediction model that can aid in predicting rates clinical remission in patients with recurrent pericarditis as well as aid in stratification into low, medium, and high risk groups. This risk prediction model may be helpful in diagnosis and management of these complicated patients.Figure 1

Research on Explainability of Deep Neural Networks and Its Applications

In recent years, Artificial Intelligence (AI) models have surged in popularity and success, garnering much attention across many disciplines and applications. In particular, recent developments in Deep Neural Networks (DNN) have allowed them to match and even surpass human capabilities in many tasks. However, the complexity of deep neural networks causes them function like black boxes that are very difficult for humans to understand, thus making them untrustworthy and limiting their usefulness in high-risk applications. Explainable Artificial Intelligence aims to alleviate this issue by generating explanations for DNNs to allow humans to understand their reasoning and increase their transparency. This paper reviews research on neural network explainability, detailing its definition, necessity, classification, and evaluation. It categorizes current explanation methods for deep neural networks according to their number of inputs. Additionally, it introduces the principles and evaluation methods for common interpretability algorithms. Future research directions and applications of interpretable neural networks are summarized. The challenges facing interpretable neural networks are outlined, along with potential solutions to these challenges.

Part-Prototype Models in Medical Imaging: Applications and Current Challenges

Recent developments in Artificial Intelligence have increasingly focused on explainability research. The potential of Explainable Artificial Intelligence (XAI) in producing trustworthy computer-aided diagnosis systems and its usage for knowledge discovery are gaining interest in the medical imaging (MI) community to support the diagnostic process and the discovery of image biomarkers. Most of the existing XAI applications in MI are focused on interpreting the predictions made using deep neural networks, typically including attribution techniques with saliency map approaches and other feature visualization methods. However, these are often criticized for providing incorrect and incomplete representations of the black-box models’ behaviour. This highlights the importance of proposing models intentionally designed to be self-explanatory. In particular, part-prototype (PP) models are interpretable-by-design computer vision (CV) models that base their decision process on learning and identifying representative prototypical parts from input images, and they are gaining increasing interest and results in MI applications. However, the medical field has unique characteristics that could benefit from more advanced implementations of these types of architectures. This narrative review summarizes existing PP networks, their application in MI analysis, and current challenges.

Accurate detection of dilated cardiomyopathy onset through machine learning predictions from ECG data

Abstract Background Dilated cardiomyopathy (DCM) is a primary heart muscle disease characterised by left-ventricular dilatation and reduced contractility, often associated with arrhythmia and conduction disease. DCM often has a genetic component, putting family members of diagnosed patients at risk for the disease. Stratification tools to provide personalized screening advice are currently lacking. Artificial intelligence (AI) based analysis of the electrocardiogram (ECG) offers a potential solution to this challenge. Purpose Our study aims to develop and validate a novel AI-ECG model capable of prediction and early detection of DCM. Methods An AI-ECG model was developed on a United States (US) based secondary care centre dataset including 337,463 ECGs from 50,932 patients within 4 months of diagnostic echocardiography and 265,576 ECGs from 51,086 patients prior to the first diagnostic echo. The data was split 50/10/40% for training, validation, and testing, with each subset containing 9-12% DCM positive labels. The model architecture was based on a convolutional neural network with residual blocks with a modified final layer for a discrete-time survival model. The AI-ECG score was linked to changes in the median ECG within the test set to provide explainability. External validation was performed on 68,885 diagnostic ECGs (0.08% DCM positive) from the UK biobank (UKB) and a Chinese general hospital dataset with 317,854 diagnostic ECGs (2.2% DCM positive) and 76,936 predictive ECGs (1.7% DCM positive). We performed common and rare variant association studies for the AI ECG scores on quality-controlled UKB data. Results The AI-ECG score for predicted future DCM was associated with diverse ECG features (Fig. 1). In the US cohort test set, the model achieved high diagnostic accuracy with an AUC of 0.9 (95% CI 0.89-0.90) overall, and 0.89 (0.85-0.92) in a subset excluding ischemic or valvular heart disease. External datasets showed AUCs of 0.86 (0.81-0.91) in the UKB and 0.97 (96-0.97) in the Chinese dataset. The predictive performance of the AI-ECG score, adjusted for age and sex, closely approximated echocardiographic traits (LVEF and LVEDD, Fig. 2) with a concordance index (CI) of 0.79 (0.76-0.83) and 0.81 (0.77-0.85), respectively, at 10 years of follow up (FU). Combining AI-ECG scores with echocardiographic traits enhanced the CI to 0.84 (0.8-0.87). Similar findings were observed in the Chinese dataset, with a CI of 0.87 (0.85-0.89) for the AI-ECG score and 0.9 (0.88-0.92) for echocardiographic traits. Genetic analyses in UKB identified associations between the AI-ECG score and rare TTN variants as well as 25 common variant loci. Conclusion Our explainable AI ECG model for future DCM prediction demonstrates robust performance across datasets, providing comparable predictive accuracy to echocardiographic traits up to 10 years pre-diagnosis. Genetic associations support the model's validity and provide new tools for DCM genetic discovery.

Intelligent Manufacturing in Wine Barrel Production: Deep Learning-Based Wood Stave Classification

Innovative wood inspection technology is crucial in various industries, especially for determining wood quality by counting rings in each stave, a key factor in wine barrel production. (1) Background: Traditionally, human inspectors visually evaluate staves, compensating for natural variations and characteristics like dirt and saw-induced aberrations. These variations pose significant challenges for automatic inspection systems. Several techniques using classical image processing and deep learning have been developed to detect tree-ring boundaries, but they often struggle with woods exhibiting heterogeneity and texture irregularities. (2) Methods: This study proposes a hybrid approach combining classical computer vision techniques for preprocessing with deep learning algorithms for classification, designed for continuous automated processing. To enhance performance and accuracy, we employ a data augmentation strategy using cropping techniques to address intra-class variability in individual staves. (3) Results: Our approach significantly improves accuracy and reliability in classifying wood with irregular textures and heterogeneity. The use of explainable AI and model calibration offers a deeper understanding of the model’s decision-making process, ensuring robustness and transparency, and setting confidence thresholds for outputs. (4) Conclusions: The proposed system enhances the performance of automatic wood inspection technologies, providing a robust solution for industries requiring precise wood quality assessment, particularly in wine barrel production.

Predictive and Explainable Artificial Intelligence for Neuroimaging Applications.

The aim of this review is to highlight the new advance of predictive and explainable artificial intelligence for neuroimaging applications. Data came from 30 original studies in PubMed with the following search terms: "neuroimaging" (title) together with "machine learning" (title) or "deep learning" (title). The 30 original studies were eligible according to the following criteria: the participants with the dependent variable of brain image or associated disease; the interventions/comparisons of artificial intelligence; the outcomes of accuracy, the area under the curve (AUC), and/or variable importance; the publication year of 2019 or later; and the publication language of English. The performance outcomes reported were within 58-96 for accuracy (%), 66-97 for sensitivity (%), 76-98 for specificity (%), and 70-98 for the AUC (%). The support vector machine and the convolutional neural network registered the best performance (AUC 98%) for the classifications of low- vs. high-grade glioma and brain conditions, respectively. Likewise, the random forest delivered the best performance (root mean square error 1) for the regression of brain conditions. The following factors were discovered to be major predictors of brain image or associated disease: (demographic) age, education, sex; (health-related) alpha desynchronization, Alzheimer's disease stage, CD4, depression, distress, mild behavioral impairment, RNA sequencing; (neuroimaging) abnormal amyloid-β, amplitude of low-frequency fluctuation, cortical thickness, functional connectivity, fractal dimension measure, gray matter volume, left amygdala activity, left hippocampal volume, plasma neurofilament light, right cerebellum, regional homogeneity, right middle occipital gyrus, surface area, sub-cortical volume. Predictive and explainable artificial intelligence provide an effective, non-invasive decision support system for neuroimaging applications.

Identifying and Managing Risks in High-Tech Supply Networks using Explainable Artificial Intelligence

Supply chains have become vital in enabling the smooth movement of material and information in the swiftly changing high-tech industries. By utilising technological innovations like automation and AI, supply chains may improve both safety and effectiveness at every level of the process, from manufacturing to sustainability. The necessity to adjust operations, give the adoption of technology top priority for resilience, and improve supply chain management efficiency. Transformation of the Supply Chain is becoming a choice of subject, when it comes to maintain the ideal inventory levels and Just in Time service deliveries. Both AI and ML are the utmost importance for the maintenance of the inventory or service through the technology. It not only helps in reducing the loss of opportunity costs, but also it resolves the issue of overstocks of any inventory and the enhanced level of service delivery. This ultimately results in the optimum utilization of the resources through the help of technology. Moreover, it is the win-win situation for the customer and provider. Even while technology has only recently begun to be incorporated into supply chains, the manufacturing and distribution sectors have adopted it more quickly, emphasising the need for effective risk identification and mitigation techniques. Notwithstanding the possible advantages of transforming the supply chain, there are obstacles and limitations that make an entire technological shift complex. The chapter also explores the concept explainable AI which can be understood as the set of techniques that makes the decision-making processes of AI systems comprehensible to humans. Acknowledging the inherent complexities and relationships within these chains, the goal of this chapter is to explore risk identification and management in high-tech supply networks. A competent and nimble stakeholder base can manage the complex landscape of the high-tech world through the incorporation of innovative technologies, proactive actions, and constructive alliances.