AI Methods Research Articles

Unveiling the human-like similarities of automatic facial expression recognition: An empirical exploration through explainable ai

AbstractFacial expression recognition is vital for human behavior analysis, and deep learning has enabled models that can outperform humans. However, it is unclear how closely they mimic human processing. This study aims to explore the similarity between deep neural networks and human perception by comparing twelve different networks, including both general object classifiers and FER-specific models. We employ an innovative global explainable AI method to generate heatmaps, revealing crucial facial regions for the twelve networks trained on six facial expressions. We assess these results both quantitatively and qualitatively, comparing them to ground truth masks based on Friesen and Ekman’s description and among them. We use Intersection over Union (IoU) and normalized correlation coefficients for comparisons. We generate 72 heatmaps to highlight critical regions for each expression and architecture. Qualitatively, models with pre-trained weights show more similarity in heatmaps compared to those without pre-training. Specifically, eye and nose areas influence certain facial expressions, while the mouth is consistently important across all models and expressions. Quantitatively, we find low average IoU values (avg. 0.2702) across all expressions and architectures. The best-performing architecture averages 0.3269, while the worst-performing one averages 0.2066. Dendrograms, built with the normalized correlation coefficient, reveal two main clusters for most expressions: models with pre-training and models without pre-training. Findings suggest limited alignment between human and AI facial expression recognition, with network architectures influencing the similarity, as similar architectures prioritize similar facial regions.

Scenario Identification and Classification to Support the Assessment of Advanced Driver Assistance Systems

In recent years, driver assistance systems in cars, buses, and trucks have become more common and powerful. In particular, the introduction of AI methods to sensors, signal fusion, and traffic recognition allows us to step forward from actual level-2 assistance to level-3 Advanced Driver Assistance Systems (ADAS), where driving becomes autonomous and responsibility shifts from the driver to the automobile manufacturers. This, however, requires a high-precision risk assessment of failure, which can only be achieved by extensive data acquisition and statistical analysis of real traffic scenarios (which is impossible to perform by humans). Therefore, critical driving situations have to be identified and classified automatically. This paper develops and compares two different strategies—a traditional rule-based approach derived from deterministic causal considerations, and an AI-based approach trained with idealized cut-in, cut-out, and cut-through maneuvers. Application to a 10-h measurement sequence on a German highway demonstrates that the latter has the higher performance, whereas the former misses some of the safety-relevant events to be identified.

Quantum-Fuzzy Expert Timeframe Predictor for Post-TAVR Monitoring

This paper presents a novel approach to predicting specific monitoring timeframes for Permanent Pacemaker Implantation (PPMI) requirements following a Transcatheter Aortic Valve Replacement (TAVR). The method combines Quantum Ant Colony Optimization (QACO) with the Adaptive Neuro-Fuzzy Inference System (ANFIS) and incorporates expert knowledge. Although this forecast is more precise, it requires a larger number of predictors to achieve this level of accuracy. Our model deploys expert-derived insights to guarantee the clinical relevance and interpretability of the predicted outcomes. Additionally, we employ quantum computing techniques to address this complex and high-dimensional problem. Through extensive assessments, we show that our quantum-enhanced model outperforms traditional methods with notable improvement in evaluation metrics, such as accuracy, precision, recall, and F1 score. Furthermore, with the integration of eXplainable AI (XAI) methods, our solution enhances the transparency and reliability of medical predictive models, hence promoting improved clinical practice decision-making. The findings highlight how quantum computing has the potential to completely transform predictive analytics in the medical field, especially when it comes to improving patient care after TAVR.

On the Application of Artificial Intelligence/Machine Learning (AI/ML) in Late-Stage Clinical Development.

Whereas AI/ML methods were considered experimental tools in clinical development for some time, nowadays they are widely available. However, stakeholders in the health care industry still need to answer the question which role these methods can realistically play and what standards should be adhered to. Clinical research in late-stage clinical development has particular requirements in terms of robustness, transparency and traceability. These standards should also be adhered to when applying AI/ML methods. Currently there is some formal regulatory guidance available, but this is more directed at settings where a device or medical software is investigated. Here we focus on the application of AI/ML methods in late-stage clinical drug development, i.e. in a setting where currently less guidance is available. This is done via first summarizing available regulatory guidance and work done by regulatory statisticians followed by the presentation of an industry application where the influence of extensive sets of baseline characteristics on the treatment effect can be investigated by applying ML-methods in a standardized manner with intuitive graphical displays leveraging explainable AI methods. The paper aims at stimulating discussions on the role such analyses can play in general rather than advocating for a particular AI/ML-method or indication where such methods could be meaningful.

Machine Learning Prediction of Autism Spectrum Disorder From a Minimal Set of Medical and Background Information

Early identification of the likelihood of autism spectrum disorder (ASD) using minimal information is crucial for early diagnosis and intervention, which can affect developmental outcomes. To develop and validate a machine learning (ML) model for predicting ASD using a minimal set of features from background and medical information and to evaluate the predictors and the utility of the ML model. For this diagnostic study, a retrospective analysis of the Simons Foundation Powering Autism Research for Knowledge (SPARK) database, version 8 (released June 6, 2022), was conducted, including data from 30 660 participants after adjustments for missing values and class imbalances (15 330 with ASD and 15 330 without ASD). The SPARK database contains participants recruited from 31 university-affiliated research clinicals and online in 26 states in the US. All individuals with a professional ASD diagnosis and their families were eligible to participate. The model performance was validated on independent datasets from SPARK, version 10 (released July 21, 2023), and the Simons Simplex Collection (SSC), consisting of 14 790 participants, followed by phenotypic associations. Twenty-eight basic medical screening and background history items present before 24 months of age. Generalizable ML prediction models were developed for detecting ASD using 4 algorithms (logistic regression, decision tree, random forest, and eXtreme Gradient Boosting [XGBoost]). Performance metrics included accuracy, area under the receiver operating characteristics curve (AUROC), sensitivity, specificity, positive predictive value (PPV), and F1 score, offering a comprehensive assessment of the predictive accuracy of the model. Explainable AI methods were applied to determine the effect of individual features in predicting ASD as secondary outcomes, enhancing the interpretability of the best-performing model. The secondary outcome analyses were further complemented by examining differences in various phenotypic measures using nonparametric statistical methods, providing insights into the ability of the model to differentiate between different presentations of ASD. The study included 19 477 (63.5%) male and 11 183 (36.5%) female participants (mean [SD] age, 106 [62] months). The mean (SD) age was 113 (68) months for the ASD group and 100 (55) months for the non-ASD group. The XGBoost (termed AutMedAI) model demonstrated strong performance with an AUROC score of 0.895, sensitivity of 0.805, specificity of 0.829, and PPV of 0.897. Developmental milestones and eating behavior were the most important predictors. Validation on independent cohorts showed an AUROC of 0.790, indicating good generalizability. In this diagnostic study of ML prediction of ASD, robust model performance was observed to identify autistic individuals with more symptoms and lower cognitive levels. The robustness and ML model generalizability results are promising for further validation and use in clinical and population settings.

Advanced AI and renewable energy sources for unified rotor angle stability control

Maintaining a reliable electricity supply amidst the integration of diverse energy sources necessitates optimizing the stability of power systems. This paper introduces a groundbreaking method to enhance the efficiency and resilience of power grids. The increasing dependence on renewable energy sources poses significant challenges to traditional power networks, thereby demanding innovative solutions to uphold their stability and security. To address these challenges, we propose an architecture that seamlessly unifies dynamic, transient, and static rotor angle stability (RAS) controls into a single, streamlined system. Utilizing reinforcement learning and real-time decision-making, we present Lazy Deep Q Networks (LDQNs) as a novel approach to RAS control. LDQNs provide real-time rotor angle instructions to RAS devices, enabling precise and efficient stability management. The incorporation of mass-distributed energy storage further augments the system's responsiveness and flexibility, mitigating fluctuations and promoting overall stability. This study advances the application of AI methods to RAS control, building on prior research in frequency and voltage stability frameworks. The proposed system outperforms conventional RAS control methods by integrating LDQNs with mass-distributed energy storage, offering superior performance and adaptability. Case studies validate the effectiveness of the unified RAS framework, demonstrating its advantages over traditional approaches across various power system configurations.

Towards Personalized AI-Based Diabetes Therapy: A Review.

Insulin pumps and other smart devices have recently made significant advancements in the treatment of diabetes, a disorder that affects people all over the world. The development of medical AI has been influenced by AI methods designed to help physicians make diagnoses, choose a course of therapy, and predict outcomes. In this article, we thoroughly analyse how AI is being used to enhance and personalize diabetes treatment. The search turned up 77 original research papers, from which we've selected the most crucial information regarding the learning models employed, the data typology, the deployment stage, and the application domains. We identified two key trends, enabled mostly by AI: patient-based therapy personalization and therapeutic algorithm optimization. In the meanwhile, we point out various shortcomings in the existing literature, like a lack of multimodal database analysis or a lack of interpretability. The rapid improvements in AI and the expansion of the amount of data already available offer the possibility to overcome these difficulties shortly and enable a wider deployment of this technology in clinical settings.

Comparing AI methods for forecasting polyester fabric tensile property

Comparing AI methods for forecasting polyester fabric tensile property

Application of Artificial Intelligence in Glacier Studies: A State-of-the-Art Review

Assessing glaciers using recent and historical data and predicting the future impacts on them due to climate change are crucial for understanding global glacier mass balance, regional water resources, and downstream hydrology. Computational methods are crucial for analyzing current conditions and forecasting glacier changes using remote sensing and other data sources. Due to the complexity and large data volumes, there is a strong demand for accelerated computing. AI-based approaches are increasingly being adopted for their efficiency and accuracy in these tasks. Thus, in the current state-of-the-art review work, available research results on the application of AI methods for glacier studies are addressed. Using selected search terms, AI-based publications are collected from research databases. They are further classified in terms of their geographical locations and glacier-related research purposes. It was found that the majority of AI-based glacier studies focused on inventorying and mapping glaciers worldwide. AI techniques like U-Net, Random forest, CNN, and DeepLab are mostly utilized in glacier mapping, demonstrating their adaptability and scalability. Other AI-based glacier studies such as glacier evolution, snow/ice differentiation, and ice dynamic modeling are reviewed and classified, Overall, AI methods are predominantly based on supervised learning and deep learning approaches, and these methods have been used almost evenly in glacier publications over the years since the beginning of this research area. Thus, the integration of AI in glacier research is advancing, promising to enhance our comprehension of glaciers amid climate change and aiding environmental conservation and resource management.

IoT-based AI Methods for Indoor Air Quality Monitoring Systems: A Systematic Review

IoT-based AI Methods for Indoor Air Quality Monitoring Systems: A Systematic Review

Integrating AI methods for comprehensive climate change biology studies

Integrating AI methods for comprehensive climate change biology studies

Revisiting Mobile Payment Risk-Reduction Strategies: A Cross-Country Analysis

ABSTRACT Mobile payment risk has become a critical cybersecurity factor in the cashless society. The outbreak of COVID-19 helped proliferate mobile payments that also bring significant risks to users. Using AI methods, this study analyzed six dimensions of mobile payment risks (financial, privacy, performance, psychology, time, and security) in a survey of 748 respondents from three countries (UK, Taiwan, and Mozambique). The decision tree method was employed to identify and analyze critical perceived risks. The ANOVA test provided insights on the perceived risks between countries. The ANOVA test showed that UK users were concerned about financial and time risks; those in Mozambique were concerned about performance, psychological, and security risks; and those in Taiwan were concerned about privacy risks. The results revealed that decision trees outperformed other methods (such as neural networks, logistic regression, support vector machine (SVM), random forest, and Naïve Bayes models). Performance risk (Taiwan and Mozambique) and security risk (UK) are the most significant factors. Cultural differences influence mobile payment risk perception in different countries. The risk-reduction strategies were also matched to the critical factors by the decision tree. This showed that simplification and risk-sharing strategies were the major tactics in all three countries. The clarification strategy works for Taiwan and Mozambique, which focuses on the benefits of using mobile payments. The results suggest that enterprises should improve and simplify the mobile payment process and collaborate with the third parties to reduce and share cybersecurity risk.

Reassuring, Misleading, Debunking: Comparing Effects of XAI Methods on Human Decisions

Trust calibration is essential in AI-assisted decision-making. If human users understand the rationale on which an AI model has made a prediction, they can decide whether they consider this prediction reasonable. Especially in high-risk tasks such as mushroom hunting (where a wrong decision may be fatal), it is important that users make correct choices to trust or overrule the AI. Various explainable AI (XAI) methods are currently being discussed as potentially useful for facilitating understanding and subsequently calibrating user trust. So far, however, it remains unclear which approaches are most effective. In this article, the effects of XAI methods on human AI-assisted decision-making in the high-risk task of mushroom picking were tested. For that endeavor, the effects of (i) Grad-CAM attributions, (ii) nearest-neighbor examples, and (iii) network-dissection concepts were compared in a between-subjects experiment with \(N=501\) participants representing end-users of the system. In general, nearest-neighbor examples improved decision correctness the most. However, varying effects for different task items became apparent. All explanations seemed to be particularly effective when they revealed reasons to (i) doubt a specific AI classification when the AI was wrong and (ii) trust a specific AI classification when the AI was correct. Our results suggest that well-established methods, such as Grad-CAM attribution maps, might not be as beneficial to end users as expected and that XAI techniques for use in real-world scenarios must be chosen carefully.

Self-learning AI method for anomaly detection in ambient assisted living environments

Self-learning AI method for anomaly detection in ambient assisted living environments

Artificial intelligence for optimum tissue excision with indocyanine green fluorescence angiography for flap reconstructions: Proof of concept

BackgroundIndocyanine green fluorescence angiography (ICGFA) is gaining popularity as an intraoperative tool to assess flap perfusion. However, it needs interpretation and there is concern regarding a potential for over-debridement with its use. Here we describe an artificial intelligence (AI) method that indicates the extent of flap trimming required. MethodsOperative ICGFA recordings from ten consenting patients undergoing flap reconstruction without subsequent partial/total necrosis as part of an approved prospective study (NCT 04220242, Institutional Review Board Ref:1/378/2092), provided the training-testing datasets. Drawing from prior similar experience with ICGFA intestinal perfusion signal analysis, five fluorescence intensity and time-related features were analysed (MATLAB R2024a) from stabilised ICGFA imagery. Machine learning model training (with ten-fold cross-validation application) was grounded on the actual trimming by a consultant plastic surgeon (S.P.) experienced in ICGFA. MATLAB classification learner app was used to identify the most important feature and generate partial dependence plots for interpretability during training. Testing involved post-hoc application to unseen videos blinded to surgeon ICGFA interpretation. ResultsTraining:testing datasets comprised 7:3 ICGFA videos with 28 and 3 sampled lines respectively. Validation and testing accuracy were 99.9 % and 99.3 % respectively. Maximum fluorescence intensity identified as the most important predictive curve feature. Partial dependence plotting revealed a threshold of 22.1 grayscale units and regions with maximum intensity less then threshold being more likely to be predicted as “excise”. ConclusionThe AI method proved discriminative regarding indicating whether to retain or excise peripheral flap portions. Additional prospective patients and expert references are needed to validate generalisability.

Transparent Insights into Alzheimer’s Progression: A Time-Aware Approach with Explainable AI

Alzheimer’s disease, acknowledged for its intricate and degenerative characteristics, presents considerable challenges, particularly among the elderly population. The relentless nature of Alzheimer’s, marked by the gradual deterioration of cognitive function, underscores the urgency to develop effective strategies for early diagnosis and intervention. Artificial intelligence has played a significant role in terms of disease diagnosis and treatments. However, it has got limited acceptance in the medical community due to its lack of transparency. This study aims to advance the understanding and prediction of Alzheimer’s disease progression through the integration of time-aware modeling and Explainable AI techniques. It makes significant contributions by addressing two key objectives in the context of Alzheimer’s disease. First, by including temporal aspects, it accurately depicts the pace at which relevant predictors change over time, thereby capturing the dynamic nature of Alzheimer’s disease. Second, by giving interpretable insights into the algorithm’s decision-making process, the study hopes to empower researchers and physicians. This approach not only enhances transparency but also builds trust in the model’s outcomes. The ADNI dataset, comprising 2980 observations, was employed for developing a prediction model using various machine learning classifiers. Among these classifiers, the Random Forest model emerged as the top performer, exhibiting superior accuracy, a high Coefficient of Determination (R2), and an impressive F1 score. To enhance interpretability, subsequent analyses utilized LIME and SHAP techniques. By combining time-aware modeling with Explainable AI methods, we seek to unravel the dynamic relationships within the dataset, providing transparent insights into the temporal evolution of Alzheimer’s disease. Thus, this paper contributes to the creation of a clinically relevant and practical model for monitoring Alzheimer’s disease progression that holds the potential for a deeper understanding of the evolving nature of the disease and paving the way for personalized and timely interventions.

MixImages: An Urban Perception AI Method Based on Polarization Multimodalities.

Intelligent urban perception is one of the hot topics. Most previous urban perception models based on semantic segmentation mainly used RGB images as unimodal inputs. However, in natural urban scenes, the interplay of light and shadow often leads to confused RGB features, which diminish the model's perception ability. Multimodal polarization data encompass information dimensions beyond RGB, which can enhance the representation of shadow regions, serving as additional data for assistance. Additionally, in recent years, transformers have achieved outstanding performance in visual tasks, and their large, effective receptive field can provide more discriminative cues for shadow regions. For these reasons, this study proposes a novel semantic segmentation model called MixImages, which can combine polarization data for pixel-level perception. We conducted comprehensive experiments on a polarization dataset of urban scenes. The results showed that the proposed MixImages can achieve an accuracy advantage of 3.43% over the control group model using only RGB images in the unimodal benchmark while gaining a performance improvement of 4.29% in the multimodal benchmark. Additionally, to provide a reference for specific downstream tasks, we also tested the impact of different combinations of polarization types on the overall segmentation accuracy. The proposed MixImages can be a new option for conducting urban scene perception tasks.

Artificial Intelligence Models for the Detection of Microsatellite Instability from Whole-Slide Imaging of Colorectal Cancer.

With the advent of whole-slide imaging (WSI), a technology that can digitally scan whole slides in high resolution, pathology is undergoing a digital revolution. Detecting microsatellite instability (MSI) in colorectal cancer is crucial for proper treatment, as it identifies patients responsible for immunotherapy. Even though universal testing for MSI is recommended, particularly in patients affected by colorectal cancer (CRC), many patients remain untested, and they reside mainly in low-income countries. A critical need exists for accessible, low-cost tools to perform MSI pre-screening. Here, the potential predictive role of the most relevant artificial intelligence-driven models in predicting microsatellite instability directly from histology alone is discussed, focusing on CRC. The role of deep learning (DL) models in identifying the MSI status is here analyzed in the most relevant studies reporting the development of algorithms trained to this end. The most important performance and the most relevant deficiencies are discussed for every AI method. The models proposed for algorithm sharing among multiple research and clinical centers, including federal learning (FL) and swarm learning (SL), are reported. According to all the studies reported here, AI models are valuable tools for predicting MSI status on WSI alone in CRC. The use of digitized H&E-stained sections and a trained algorithm allow the extraction of relevant molecular information, such as MSI status, in a short time and at a low cost. The possible advantages related to introducing DL methods in routine surgical pathology are underlined here, and the acceleration of the digital transformation of pathology departments and services is recommended.

Application of Generative Artificial Intelligence in Predicting Membrane Partitioning of Drugs: Combining Denoising Diffusion Probabilistic Models and MD Simulations Reduces the Computational Cost to One-Third.

The optimal interaction of drugs with plasma membranes and membranes of subcellular organelles is a prerequisite for desirable pharmacology. Importantly, for drugs targeting the transmembrane lipid-facing sites of integral membrane proteins, the relative affinity of a drug to the bilayer lipids compared to the surrounding aqueous phase affects the partitioning, access, and binding of the drug to the target site. Molecular dynamics (MD) simulations, including enhanced sampling techniques such as steered MD, umbrella sampling (US), and metadynamics, offer valuable insights into the interactions of drugs with the membrane lipids and water in atomistic detail. However, these methods are computationally prohibitive for the high-throughput screening of drug candidates. This study shows that applying denoising diffusion probabilistic models (DDPMs), a generative AI method, to US simulation data reduces the computational cost significantly. Specifically, the models used only partial (one-third) data from the US simulations and reproduced the complete potential of mean force (PMF) profiles for three FDA-approved drugs (β2-adrenergic agonists) and ∼20 biologically relevant chemicals with known experimentally characterized bilayer locations. Intriguingly, the model can predict the solvation-free energies for partitioning and crossing the bilayer, preferred bilayer locations (low-energy well), and orientations of the ligands with high accuracy. The results indicate that DDPMs can be used to characterize the complete membrane partitioning profile of drug molecules using fewer umbrella sampling simulations at select positions along the bilayer normal (z-axis), irrespective of their amphiphilic-lipophilic-cephalophilic characteristics.

Explainable AI for Interpretation of Ovarian Tumor Classification Using Enhanced ResNet50.

Deep learning architectures like ResNet and Inception have produced accurate predictions for classifying benign and malignant tumors in the healthcare domain. This enables healthcare institutions to make data-driven decisions and potentially enable early detection of malignancy by employing computer-vision-based deep learning algorithms. These CNN algorithms, in addition to requiring huge amounts of data, can identify higher- and lower-level features that are significant while classifying tumors into benign or malignant. However, the existing literature is limited in terms of the explainability of the resultant classification, and identifying the exact features that are of importance, which is essential in the decision-making process for healthcare practitioners. Thus, the motivation of this work is to implement a custom classifier on the ovarian tumor dataset, which exhibits high classification performance and subsequently interpret the classification results qualitatively, using various Explainable AI methods, to identify which pixels or regions of interest are given highest importance by the model for classification. The dataset comprises CT scanned images of ovarian tumors taken from to the axial, saggital and coronal planes. State-of-the-art architectures, including a modified ResNet50 derived from the standard pre-trained ResNet50, are implemented in the paper. When compared to the existing state-of-the-art techniques, the proposed modified ResNet50 exhibited a classification accuracy of 97.5 % on the test dataset without increasing the the complexity of the architecture. The results then were carried for interpretation using several explainable AI techniques. The results show that the shape and localized nature of the tumors play important roles for qualitatively determining the ability of the tumor to metastasize and thereafter to be classified as benign or malignant.