Segui questo link per vedere altri tipi di pubblicazioni sul tema: Interpretability of AI Models for Parkinson's Disease Detection.
Articoli di riviste sul tema "Interpretability of AI Models for Parkinson's Disease Detection"
Cita una fonte nei formati APA, MLA, Chicago, Harvard e in molti altri stili
Vedi i top-50 articoli di riviste per l'attività di ricerca sul tema "Interpretability of AI Models for Parkinson's Disease Detection".
Accanto a ogni fonte nell'elenco di riferimenti c'è un pulsante "Aggiungi alla bibliografia". Premilo e genereremo automaticamente la citazione bibliografica dell'opera scelta nello stile citazionale di cui hai bisogno: APA, MLA, Harvard, Chicago, Vancouver ecc.
Puoi anche scaricare il testo completo della pubblicazione scientifica nel formato .pdf e leggere online l'abstract (il sommario) dell'opera se è presente nei metadati.
Vedi gli articoli di riviste di molte aree scientifiche e compila una bibliografia corretta.
1
Samuel Fanijo, Uyok Hanson, Taiwo Akindahunsi, Idris Abijo e Tinuade Bolutife Dawotola. "Artificial intelligence-powered analysis of medical images for early detection of neurodegenerative diseases". World Journal of Advanced Research and Reviews 19, n. 2 (30 agosto 2023): 1578–87. http://dx.doi.org/10.30574/wjarr.2023.19.2.1432.
Testo completoAbstract (sommario):
Neurodegenerative diseases including Alzheimer's, Parkinson's, and Huntington's offer serious health issues to people all over the world, due to their progressive nature and lack of effective therapies. In order to improve patient outcomes and allow for prompt action to limit the progression of the disease, early identification is essential. With a focus on deep learning methods, this study investigates the use of AI-powered analysis of medical images for the early detection of neurodegenerative disorders. The use of several medical imaging modalities, such as PET, CT, and MRI, in identifying disease biomarkers at an early stage is investigated. The usefulness of deep learning techniques to automate feature extraction, categorize illness states, and track disease progression is highlighted. These techniques include convolutional neural networks [CNNs], recurrent neural networks [RNNs], and generative adversarial networks [GANs]. The study also discusses the difficulties in using AI implementation, including data quality, image variability, and the interpretability of AI models. Furthermore, the study explores possible regulatory and ethical considerations in clinical adoption. It also examines AI's growing role in clinical settings and its ability to work with personalized medicine which present promising opportunities for improving the diagnosis and management neurodegenerative disease. The final section of this paper outlines important future directions for increasing the use of AI in clinical care, including multi-modal fusion and transfer learning.
2
Adeniran, Opeyemi Taiwo, Blessing Ojeme, Temitope Ezekiel Ajibola, Ojonugwa Oluwafemi Ejiga Peter, Abiola Olayinka Ajala, Md Mahmudur Rahman e Fahmi Khalifa. "Explainable MRI-Based Ensemble Learnable Architecture for Alzheimer’s Disease Detection". Algorithms 18, n. 3 (13 marzo 2025): 163. https://doi.org/10.3390/a18030163.
Testo completoAbstract (sommario):
With the advancements in deep learning methods, AI systems now perform at the same or higher level than human intelligence in many complex real-world problems. The data and algorithmic opacity of deep learning models, however, make the task of comprehending the input data information, the model, and model’s decisions quite challenging. This lack of transparency constitutes both a practical and an ethical issue. For the present study, it is a major drawback to the deployment of deep learning methods mandated with detecting patterns and prognosticating Alzheimer’s disease. Many approaches presented in the AI and medical literature for overcoming this critical weakness are sometimes at the cost of sacrificing accuracy for interpretability. This study is an attempt at addressing this challenge and fostering transparency and reliability in AI-driven healthcare solutions. The study explores a few commonly used perturbation-based interpretability (LIME) and gradient-based interpretability (Saliency and Grad-CAM) approaches for visualizing and explaining the dataset, models, and decisions of MRI image-based Alzheimer’s disease identification using the diagnostic and predictive strengths of an ensemble framework comprising Convolutional Neural Networks (CNNs) architectures (Custom multi-classifier CNN, VGG-19, ResNet, MobileNet, EfficientNet, DenseNet), and a Vision Transformer (ViT). The experimental results show the stacking ensemble achieving a remarkable accuracy of 98.0% while the hard voting ensemble reached 97.0%. The findings present a valuable contribution to the growing field of explainable artificial intelligence (XAI) in medical imaging, helping end users and researchers to gain deep understanding of the backstory behind medical image dataset and deep learning model’s decisions.
3
Hamza, Naeem, Nuaman Ahmed e Naeema Zainaba. "A Comparative Analysis of Traditional and AI-Driven Methods for Disease Detection: Novel Approaches, Methodologies, and Challenges". Journal of Medical Health Research and Psychiatry 01, n. 02 (2024): 01–03. https://doi.org/10.70844/jmhrp.2024.1.2.28.
Testo completoAbstract (sommario):
Background: Accurate and early disease detection is crucial for improving patient outcomes. Traditional methods have relied on manual medical data analysis, which can be labor-intensive and error prone. Methods: This comparative review examines traditional versus AI-driven detection methods, highlighting their applications, advantages, and limitations. We employed PRISMA guidelines to systematically review the literature, using strict inclusion and exclusion criteria to evaluate relevant studies. Results: Our findings suggest that while AI-driven methods outperform traditional approaches in terms of speed and accuracy, challenges such as algorithm interpretability and data quality remain significant barriers. Conclusions: Novel aspects of this study include an in-depth comparison of AI models, their integration into clinical practice, and the challenges of data quality and regulatory frameworks. Overall, AI-driven methods have the potential to revolutionize disease detection, but addressing the challenges of algorithm interpretability and data quality is crucial for their successful integration into clinical practice.
4
Fatima, Shereen, Hidayatullah Shaikh, Attaullah Sahito e Asadullah Kehar. "A Review of Skin Disease Detection Using Deep Learning". VFAST Transactions on Software Engineering 12, n. 4 (31 dicembre 2024): 220–38. https://doi.org/10.21015/vtse.v12i4.2022.
Testo completoAbstract (sommario):
Amid increasing concerns about skin diseases exacerbated by climate change or lifestyle, some diseases are undiagnosed or misdiagnosed due to limited healthcare facilities. The worldwide health burden emphasizes the need for innovative diagnostics. This study explores the evolutionary role of deep learning in skin disease detection, providing the most advanced and effective research approaches, model achievements, and dataset usage exclusively. The review adapts data from 30 research papers and many datasets to address imbalanced class and various efficiency factors. The developments in CNN models like MobileNet or EfficientNet, have strengthened computational potential, while hybrid models have accommodated local and global features. Furthermore, Explainable AI (EXI) and augmented datasets have overcome the challenges including noisy, biased datasets and the less interpretable AI models. This study declares the innovative capacity of deep learning in dermatological analysis, highlighting its scalability and performance. Future research is required to consider dataset diversity, interpretability, and incorporating medical metadata to enhance model performances.
5
Hasan Saif, Fatima, Mohamed Nasser Al-Andoli e Wan Mohd Yaakob Wan Bejuri. "Explainable AI for Alzheimer Detection: A Review of Current Methods and Applications". Applied Sciences 14, n. 22 (5 novembre 2024): 10121. http://dx.doi.org/10.3390/app142210121.
Testo completoAbstract (sommario):
Alzheimer’s disease (AD) is the most common cause of dementia, marked by cognitive decline and memory loss. Recently, machine learning and deep learning techniques have introduced promising solutions for improving AD detection through MRI, especially in settings where specialists may not be readily available. These techniques offer the potential to assist general practitioners and non-specialists in busy clinical environments. However, the ‘black box’ nature of many AI techniques makes it challenging for non-expert physicians to fully trust their diagnostic accuracy. In this review, we critically evaluate current explainable AI (XAI) methods applied to AD detection and highlight their limitations. In addition, a new interpretability framework, called “Feature-Augmented”, was theoretically designed to improve model interpretability. This approach remains underexplored, primarily due to the scarcity of explainable AD-specific datasets. Furthermore, we underscore the importance of AI models being accurate and explainable, which enhance diagnostic confidence and patient care outcomes.
6
Rakhi Raghukumar, Aswathi V Nair, Amrutha Raju, Aina S Dcruz e Susheel George Joseph. "AI Used to Predict Alzheimer’s Disease". International Research Journal on Advanced Engineering and Management (IRJAEM) 2, n. 12 (12 dicembre 2024): 3647–51. https://doi.org/10.47392/irjaem.2024.0541.
Testo completoAbstract (sommario):
Alzheimer's disease (AD) is a progressive neurodegenerative disorder that leads to cognitive decline and memory loss, severely affecting millions worldwide. Early detection and accurate prediction of Alzheimer's are critical for timely interventions. This paper explores the application of Artificial Intelligence (AI) in predicting Alzheimer's disease, focusing on machine learning (ML) models, neural networks, and deep learning (DL) techniques. By analyzing a combination of neuroimaging data, genetic information, and cognitive test results, AI systems can identify subtle patterns and biomarkers that indicate the onset of AD even before the appearance of clinical symptoms. The paper discusses the integration of AI with brain imaging technologies, such as MRI and PET scans, as well as the role of natural language processing (NLP) in evaluating speech and text patterns. Key challenges such as data quality, interpretability, and the need for large, diverse datasets are also addressed. The potential for AI to enhance diagnostic accuracy and facilitate personalized treatment approaches in Alzheimer’s care is highlighted, along with future directions for research in this field. The results suggest that AI has the capacity to significantly improve early detection and intervention strategies, ultimately advancing the fight against Alzheimer's disease.
7
Ismail Y e Vijaya Kumar Voleti. "A Review on Usage of Artificial Intelligence for Early Detection and Management of Alzheimer's Disease". Journal of Pharma Insights and Research 2, n. 5 (4 ottobre 2024): 007–13. http://dx.doi.org/10.69613/06tz7453.
Testo completoAbstract (sommario):
Artificial Intelligence (AI) has emerged as a powerful tool in Alzheimer's disease (AD) research and clinical practice. This review discusses about the recent advances in AI applications for AD, focusing on neuroimaging analysis, biomarker discovery, cognitive assessment, and predictive modeling. AI techniques, particularly deep learning algorithms, have significantly improved the accuracy and efficiency of brain imaging interpretation, enabling earlier detection of AD-related structural and functional changes. In biomarker research, AI has accelerated the identification of novel blood-based and CSF markers, potentially leading to less invasive and more cost-effective diagnostic methods. AI-driven cognitive assessment tools, including computerized tests and speech analysis, offer more sensitive measures of cognitive decline. Additionally, AI-based predictive models integrating multiple data types show promise in personalized risk assessment and disease progression forecasting. Despite these advancements, challenges remain in data standardization, model interpretability, and ethical considerations. This review explains about the current state of AI in AD research, its potential impact on patient care, and areas requiring further investigation to fully realize the benefits of AI in combating Alzheimer's disease
8
Paul, Tanmoy, Omiya Hassan, Christina S. McCrae, Syed Kamrul Islam e Abu Saleh Mohammad Mosa. "An Explainable Fusion of ECG and SpO2-Based Models for Real-Time Sleep Apnea Detection". Bioengineering 12, n. 4 (3 aprile 2025): 382. https://doi.org/10.3390/bioengineering12040382.
Testo completoAbstract (sommario):
Obstructive sleep apnea (OSA) is a common disorder characterized by disrupted breathing during sleep, leading to serious health consequences such as daytime fatigue, hypertension, metabolic issues, and cardiovascular disease. Polysomnography (PSG) is the standard diagnostic method but is costly and uncomfortable for patients, which has led to interest in artificial intelligence (AI) for automated OSA detection. To develop an explainable AI model that utilizes electrocardiogram (ECG) and blood oxygen saturation (SpO2) data for real-time apnea detection, providing visual explanations to enhance interpretability and support clinical decisions. It emphasizes giving visual explanations to show how specific segments of the signal contribute to the AI’s conclusions. Furthermore, it explores the combination of individual models to improve detection accuracy. The fusion of individual models demonstrates an enhanced performance in detection accuracy. Visual explanations for AI decisions highlight the importance of certain signal features, making the model’s operations transparent to healthcare providers. The proposed AI model addresses the crucial need for transparent and interpretable AI in healthcare. By providing real-time, explainable OSA detection, this approach represents a significant advancement in the field, potentially improving patient care and aiding in the early identification and management of OSA.
9
Sarma Borah, Proyash Paban, Devraj Kashyap, Ruhini Aktar Laskar e Ankur Jyoti Sarmah. "A Comprehensive Study on Explainable AI Using YOLO and Post Hoc Method on Medical Diagnosis". Journal of Physics: Conference Series 2919, n. 1 (1 dicembre 2024): 012045. https://doi.org/10.1088/1742-6596/2919/1/012045.
Testo completoAbstract (sommario):
Abstract Medical imaging plays a pivotal role in disease detection and intervention. The black-box nature of deep learning models, such as YOLOv8, creates challenges in interpreting their decisions. This paper presents a toolset to enhance interpretability in AI based diagnostics by integrating Explainable AI (XAI) techniques with YOLOv8. This paper explores implementation of post hoc methods, including Grad-CAM and Eigen CAM, to assist end users in understanding the decision making of the model. This comprehensive evaluation utilises CT-Datasets, demonstrating the efficacy of YOLOv8 for object detection in different medical fields. This paper compares the interpretability offered by different post hoc methods, shedding light on abnormalities detected by the model. Moreover, this paper introduces a user-friendly interface for end users, incorporating the generated heat maps for intuitive understanding using different CAM algorithms. These findings underscore the importance of XAI in medical image analysis and offer a practical framework for improving interpretability in X-ray diagnostics. The comparison of the different CAM methods can offer a choice for end users to determine the best fit for deployable tools. This work contributes to bridging the gap between sophisticated deep learning models and actionable insights for professionals. Access at https://spritan.github.io/YOLOv8_Explainer/
10
Gupta, Ayush, Jeya Mala D., Vishal Kumar Yadav e Mayank Arora. "Dissecting Retinal Disease: A Multi-Modal Deep Learning Approach with Explainable AI for Disease Classification across Various Classes". International Journal of Online and Biomedical Engineering (iJOE) 21, n. 02 (17 febbraio 2025): 38–51. https://doi.org/10.3991/ijoe.v21i02.51409.
Testo completoAbstract (sommario):
This study investigates the efficacy of various deep learning (DL) models in detecting retinal diseases, specifically focusing on cataract detection. Utilizing a pre-processed fundus images data set classified into normal and cataract classes, we evaluate the performance of ResNet, VGG-16 and VGG-19 models based on accuracy, sensitivity, and specificity in classifying fundus images. The primary objective of this work is to provide explanations on the predictions done by the said DL models in order to ensure the ground-truth verification. The explanation is achieved using the explainable artificial intelligence (XAI) model namely gradient-weighted class activation mapping (Grad-CAM), which helps to visualize and interpret the decision-making process of these models. Through a comprehensive exploratory data analysis (EDA), model training, and evaluation, VGG-19 emerged as the superior model, achieving the highest accuracy, precision, and recall. Grad-CAM heat maps provide insights into the models’ attention in image features, highlighting the impact of cataracts on retinal structure. The study underscores the potential of DL in retinal disease detection and the pivotal role of explainable artificial intelligence (XAI) in enhancing model interpretability. Future directions include exploring more advanced DL architectures and furthering the application of XAI techniques to improve detection systems’ accuracy and transparency.
11
Khan, Mohammad Badhruddouza, Salwa Tamkin, Jinat Ara, Mobashwer Alam e Hanif Bhuiyan. "CropsDisNet: An AI-Based Platform for Disease Detection and Advancing On-Farm Privacy Solutions". Data 10, n. 2 (18 febbraio 2025): 25. https://doi.org/10.3390/data10020025.
Testo completoAbstract (sommario):
Crop failure is defined as crop production that is significantly lower than anticipated, resulting from plants that are harmed, diseased, destroyed, or influenced by climatic circumstances. With the rise in global food security concern, the earliest detection of crop diseases has proven to be pivotal in agriculture industries to address the needs of the global food crisis and on-farm data protection, which can be met with a privacy-preserving deep learning model. However, deep learning seems to be a largely complex black box to interpret, necessitating a prerequisite for the groundwork of the model’s interpretability. Considering this, the aim of this study was to follow up on the establishment of a robust deep learning custom model named CropsDisNet, evaluated on a large-scale dataset named “New Bangladeshi Crop Disease Dataset (corn, potato and wheat)”, which contains a total of 8946 images. The integration of a differential privacy algorithm into our CropsDisNet model could establish the benefits of automated crop disease classification without compromising on-farm data privacy by reducing training data leakage. To classify corn, potato, and wheat leaf diseases, we used three representative CNN models for image classification (VGG16, Inception Resnet V2, Inception V3) along with our custom model, and the classification accuracy for these three different crops varied from 92.09% to 98.29%. In addition, demonstration of the model’s interpretability gave us insight into our model’s decision making and classification results, which can allow farmers to understand and take appropriate precautions in the event of early widespread harvest failure and food crises.
12
Jafar, Abbas, e Myungho Lee. "Enhancing Kidney Disease Diagnosis Using ACO-Based Feature Selection and Explainable AI Techniques". Applied Sciences 15, n. 6 (10 marzo 2025): 2960. https://doi.org/10.3390/app15062960.
Testo completoAbstract (sommario):
Kidney disease is a global health concern, impacting a substantial part of the overall population and contributing to high morbidity and mortality rates. The initially diagnosed phases of kidney disease are often present without noticeable indications, leading to delayed diagnosis and treatment. Therefore, early detection is crucial to reducing complications and improving the lives of those impacted. However, the performance of previous automated approaches has often been hindered by suboptimal feature selection and algorithms’ “black-box” nature, which adversely affect their interpretability and clinical applicability. This paper aims to address these limitations by creating an effective machine-learning-based approach that integrates ant colony metaheuristic optimization algorithms for feature selection and explainable artificial intelligence techniques such as SHAP and LIME for model interpretation. The ant colony optimization method identified the most relevant feature subsets using a clinical dataset, reducing model complexity while preserving predictive accuracy. Performance evaluation shows that the extra trees classifier, when using optimized selected features, achieved the highest performance with an accuracy of 97.70% and an area under the curve of 99.55%, outperforming previous models trained on raw and complete processed feature sets. To enhance interpretability, the SHAP and LIME explainable techniques were employed, providing detailed insights into the contribution of key features such as TimeToEventMonths, HistoryDiabetes, and Age. This comprehensive framework, combining advanced feature selection with explainable models, improves clinical decision-making and fosters trust in machine learning applications for healthcare.
13
Alborzi, Melina, e Parsa Abadi. "OSTEO-AI: A Systematic Review and Meta-Analysis of Artificial Intelligence Models for Osteoarthritis and Osteoporosis Detection and Prognosis". Undergraduate Research in Natural and Clinical Science and Technology (URNCST) Journal 9, n. 2 (14 febbraio 2025): 1–14. https://doi.org/10.26685/urncst.783.
Testo completoAbstract (sommario):
Introduction: Osteoarthritis (OA) and osteoporosis are leading degenerative bone diseases that diminish quality of life and impose significant socioeconomic costs. Traditional diagnostic approaches, including imaging and bone density assessments, often fail to detect disease in its early stages, delaying critical interventions. Emerging artificial intelligence (AI) techniques, particularly those employing machine learning (ML) and deep learning (DL), offer promising avenues for early detection and more accurate prognostication. Methods: We conducted a systematic review of AI models developed between 2018 and 2024, assessing their performance in diagnosing and predicting the progression of OA and osteoporosis. Studies utilizing supervised or unsupervised methods applied to imaging modalities (e.g., X-ray, MRI, DXA) or clinical data were included. We evaluated model accuracy, reliability, clinical applicability, and generalizability. Quality and risk of bias were assessed using a modified CLAIM framework, ensuring alignment with transparency, validity, and clinical integration standards. Results: Of 2,300 identified articles, 33 studies met the inclusion criteria. Top-performing models for OA reached up to 97% accuracy, with one study achieving an AUC of 0.93 for MRI-based progression prediction. For osteoporosis, the strongest models attained a C-index of 0.90 using DXA imaging, indicating robust fracture risk prediction. Nevertheless, many studies relied on geographically or demographically homogeneous datasets, limiting broader applicability. Only 15% included external validation, and a substantial proportion lacked interpretability features essential for clinical adoption. Discussion: AI-driven models outperformed conventional diagnostic tools in accuracy and early disease detection. However, the limited dataset diversity, infrequent external validation, and insufficient model interpretability pose barriers to clinical integration. The reliance on male-dominant datasets for osteoporosis and geographically narrow cohorts for OA underscores the need for broader data representation. Standardizing evaluation metrics and improving explainability will enhance cross-study comparisons and support adoption in practice. Conclusion: AI holds transformative potential for improving OA and osteoporosis diagnostics, facilitating earlier interventions, and informing personalized patient management. Future work should prioritize diverse, well-validated datasets; transparent, clinician-friendly interfaces; and standardized performance metrics. Addressing these challenges will enable AI to evolve from a promising innovation into a cornerstone of global musculoskeletal healthcare.
14
Singh, Jaswinder, e Gaurav Dhiman. "A Review on Predictive Analytics for Early Disease Detection in Neonatal Healthcare using Artificial Intelligence". Journal of Neonatal Surgery 14, n. 5S (15 marzo 2025): 831–42. https://doi.org/10.52783/jns.v14.2158.
Testo completoAbstract (sommario):
The early detection of diseases plays a crucial role in improving patient outcomes, reducing healthcare costs, and enabling timely interventions. In recent years, the integration of Artificial Intelligence (AI) and Predictive Analytics (PA) has emerged as a transformative approach in healthcare, offering significant advancements in detecting diseases at their earliest stages. This paper provides a comprehensive review of the application of AI-driven predictive analytics in early disease detection, focusing on various AI techniques such as machine learning (ML), deep learning (DL), natural language processing (NLP), and neural networks. These techniques have shown exceptional promise in identifying patterns and correlations within medical data—including electronic health records (EHRs), medical imaging, genetic data, and wearable devices—that can signal the onset of diseases before they become clinically evident. The paper discusses the effectiveness of AI-based predictive models in detecting a wide range of diseases, including cancer, cardiovascular diseases, diabetes, neurological disorders, neonatal conditions, and infectious diseases. Special attention is given to AI applications in neonatal healthcare, where early detection of conditions such as neonatal sepsis, respiratory distress syndrome, and congenital anomalies can significantly improve survival rates and long-term health outcomes. By leveraging large datasets and advanced algorithms, AI systems can provide accurate predictions, risk assessments, and personalized treatment plans, leading to improved early diagnosis and targeted interventions. However, the integration of AI in disease detection also presents challenges such as data privacy concerns, model interpretability, ethical issues, and the need for robust regulatory frameworks. Furthermore, the paper highlights key advancements in AI technologies that have contributed to the success of predictive analytics in healthcare, along with real-world applications, case studies, and examples of AI models that have been implemented in clinical settings. The limitations and potential solutions to these challenges are also examined, with an emphasis on the importance of high-quality, representative datasets and continuous collaboration between AI researchers, clinicians, and regulatory bodies. This review aims to provide a thorough understanding of the current landscape of AI-powered predictive analytics for early disease detection and to highlight future directions in the field. As AI technologies continue to evolve, their role in enhancing early disease detection, particularly in neonatal care, improving patient outcomes, and enabling preventive healthcare will become increasingly significant, ultimately leading to a more efficient, effective, and equitable healthcare system.
15
Abbas, Shahid, Abdul Sattar, Syeda Hina Shah, Sidrah Hafeez, Waqas Mahmood, Raza iqbal, Keziah Shaheen, Pervaiz Azam e Tazeem Shahbaz. "THE ROLE OF ARTIFICIAL INTELLIGENCE IN PERSONALIZED MEDICINE AND PREDICTIVE DIAGNOSTICS – A NARRATIVE REVIEW". Insights-Journal of Health and Rehabilitation 3, n. 1 (Health & Allied) (24 febbraio 2025): 624–31. https://doi.org/10.71000/k6cga886.
Testo completoAbstract (sommario):
Background: Artificial intelligence (AI) has revolutionized personalized medicine and predictive diagnostics by enabling data-driven, individualized healthcare strategies. AI-powered models leverage vast datasets, including genomic, proteomic, and clinical information, to improve disease detection, optimize treatment selection, and enhance patient outcomes. With the increasing burden of chronic diseases and the growing demand for precision medicine, AI presents significant opportunities to transform traditional healthcare paradigms. However, challenges related to clinical implementation, algorithmic bias, and regulatory considerations necessitate a critical evaluation of its applications. Objective: This narrative review aims to explore the role of AI in personalized medicine and predictive diagnostics, analyzing its clinical applications, benefits, limitations, and future directions. The review synthesizes current evidence on AI-driven advancements in disease diagnosis, risk stratification, and treatment optimization while addressing key challenges hindering its widespread adoption. Main Discussion Points: AI has demonstrated superior diagnostic accuracy in various medical domains, including oncology, cardiology, and neurology, through deep learning and machine learning algorithms. Predictive models enhance risk assessment, enabling early intervention and personalized therapeutic approaches. Despite these advancements, methodological limitations, variability in outcome measurement, and concerns regarding data standardization and interpretability pose significant barriers. Ethical considerations, regulatory frameworks, and the need for unbiased, transparent AI models remain critical challenges in integrating AI into routine clinical practice. Conclusion: AI holds immense potential in advancing personalized medicine and predictive diagnostics, yet its real-world application requires rigorous validation, standardized protocols, and ethical oversight. Future research should focus on developing explainable AI models, conducting large-scale randomized controlled trials, and ensuring equitable healthcare access to maximize AI’s impact on patient care.
16
Nasheet Tarik. "Bridging the Gaps in AI-Driven Healthcare: Enhancing Interpretability, Affordability, and Security for Scalable Patient-Centered Solutions". Journal of Information Systems Engineering and Management 10, n. 19s (12 marzo 2025): 74–86. https://doi.org/10.52783/jisem.v10i19s.2977.
Testo completoAbstract (sommario):
Background: The integration of Artificial Intelligence (AI) and machine-assisted technologies in healthcare has significantly transformed diagnostics, treatment, and patient monitoring. AI-driven solutions, including deep learning for medical imaging, robotic-assisted surgery, and real-time patient analytics, have improved clinical decision-making and patient outcomes. However, challenges such as lack of interpretability, high implementation costs, and data security concerns hinder the full-scale adoption of these technologies. Purpose: This study aims to analyze the role of AI and robotics in modern healthcare, highlighting advancements in disease diagnosis, robotic rehabilitation, and predictive analytics. The research also identifies existing limitations and proposes methodologies to improve AI model transparency, reduce costs, and enhance real-time patient monitoring. Methods: A systematic literature review was conducted, analyzing AI-driven healthcare applications in diagnostics, robotic-assisted treatment, and predictive analytics. The study integrates Explainable AI (XAI), federated learning, Internet of Medical Things (IoMT), and blockchain-based security frameworks to propose solutions for current challenges. Comparative performance analysis of AI models and robotic frameworks was carried out to assess efficiency, cost-effectiveness, and adaptability in clinical environments. Results: Findings indicate that AI-based medical imaging improves disease detection accuracy by 92.5%, while robotic-assisted treatments enhance patient recovery rates by 78.2%. IoMT-powered real-time patient monitoring has demonstrated 85.1% efficiency in detecting early signs of critical conditions. However, challenges such as high computational costs, lack of standardized AI frameworks, and ethical concerns remain significant barriers to adoption. Conclusion: AI-driven healthcare solutions offer immense potential in improving medical diagnostics, precision surgeries, and patient monitoring. However, addressing issues related to model interpretability, cost reduction, and ethical AI deployment is crucial for broader implementation. Future research should focus on developing scalable, secure, and real-time adaptive AI-driven healthcare systems to optimize patient outcomes worldwide.
17
Alhassun, Wejdan H., Abdulaziz S. Alothman e Sultan A. Alfawaz. "The Role of AI in Early Detection of Alzheimer's and Parkinson's Diseases: A Literature Survey". Asian Journal of Research in Computer Science 18, n. 2 (4 febbraio 2025): 186–96. https://doi.org/10.9734/ajrcos/2025/v18i2570.
Testo completoAbstract (sommario):
Early detection of neurodegenerative diseases like Alzheimer’s and Parkinson’s is crucial for improving patient care and enabling timely interventions. Artificial intelligence (AI) offers innovative approaches to analyzing complex medical datasets, revolutionizing the detection of these diseases at early stages. This review discusses key AI methodologies, including machine learning (ML), deep learning (DL), natural language processing (NLP), and reinforcement learning (RL), and their applications in early diagnosis. ML models excel in predicting disease risk and classifying imaging and biometric data, while DL techniques, such as convolutional and recurrent neural networks, are effective in processing unstructured data like images and speech. NLP facilitates extracting critical insights from clinical notes and patient narratives, and RL enhances decision-making in diagnostic workflows. Integrating multimodal data—such as genomics, neuroimaging, wearable device metrics, and electronic health records—further strengthens diagnostic precision. Despite its promise, the widespread implementation of AI faces challenges, including the need for standardized data, ethical considerations, and clinical validation. Overcoming these obstacles is essential for AI to transform early detection and management of neurodegenerative diseases. This review emphasizes the significance of interdisciplinary efforts and sustained research to unlock AI’s full potential in medical applications.
18
Venugopal Boppana. "Plant Disease Detection Using Hybrid MobileNetV2- Compact CNN Architecture with LIME Integration". Journal of Information Systems Engineering and Management 10, n. 13s (10 febbraio 2025): 554–68. https://doi.org/10.52783/jisem.v10i13s.2111.
Testo completoAbstract (sommario):
This paper presents an advanced approach to plant disease detection by implementing explainable AI techniques that combine MobileNetV2 architecture with transfer learning and compact convolutional neural networks (CNN). The study compares three distinct models' performance on a plant leaf disease dataset, revealing MobileNetV2's superior accuracy of 95% with 94% precision in disease classification, despite requiring 850 seconds for training. The Compact CNN achieved 82% accuracy with minimal training time of 420 seconds, demonstrating its efficiency for resource-constrained applications. Disease-specific analysis showed exceptional detection rates for common plant diseases, with Apple Scab at 96.5%, Black Rot at 94.8%, and Cedar Rust at 95.2%. The integration of LIME (Local Interpretable Model-agnostic Explanations) provided transparent insights into the model's decision-making process, while the Compact CNN demonstrated 45% reduced memory usage compared to MobileNetV2. This implementation establishes a robust framework for practical agricultural applications, balancing high accuracy with computational efficiency and interpretability.
19
Meena Jindal e Khushwant Kaur. "Enhancing Agricultural Sustainability Through AI-Powered Image Processing: Review Study on Plant Disease Detection". International Journal of Scientific Research in Science and Technology 11, n. 6 (10 dicembre 2024): 490–96. https://doi.org/10.32628/ijsrst24114312.
Testo completoAbstract (sommario):
The agricultural field is encountering multiple problems with the climate changing, population booming and overusing chemical pesticide which all lead to unsustainable agriculture. Affecting quality and yield, plant diseases account for a heavy loss from the final production. Conventional plant disease detection is definitely the aforementioned matter as well, profound education analyzing with labor-intensive and time-consuming procedure yet not so accurate. By merging artificial intelligence (AI) with image processing, plant disease diagnosis can be automated quickly and efficiently. It uses machine learning algorithms, combined with high-resolution imagery to detect disease symptoms in the early stage of infestation thereby making the treatment process largely dependent on chemical control. In this paper, we reviewed state-of-the-art methods which have experience significant improvement and development in terms of image processing approaches using AI for plant disease recognition. We made a lot of progress however there are still many gaps to fill like other data types, real-time processing and generalizability models that need to be incorporated with farming practices as well accessibility considering all the factors is important for economic viability. Overcoming these gaps requires a holistic approach by combining AI innovations with perspectives from the fields of agronomy and agricultural economics. Future research could potentially concentrate in improving the real-time process, increasing model interpretability and integration with current agricultural systems. Overcoming these challenges, AI-powered image processing can be the backbone of precision agriculture that could secure our food supply and make farming more sustainable.
20
Abukaresh, ALaa Im, e Ali Okatan. "AI-Based Early Detection of Parkinson's Disease using Mri: A Comparative Analysis of Densenet121 and Resnet Models". EURAS Journal of Engineering and Applied Sciences 4, n. 2 (2021): 81–117. http://dx.doi.org/10.17932/ejeas.2021.024/ejeas_v04i2003.
Testo completo
21
Junior, Kamese Jordan, Kouayep Sonia Carole, Tagne Poupi Theodore Armand, Hee-Cheol Kim e The Alzheimer’s Disease Neuroimaging Initiative The Alzheimer’s Disease Neuroimaging Initiative. "Alzheimer’s Multiclassification Using Explainable AI Techniques". Applied Sciences 14, n. 18 (14 settembre 2024): 8287. http://dx.doi.org/10.3390/app14188287.
Testo completoAbstract (sommario):
In this study, we address the early detection challenges of Alzheimer’s disease (AD) using explainable artificial intelligence (XAI) techniques. AD, characterized by amyloid plaques and tau tangles, leads to cognitive decline and remains hard to diagnose due to genetic and environmental factors. Utilizing deep learning models, we analyzed brain MRI scans from the ADNI database, categorizing them into normal cognition (NC), mild cognitive impairment (MCI), and AD. The ResNet-50 architecture was employed, enhanced by a channel-wise attention mechanism to improve feature extraction. To ensure model transparency, we integrated local interpretable model-agnostic explanations (LIMEs) and gradient-weighted class activation mapping (Grad-CAM), highlighting significant image regions contributing to predictions. Our model achieved 85% accuracy, effectively distinguishing between the classes. The LIME and Grad-CAM visualizations provided insights into the model’s decision-making process, particularly emphasizing changes near the hippocampus for MCI. These XAI methods enhance the interpretability of AI-driven AD diagnosis, fostering trust and aiding clinical decision-making. Our approach demonstrates the potential of combining deep learning with XAI for reliable and transparent medical applications.
22
Shakir, Daniah Abdul Qahar, e Eman Turki Mahdi. "Machine Learning Techniques for Skin Fungal Infection Detection -A Review". Mesopotamian Journal of Artificial Intelligence in Healthcare 2024 (5 dicembre 2024): 177–83. https://doi.org/10.58496/mjaih/2024/018.
Testo completoAbstract (sommario):
Skin diseases are the most common than other diseases. Skin diseases may be caused by fungal infection, allergy, bacteria, viruses, etc. Fungal disease affects more than billions people worldwide. The accurately of diagnosing skin fungal infections can be challenging in their early stages or present with symptoms, which mimic other dermatological disorders. The identification and categorization of skin fungal infections could be improved by recent developments in deep learning and artificial intelligence (AI). It offers a more effective and dependable substitute for conventional diagnostic techniques. Here we tried to focus on various methods for identifying fungal skin that rely on deep learning (such as transformers, convolutional neural networks, and hybrid models), the difficulties related to data diversity and availability, going over the shortcomings of current datasets, how data augmentation and synthetic data creation which might help close these gaps. We also investigate how improving interpretability and usability can help clinical uptake of AI-based diagnostic systems. Finally, the study concludes with suggestions for further research, highlighting the revolutionary potential of deep learning in dermatology and stressing the necessity of sophisticated model architectures, a wide range of high-quality datasets, and thorough clinical validation.
23
Roshan, M. K. Gagan. "Multiclass Medical X-ray Image Classification using Deep Learning with Explainable AI". International Journal for Research in Applied Science and Engineering Technology 10, n. 6 (30 giugno 2022): 4518–26. http://dx.doi.org/10.22214/ijraset.2022.44541.
Testo completoAbstract (sommario):
Abstract: COVID-19, also known as novel coronavirus, created a colossal health crisis worldwide. This virus is a disease that basically comes from Severe Acute Respiratory Syndrome (SARS) and Middle East Respiratory Syndrome (MERS). A novel coronavirus, COVID-19, is the infection caused by SARS-CoV-2. The early detection of COVID-19 and the accurate separation of non-COVID-19 cases at the lowest cost and in the early stages of the disease are among the main challenges in the current COVID-19 pandemic. As pneumonia also is a significant indication of COVID-19, then it is necessary to detect it in the early stages. Another challenge is that it is very elusive to classify the chest xray between COVID-19 and pneumonia as the visual indications for both the labels are quite similar. The application of deep learning in the field of radiologic image processing reduces false-positive and negative errors in the detection of this disease and could offer a unique opportunity to provide fast, cheap, and safe diagnostic services to patients. Also, the Deep learning models are considered to be the “black boxes”. According to the ethics of AI in radiology, “transparency, interpretability, and explainability are necessary to build patient and provider trust”
24
Muriithi, Dennis Kariuki, Victor Wandera Lumumba, Olushina Olawale Awe e Daniel Mwangi Muriithi. "An Explainable Artificial Intelligence Models for Predicting Malaria Risk in Kenya". European Journal of Artificial Intelligence and Machine Learning 4, n. 1 (28 febbraio 2025): 1–8. https://doi.org/10.24018/ejai.2025.4.1.47.
Testo completoAbstract (sommario):
The article aims to develop interpretable Machine Learning models using R statistical programming language for malaria risk prediction in Kenya, emphasizing leveraging Explainable AI (XAI) techniques to support targeted interventions and improve early detection mechanisms. The methodology involved using synthetic data with 1000 observations, employing over-sampling to address class imbalance, utilizing two machine learning algorithms (Random Forest and Extreme Gradient Boosting), applying cross-validation techniques, Hyper-parameter tuning and implementing feature importance and SHAP (Shapley Additive Explanations) for model interpretability. The findings revealed that Random Forest outperformed Extreme Gradient Boosting with 98% accuracy. Critical prediction features included clinical symptoms such as nausea, muscle aches, and fever, plasmodium species identification, and environmental factors like rainfall and temperature. Both models demonstrated strong sensitivity in detecting malaria cases. This promotes trust in model predictions by clearly outlining the decision process for individual outcomes. The research concluded that integrating Explainable AI into malaria risk prediction represents a transformative approach to public health management. Through providing transparent, interpretable models, the research offers a robust, data-driven approach to predicting malaria risks, potentially empowering healthcare providers and policymakers to deploy resources more effectively and reduce the disease burden in endemic regions.
25
Isiaka, Salman O., Ronke S. Babatunde e Rafiu M. Isiaka. "Exploring Artificial Intelligence (AI) Technologies in Predictive Medicine: A Systematic Review". Kasu Journal of Computer Science 1, n. 2 (30 giugno 2024): 366–77. http://dx.doi.org/10.47514/kjcs/2024.1.2.0014.
Testo completoAbstract (sommario):
Artificial Intelligence (AI) technologies, including cognitive computing, machine learning (ML), and deep learning (DL), have the potential to transform healthcare. They enhance patient care, improve symptom differentiation, address medication adherence issues, and facilitate continuous patient support. AI's promise in electronic health records (EHRs) includes extracting valuable insights and improving healthcare infrastructure through advanced disease detection and diagnostics. This study reviews and analyzes AI applications in predictive medicine, focusing on emerging pattern recognition techniques and prediction model algorithms. The goal is to assess the current state of AI applications, identify gaps, and develop models that tackle challenges and ethical dilemmas in predictive medicine. A systematic review was conducted using five major scientific literature databases (ScienceDirect, PubMed, Scopus, IEEE Xplore, and Web of Science). The review covered publications from 2015 to May 2024, employing comprehensive search strings to identify relevant studies. Articles were filtered using inclusion and exclusion criteria, and data were extracted and classified based on AI algorithms, application contexts, and key results. Performance metrics of various AI models and techniques were analyzed following PRISMA guidelines. The review found that deep learning models, particularly Convolutional Neural Networks (CNNs) and Gradient Boosting Machines (GBMs), performed best in personalized medicine. Recurrent Neural Networks (RNNs) and Random Forests (RFs) excelled in disease prediction. Machine learning algorithms, feature selection techniques, data fusion methods, and natural language processing (NLP) techniques improved hospital discharge decision-making. Data fusion methods achieved the highest performance metrics, though challenges such as model interpretability and data complexity were identified.
26
Tulsani, Vijya, Prashant Sahatiya, Jignasha Parmar e Jayshree Parmar. "XAI Applications in Medical Imaging: A Survey of Methods and Challenges". International Journal on Recent and Innovation Trends in Computing and Communication 11, n. 9 (27 ottobre 2023): 181–86. http://dx.doi.org/10.17762/ijritcc.v11i9.8332.
Testo completoAbstract (sommario):
Medical imaging plays a pivotal role in modern healthcare, aiding in the diagnosis, monitoring, and treatment of various medical conditions. With the advent of Artificial Intelligence (AI), medical imaging has witnessed remarkable advancements, promising more accurate and efficient analysis. However, the black-box nature of many AI models used in medical imaging has raised concerns regarding their interpretability and trustworthiness. In response to these challenges, Explainable AI (XAI) has emerged as a critical field, aiming to provide transparent and interpretable solutions for medical image analysis. This survey paper comprehensively explores the methods and challenges associated with XAI applications in medical imaging. The survey begins with an introduction to the significance of XAI in medical imaging, emphasizing the need for transparent and interpretable AI solutions in healthcare. We delve into the background of medical imaging in healthcare and discuss the increasing role of AI in this domain. The paper then presents a detailed survey of various XAI techniques, ranging from interpretable machine learning models to deep learning approaches with built-in interpretability and post hoc interpretation methods. Furthermore, the survey outlines a wide range of applications where XAI is making a substantial impact, including disease diagnosis and detection, medical image segmentation, radiology reports, surgical planning, and telemedicine. Real-world case studies illustrate successful applications of XAI in medical imaging. The challenges associated with implementing XAI in medical imaging are thoroughly examined, addressing issues related to data quality, ethics, regulation, clinical integration, model robustness, and human-AI interaction. The survey concludes by discussing emerging trends and future directions in the field, highlighting the ongoing efforts to enhance XAI methods for medical imaging and the critical role XAI will play in the future of healthcare. This survey paper serves as a comprehensive resource for researchers, clinicians, and policymakers interested in the integration of Explainable AI into medical imaging, providing insights into the latest methods, successful applications, and the challenges that lie ahead.
27
Khushi Jha e Awadhesh Kumar. "Role of Artificial Intelligence in Detecting Neurological Disorders". International Research Journal on Advanced Engineering Hub (IRJAEH) 2, n. 02 (23 febbraio 2024): 73–79. http://dx.doi.org/10.47392/irjaeh.2024.0015.
Testo completoAbstract (sommario):
AI plays a pivotal role in detecting neurological disorders by leveraging advanced technologies to analyze vast amounts of data and aid in diagnosis. Here are several key roles AI plays. Artificial Intelligence (AI) has emerged as a revolutionary tool in the realm of healthcare, particularly in the early detection and accurate diagnosis of neurological disorders. The present paper delves into the multifaceted applications of AI specifically tailored to identify and discern various neurological conditions. AI's prowess in medical imaging analysis has significantly advanced the field by enabling nuanced and precise identification of neurological anomalies. By meticulously analyzing MRI scans, CT scans, and X-rays, AI-driven algorithms excel in detecting subtle patterns indicative of diverse neurological disorders such as Alzheimer's disease, Parkinson's disease, multiple sclerosis, and brain tumors. These technologies not only enhance diagnostic accuracy but also enable early intervention and improved patient outcomes. Moreover, AI leverages extensive datasets encompassing clinical records, genetic information, and biosensor data to predict and assess an individual's susceptibility to neurological disorders. Predictive analytics powered by machine learning models, aid in risk assessment, paving the way for personalized medicine and proactive healthcare strategies. Ethical considerations underscore the implementation of AI in neurological disorder detection, emphasizing the need for transparent algorithms, stringent data privacy protocols, and unbiased AI systems to ensure patient confidentiality and trust in healthcare. The evolving landscape of AI in neuroscience presents an exciting frontier, fostering collaborations between AI experts and neuroscientists. Together, they aim to unravel the intricacies of neurological disorders, pushing the boundaries of innovation and paving the path toward early detection, targeted treatments, and improved quality of life for individuals affected by these conditions. This paper highlights the transformative impact of AI in detecting neurological disorders, 7emphasizing its role in early detection, personalized medicine, ethical considerations, and the potential for collaborative advancements in neuroscience.
28
Singh, Jaswinder, e Gaurav Dhiman. "A Survey on Artificial Intelligence in Precision Medicine and Healthcare Analysis for Neonatal Surgery". Journal of Neonatal Surgery 14, n. 5S (15 marzo 2025): 799–808. https://doi.org/10.52783/jns.v14.2155.
Testo completoAbstract (sommario):
Artificial Intelligence (AI) has emerged as a transformative force in precision medicine, healthcare analysis, and neonatal surgery, enabling personalized treatment, early disease detection, and optimized clinical decision-making. This survey explores the evolving role of AI in healthcare, focusing on its applications, challenges, and future prospects. AI-driven approaches, including machine learning (ML) and deep learning (DL), have demonstrated remarkable accuracy in medical imaging, genomics, drug discovery, neonatal diagnostics, and patient risk assessment. These technologies enhance diagnostic precision, facilitate predictive analytics, and support real-time monitoring of chronic diseases and neonatal conditions. Precision medicine, which tailors treatments based on an individual’s genetic, environmental, and lifestyle factors, benefits significantly from AI-powered analytics. The integration of AI with electronic health records (EHRs), wearable devices, and biomedical data accelerates early disease identification and personalized therapeutic strategies, including those crucial for neonatal care and surgery. AI models trained on vast healthcare datasets can predict disease progression, recommend targeted therapies, and improve patient outcomes. Furthermore, natural language processing (NLP) enhances clinical documentation, reducing administrative burdens and improving efficiency in healthcare systems. Despite its potential, AI in precision medicine and neonatal surgery faces challenges, including data privacy concerns, model interpretability, and regulatory compliance. Ethical considerations, such as bias in AI models and equitable access to AI-driven healthcare, must be addressed to ensure responsible implementation. Additionally, integrating AI with traditional clinical workflows requires collaboration between healthcare professionals, data scientists, and policymakers. This survey provides a comprehensive analysis of AI applications in precision medicine, healthcare analysis, and neonatal surgery, highlighting key advancements, challenges, and future research directions. As AI continues to evolve, its role in revolutionizing healthcare will expand, paving the way for more efficient, accurate, and patient-centric medical practices. The findings of this survey aim to guide researchers, clinicians, and policymakers in leveraging AI for the next generation of precision healthcare, particularly in neonatal surgical interventions.
29
Patel, Mr Dhavalkumar Upendrabhai, e Dr Suchita Patel. "A Review, Synthesizing Frameworks, and Future Research Agenda: Use of AI & ML Models in Cardiovascular Diseases Diagnosis". International Journal of Innovative Technology and Exploring Engineering 12, n. 11 (30 ottobre 2023): 12–19. http://dx.doi.org/10.35940/ijitee.k9733.10121123.
Testo completoAbstract (sommario):
Cardiovascular diseases (CVDs) continue to be a leading cause of morbidity and mortality worldwide. Early detection and accurate diagnosis of the initial phases of CVDs are crucial for effective intervention and improved patient outcomes. In recent years, advances in intelligent automation and machine learning (ML) techniques have shown promise in enhancing the accuracy and efficiency of CVD detection. This systematic review aims to comprehensively analyze and synthesize the existing literature on the application of intelligent automation and ML adaptive classifier models in the detection of the initial phase of cardiovascular disease within the realm of medical science. The review follows a rigorous systematic methodology, including comprehensive literature search, study selection, data extraction, and quality assessment. A wide range of scholarly articles from the reputed journal were searched to identify relevant studies published over a specified period. The selected studies were critically evaluated for methodological robustness and relevance to the research objective. The synthesis of findings reveals a diverse landscape of research endeavors focused on employing intelligent automation and ML adaptive classifier models for CVD detection. The review highlights the various types of ML algorithms utilized, such as neural networks, decision trees, and support vector machines, and their potential to enhance the accuracy of diagnosis by analyzing complex and heterogeneous data sources, clinical records, and omics data. Furthermore, the review discusses challenges and limitations encountered in implementing these models, including data quality, interpretability, and ethical considerations. It also underscores the importance of interdisciplinary collaboration between medical practitioners, data scientists, and domain experts to ensure the seamless integration of these innovative technologies into clinical practice. In conclusion, this systematic review underscores the significant advancements made in the field of intelligent automation and ML adaptive classifier models in the detection of the initial phase of cardiovascular disease. While acknowledging the potential of these approaches, it also emphasizes the need for further research, standardization, and validation to harness their full capabilities and contribute to more accurate, timely and personalized cardiovascular disease diagnosis and management.
30
Kantapalli, Bhaskar, Arshia Aamena, Chebrolu Yogavarshinee, Badugu Divya Teja e Dasari Teja Sri. "OPTIMIZED SYMPTOM-BASED DEEP LEARNING FRAMEWORK FOR MONKEYPOX DIAGNOSIS WITH LIME EXPLAINABILITY". Industrial Engineering Journal 54, n. 03 (2025): 84–92. https://doi.org/10.36893/iej.2025.v54i3.009.
Testo completoAbstract (sommario):
Monkeypox is an emerging zoonotic disease that has raised global health concerns due to its increasing transmission rates. Traditional diagnostic methods rely on laboratory testing, which can be timeconsuming and inaccessible in resource-limited settings. This study presents an Optimized Deep Neural Framework (ODNF) to diagnose monkeypox based on clinical symptoms, leveraging deep learning for accurate and rapid classification. The research explores various machine learning models, including Random Forest, XG Boost, and Cat Boost, before implementing ODNF, which achieved superior performance with a 99% accuracy rate. The dataset underwent preprocessing steps, including handling imbalanced data and feature encoding, ensuring optimal learning. Additionally, Local Interpretable Model-Agnostic Explanations (LIME) was employed to enhance model interpretability, providing insights into symptom-based predictions. Comparative evaluation against traditional models demonstrated that ODNF outperforms existing approaches, making it a viable AI-based diagnostic tool for monkeypox detection.
31
Folorunsho, O., S. E. Akinsanya, O. A. Fagbuagun, S. A. Mogaji e S. K. Raji. "Explainable Ensemble Deep Learning Model for Predicting Diabetic Retinopathy Based on APTOS 2019 Eye Pack Dataset". LAUTECH Journal of Engineering and Technology 19, n. 1 (14 febbraio 2025): 1–14. https://doi.org/10.36108/laujet/5202.91.0110.
Testo completoAbstract (sommario):
Detection of diabetic retinopathy (DR) as early as possible is vital in mitigating the complicated issues associated with the disease. Recent advances in artificial intelligence (AI), particularly deep learning (DL) techniques, have led to appreciable increase in the accuracy of predicting various disease classes. However, the challenge of AI models is the difficulty in providing insights into how and why a model arrives in attaining decision-making to facilitate trust and adoption in clinical settings. Therefore, this study aimed to enhance the detection rate of DR and explain the significant regions on the image for the model's overall performance. This study utilised Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) networks, Simple Recurrent Neural Networks (SRNN), and XGBoost in an ensemble model (EM). Specifically, Shapley Additive exPlanations (SHAP), a popular Explainable Artificial Intelligence (XAI) technique was utilised to identify and provide insights to which parts of the images features that contribute to the model's overall performance. After a series of experiments using the APTOS 2019 eye pack dataset collected from the Kaggle repository to evaluate the performance of CNN, LSTM, SRNN, and XGBoost. The EM outperformed all the other models with 95.63% accuracy, 97.79% precision, 93.64% recall rate, 98.79% F1-score and 97.75% AUC score. Also, SHAP analysis revealed significant regions on the image that influenced predictions, thus showing how important interpretability was for the model. The results imply that the ensemble DL, particularly with XGBoost, enhances the detection of DR, thereby improving the efficiency of screening tests and supporting personalised treatment plans in clinical practice through integrating these advanced models with XAI tools, creating trust towards automated diagnostic systems.
32
Manepalli, Sailaja, Jobin Varghese, Akku Madslhusdhan, Gandhikota Umamahesh e Kiran Kumar Reddy Penubaka. "AI and ML in Biomedical Research: Unlocking Precision Medicine and Accelerating Discoveries". Journal of Neonatal Surgery 14, n. 11S (3 aprile 2025): 43–56. https://doi.org/10.52783/jns.v14.2940.
Testo completoAbstract (sommario):
Artificial intelligence and machine learning integration in biomedical research has tremendously benefitted precision medicine, disease diagnosis, and drug discovery. On the basis of these four advanced algorithms, this study investigates how AI-driven methodologies can be used for analysis in medical imaging, processing of genomic data and the prediction of drug response. Results from the experimental results show that traditional methods fail with a diagnostic accuracy of 82.7 % while Deep Learning-based medical imaging models attain a diagnostic accuracy of 97.3% outperforming the traditional methods by 15%. AI based genomic data mining had helped improve the mutation detection rate by 18%, which improved precision medicine approaches. Predictive models in cancer immunotherapy also increased treatment success rates by 22% in AI’s study. In addition, applying reinforcement learning in drug discovery led to compound screening efficiency of 40% improvement and reduced total drug development time. This underscores AI’s ability to increase diagnostic precision, improve treatment strategies and improve biomedical research efficiency. Meanwhile, much more attention will be needed for challenges so as cloud providers will need to meet requirements for data privacy, model interpretability as well as regulatory compliance. The future research should pursue the enhancement of AI explainability, the integration of multi-modal biomedical data, and the improvement of AI driven personalized treatment recommendations. Therefore, this study can contribute to the advancement of AI driven healthcare innovations and help create more accurate and accessible and personalized medical solutions.
33
Meher, Dinesh, Mrinal Gogoi, Pankaj Bharali, Prajna Anirvan e Shivaram Prasad Singh. "Artificial Intelligence in Small Bowel Endoscopy: Current Perspectives and Future Directions". Journal of Digestive Endoscopy 11, n. 04 (8 ottobre 2020): 245–52. http://dx.doi.org/10.1055/s-0040-1717824.
Testo completoAbstract (sommario):
AbstractArtificial intelligence (AI) is a computer system that is able to perform tasks which normally require human intelligence. The role of AI in the field of gastroenterology has been gradually evolving since its inception in the 1950s. Discovery of wireless capsule endoscopy (WCE) and balloon enteroscopy (BE) has revolutionized small gut imaging. While WCE is a relatively patient-friendly and noninvasive mode to examine the nonobstructed small gut, it is limited by a lengthy examination time and the need for expertise in reading images acquired by the capsule. Similarly, BE, despite having the advantage of therapeutic intervention, is costly, invasive, and requires general sedation. Incorporation of concepts like machine learning and deep learning has been used to handle large amounts of data and images in gastroenterology. Interestingly, in small gut imaging, the application of AI has been limited to WCE only. This review was planned to examine and summarize available published data on various AI-based approaches applied to small bowel disease. We conducted an extensive literature search using Google search engine, Google Scholar, and PubMed database for published literature in English on the application of different AI techniques in small bowel endoscopy, and have summarized the outcome and benefits of these applications of AI in small bowel endoscopy. Incorporation of AI in WCE has resulted in significant advancements in the detection of various lesions starting from dysplastic mucosa, inflammatory and nonmalignant lesions to the detection of bleeding with increasing accuracy and has shortened the lengthy review time in image analysis. As most of the studies to evaluate AI are retrospective, the presence of inherent selection bias cannot be excluded. Besides, the interpretability (black-box nature) of AI models remains a cause for concern. Finally, issues related to medical ethics and AI need to be judiciously addressed to enable its seamless use in future.
34
Obinna Nweke e Felix Adebayo Bakare. "Automated evaluation systems utilizing data science for enhanced accuracy, transparency, and decision optimization". World Journal of Advanced Research and Reviews 25, n. 2 (28 febbraio 2025): 2606–25. https://doi.org/10.30574/wjarr.2025.25.2.0667.
Testo completoAbstract (sommario):
Automated evaluation systems have emerged as a transformative approach in various industries, leveraging data science, machine learning, and artificial intelligence to enhance accuracy, transparency, and decision optimization. These systems are extensively utilized in domains such as finance, education, healthcare, and human resource management, where objective assessments and real-time data analysis are critical for decision-making. By integrating advanced analytics, statistical modeling, and natural language processing (NLP), these systems can process large volumes of structured and unstructured data, minimizing human bias and errors. In the financial sector, automated evaluation models leverage predictive analytics and anomaly detection algorithms to assess creditworthiness, fraud risks, and investment performance, ensuring data-driven decision-making. Similarly, in education and recruitment, AI-powered grading and skill assessment platforms optimize the evaluation process by identifying knowledge gaps and predicting candidate success. The healthcare sector benefits from AI-driven diagnostic tools that analyze patient data, improving disease detection rates and treatment recommendations. A key challenge in automated evaluation systems is ensuring fairness, explainability, and compliance with regulatory standards. Bias in training datasets and model interpretability issues often raise concerns about ethical AI deployment. Recent advancements in explainable AI (XAI) and fairness-aware machine learning algorithms have significantly improved transparency, allowing stakeholders to audit, interpret, and validate evaluation results with greater confidence. This paper explores the evolving landscape of automated evaluation systems, emphasizing the role of big data, deep learning, and decision optimization frameworks in refining predictive accuracy and operational efficiency. Furthermore, it highlights best practices and future directions for enhancing accountability, ethical compliance, and adaptive learning models within automated decision-making infrastructures.
35
Nafisat Temilade Popoola e Felix Adebayo Bakare. "Advanced computational forecasting techniques to strengthen risk prediction, pattern recognition, and compliance strategies". International Journal of Science and Research Archive 12, n. 2 (30 agosto 2024): 3033–54. https://doi.org/10.30574/ijsra.2024.12.2.1412.
Testo completoAbstract (sommario):
In an era defined by data-driven decision-making, advanced computational forecasting techniques have emerged as powerful tools for strengthening risk prediction, pattern recognition, and compliance strategies. These techniques leverage artificial intelligence (AI), machine learning (ML), and big data analytics to enhance accuracy, efficiency, and reliability in risk assessment across diverse industries. Traditional risk prediction models often rely on historical data and statistical methods, which, while effective, struggle to capture complex, non-linear patterns in evolving datasets. Advanced computational techniques, such as deep learning, ensemble learning, and reinforcement learning, have significantly improved predictive capabilities by identifying intricate correlations and anomalies in vast datasets. Pattern recognition plays a crucial role in cybersecurity, fraud detection, and financial risk management, where real-time anomaly detection enables organizations to preemptively mitigate threats. Predictive analytics models integrated with neural networks and natural language processing (NLP) have further revolutionized compliance strategies, ensuring adherence to regulatory frameworks and minimizing operational risks. In financial institutions, computational forecasting optimizes credit risk assessment and anti-money laundering (AML) monitoring, while in healthcare, it enhances disease outbreak predictions and patient care strategies. Despite these advancements, challenges such as algorithmic biases, data privacy concerns, and interpretability issues remain. Regulatory bodies are increasingly scrutinizing AI-driven decision systems to ensure transparency, fairness, and accountability. This study provides a comprehensive analysis of the latest computational forecasting techniques, their applications in risk management, and the evolving regulatory landscape. By addressing existing challenges and optimizing these techniques, industries can leverage AI-driven forecasting to enhance resilience, mitigate risks, and maintain regulatory compliance in an increasingly complex digital ecosystem.
36
Tonti, Emanuele, Sofia Tonti, Flavia Mancini, Chiara Bonini, Leopoldo Spadea, Fabiana D’Esposito, Caterina Gagliano, Mutali Musa e Marco Zeppieri. "Artificial Intelligence and Advanced Technology in Glaucoma: A Review". Journal of Personalized Medicine 14, n. 10 (16 ottobre 2024): 1062. http://dx.doi.org/10.3390/jpm14101062.
Testo completoAbstract (sommario):
Background: Glaucoma is a leading cause of irreversible blindness worldwide, necessitating precise management strategies tailored to individual patient characteristics. Artificial intelligence (AI) holds promise in revolutionizing the approach to glaucoma care by providing personalized interventions. Aim: This review explores the current landscape of AI applications in the personalized management of glaucoma patients, highlighting advancements, challenges, and future directions. Methods: A systematic search of electronic databases, including PubMed, Scopus, and Web of Science, was conducted to identify relevant studies published up to 2024. Studies exploring the use of AI techniques in personalized management strategies for glaucoma patients were included. Results: The review identified diverse AI applications in glaucoma management, ranging from early detection and diagnosis to treatment optimization and prognosis prediction. Machine learning algorithms, particularly deep learning models, demonstrated high accuracy in diagnosing glaucoma from various imaging modalities such as optical coherence tomography (OCT) and visual field tests. AI-driven risk stratification tools facilitated personalized treatment decisions by integrating patient-specific data with predictive analytics, enhancing therapeutic outcomes while minimizing adverse effects. Moreover, AI-based teleophthalmology platforms enabled remote monitoring and timely intervention, improving patient access to specialized care. Conclusions: Integrating AI technologies in the personalized management of glaucoma patients holds immense potential for optimizing clinical decision-making, enhancing treatment efficacy, and mitigating disease progression. However, challenges such as data heterogeneity, model interpretability, and regulatory concerns warrant further investigation. Future research should focus on refining AI algorithms, validating their clinical utility through large-scale prospective studies, and ensuring seamless integration into routine clinical practice to realize the full benefits of personalized glaucoma care.
37
Singh, Neelu, Swagatika Lenka e Akansha Sharma. "Healthcare Prediction based on ML and Convolutional Neural Network". Journal of Neonatal Surgery 14, n. 6S (17 marzo 2025): 440–53. https://doi.org/10.52783/jns.v14.2252.
Testo completoAbstract (sommario):
The integration of machine learning (ML) and deep learning techniques, particularly Convolutional Neural Networks (CNNs), has significantly transformed healthcare by enhancing predictive capabilities in disease diagnosis, medical imaging, and personalized treatment. This paper explores the application of ML and CNN-based models in healthcare prediction, focusing on their ability to analyze complex medical data, detect patterns, and improve early diagnosis. CNNs, renowned for their efficacy in image recognition, play a pivotal role in medical imaging tasks, such as tumor detection, diabetic retinopathy classification, and organ segmentation. Additionally, ML algorithms, including decision trees, support vector machines, and deep neural networks, complement CNNs by processing non-image-based medical data, aiding in patient risk assessment and prognosis prediction. Despite their promising contributions, ML and CNN-based healthcare models face challenges, including data scarcity, class imbalance, model interpretability, and ethical concerns regarding patient privacy. Addressing these issues through robust data augmentation techniques, explainable AI models, and federated learning can enhance the reliability and applicability of predictive healthcare solutions. Furthermore, integrating electronic health records (EHRs), genomic data, and wearable sensor information with ML models can pave the way for more personalized and data-driven healthcare systems. This paper provides a comprehensive analysis of recent advancements in ML and CNN-based healthcare prediction models, discussing their strengths, limitations, and future research directions. By leveraging AI-driven techniques, healthcare professionals can achieve improved diagnostic accuracy, reduced human error, and enhanced patient outcomes, ultimately advancing the field of precision medicine.
38
Idroes, Ghazi Mauer, Teuku Rizky Noviandy, Talha Bin Emran e Rinaldi Idroes. "Explainable Deep Learning Approach for Mpox Skin Lesion Detection with Grad-CAM". Heca Journal of Applied Sciences 2, n. 2 (19 settembre 2024): 54–63. http://dx.doi.org/10.60084/hjas.v2i2.216.
Testo completoAbstract (sommario):
Mpox is a viral zoonotic disease that presents with skin lesions similar to other conditions like chickenpox, measles, and hand-foot-mouth disease, making accurate diagnosis challenging. Early and precise detection of mpox is critical for effective treatment and outbreak control, particularly in resource-limited settings where traditional diagnostic methods are often unavailable. While deep learning models have been applied successfully in medical imaging, their use in mpox detection remains underexplored. To address this gap, we developed a deep learning-based approach using the ResNet50v2 model to classify mpox lesions alongside five other skin conditions. We also incorporated Grad-CAM (Gradient-weighted Class Activation Mapping) to enhance model interpretability. The results show that the ResNet50v2 model achieved an accuracy of 99.33%, precision of 99.34%, sensitivity of 99.33%, and an F1-score of 99.32% on a dataset of 1,594 images. Grad-CAM visualizations confirmed that the model focused on relevant lesion areas for its predictions. While the model performed exceptionally well overall, it struggled with misclassifications between visually similar diseases, such as chickenpox and mpox. These results demonstrate that AI-based diagnostic tools can provide reliable, interpretable support for clinicians, particularly in settings with limited access to specialized diagnostics. However, future work should focus on expanding datasets and improving the model's capacity to distinguish between similar conditions.
39
DURGA SAI SIVA VARA PRASAD RAJU, NIDADAVOLU VENKAT, e PENMETSA NAVEENA DEVI. "AI-Assisted Medical Imaging and Heart Disease Diagnosis: A Deep Learning Approach for Automated Analysis and Enhanced Prediction Using Ensemble Classifiers". Journal of Artificial Intelligence General science (JAIGS) ISSN:3006-4023 6, n. 1 (24 ottobre 2024): 210–29. http://dx.doi.org/10.60087/jaigs.v6i1.242.
Testo completoAbstract (sommario):
In the medical field, early detection of cardiovascular problems is a challenging task. This research aims to improve the accuracy of heart disease prediction through the application of machine learning techniques. Cardiovascular diseases (CVDs), including coronary artery disease, stroke, and peripheral artery disease, are the leading cause of mortality worldwide. Early identification of individuals at high risk of developing CVDs is crucial for preventing adverse cardiovascular events through medical interventions and lifestyle modifications. Machine learning (ML) offers innovative techniques to build predictive models that can accurately estimate CVD risk based on patient data. This review provides a comprehensive overview of recent research on applying ML algorithms for CVD risk assessment. The paper begins with background on CVD epidemiology and risk factors, followed by sections on ML methodology, feature selection techniques, model evaluation metrics, public CVD datasets, and ethical considerations. The main focus is a critical analysis of over 50 studies from 2015-2022 that developed ML models for predicting various CVD outcomes. The performance of classical ML algorithms like logistic regression and random forest is compared with deep learning methods like convolution and recurrent neural networks across diverse patient cohorts. Challenges and limitations around model interpretability, data quality, feature engineering, and external validation are discussed. Overall, the review demonstrates that ML has strong potential to enhance individualized CVD risk estimation and enable personalized preventive care, although more methodological refinement and clinical validation are warranted before full-scale clinical implementation.
40
Rahmathullah, Rahmathullah, S. Nagakishore Bhavanam e Vasujadevi Midasala. "Deep Learning-Powered Cardiovascular Disease Prediction: A Novel Approach to Early Diagnosis and Risk Assessment". Journal of Neonatal Surgery 14, n. 4 (21 marzo 2025): 21–31. https://doi.org/10.52783/jns.v14.2421.
Testo completoAbstract (sommario):
Cardiovascular disease (CVD) continues to be a leading cause of death and disability worldwide, underscoring the critical need for improved risk prediction and early diagnosis. Traditional risk models, such as the Framingham Risk Score, provide valuable insights but are limited in their ability to incorporate the diverse, multi-dimensional data necessary for personalized healthcare. In response to this challenge, we propose a novel deep learning-based framework that integrates clinical, genetic, and imaging data to enhance CVD prediction and risk stratification. The proposed model utilizes Convolutional Neural Networks (CNNs) for analyzing cardiovascular imaging and Recurrent Neural Networks (RNNs)/Long Short-Term Memory (LSTM) for processing sequential data from electronic health records (EHRs). By employing attention mechanisms, the model effectively combines these diverse data types to provide a more comprehensive evaluation of risk factors. The model was trained on large-scale datasets, including MIMIC-III and UK Biobank, and transfer learning techniques were applied to improve generalizability across various patient populations. Additionally, we incorporate Explainable AI (XAI) tools, such as SHAP and Grad-CAM, to facilitate clinical interpretability, enabling healthcare professionals to understand and trust the model’s predictions. Experimental results demonstrate that our deep learning framework significantly outperforms traditional machine learning models, achieving higher accuracy, sensitivity, and specificity in predicting the onset of CVD. Furthermore, the model shows robust generalizability across diverse demographic groups and offers real-time monitoring potential through integration with wearable devices. To ensure data privacy, we introduce federated learning, allowing the model to train across multiple institutions without sharing sensitive patient data. This study represents a significant advancement in the field of AI-driven precision cardiology, providing a scalable solution for early detection, personalized treatment, and clinical decision support. Future work will focus on refining model generalization, incorporating real-time data from wearables, and addressing regulatory and ethical considerations to promote widespread adoption.
41
Galić, Irena, Marija Habijan, Hrvoje Leventić e Krešimir Romić. "Machine Learning Empowering Personalized Medicine: A Comprehensive Review of Medical Image Analysis Methods". Electronics 12, n. 21 (25 ottobre 2023): 4411. http://dx.doi.org/10.3390/electronics12214411.
Testo completoAbstract (sommario):
Artificial intelligence (AI) advancements, especially deep learning, have significantly improved medical image processing and analysis in various tasks such as disease detection, classification, and anatomical structure segmentation. This work overviews fundamental concepts, state-of-the-art models, and publicly available datasets in the field of medical imaging. First, we introduce the types of learning problems commonly employed in medical image processing and then proceed to present an overview of commonly used deep learning methods, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and generative adversarial networks (GANs), with a focus on the image analysis task they are solving, including image classification, object detection/localization, segmentation, generation, and registration. Further, we highlight studies conducted in various application areas, encompassing neurology, brain imaging, retinal analysis, pulmonary imaging, digital pathology, breast imaging, cardiac imaging, bone analysis, abdominal imaging, and musculoskeletal imaging. The strengths and limitations of each method are carefully examined, and the paper identifies pertinent challenges that still require attention, such as the limited availability of annotated data, variability in medical images, and the interpretability issues. Finally, we discuss future research directions with a particular focus on developing explainable deep learning methods and integrating multi-modal data.
42
Cai, Jingxun, Zne-Jung Lee, Zhihxian Lin e Ming-Ren Yang. "A Novel SHAP-GAN Network for Interpretable Ovarian Cancer Diagnosis". Mathematics 13, n. 5 (6 marzo 2025): 882. https://doi.org/10.3390/math13050882.
Testo completoAbstract (sommario):
Ovarian cancer stands out as one of the most formidable adversaries in women’s health, largely due to its typically subtle and nonspecific early symptoms, which pose significant challenges to early detection and diagnosis. Although existing diagnostic methods, such as biomarker testing and imaging, can help with early diagnosis to some extent, these methods still have limitations in sensitivity and accuracy, often leading to misdiagnosis or missed diagnosis. Ovarian cancer’s high heterogeneity and complexity increase diagnostic challenges, especially in disease progression prediction and patient classification. Machine learning (ML) has outperformed traditional methods in cancer detection by processing large datasets to identify patterns missed by conventional techniques. However, existing AI models still struggle with accuracy in handling imbalanced and high-dimensional data, and their “black-box” nature limits clinical interpretability. To address these issues, this study proposes SHAP-GAN, an innovative diagnostic model for ovarian cancer that integrates Shapley Additive exPlanations (SHAP) with Generative Adversarial Networks (GANs). The SHAP module quantifies each biomarker’s contribution to the diagnosis, while the GAN component optimizes medical data generation. This approach tackles three key challenges in medical diagnosis: data scarcity, model interpretability, and diagnostic accuracy. Results show that SHAP-GAN outperforms traditional methods in sensitivity, accuracy, and interpretability, particularly with high-dimensional and imbalanced ovarian cancer datasets. The top three influential features identified are PRR11, CIAO1, and SMPD3, which exhibit wide SHAP value distributions, highlighting their significant impact on model predictions. The SHAP-GAN network has demonstrated an impressive accuracy rate of 99.34% on the ovarian cancer dataset, significantly outperforming baseline algorithms, including Support Vector Machines (SVM), Logistic Regression (LR), and XGBoost. Specifically, SVM achieved an accuracy of 72.78%, LR achieved 86.09%, and XGBoost achieved 96.69%. These results highlight the superior performance of SHAP-GAN in handling high-dimensional and imbalanced datasets. Furthermore, SHAP-GAN significantly alleviates the challenges associated with intricate genetic data analysis, empowering medical professionals to tailor personalized treatment strategies for individual patients.
43
Mahmoud, Akeel Shaker, Olfa Lamouchi e Safya Belghith. "Advancements in Machine Learning and Deep Learning for Early Diagnosis of Chronic Kidney Diseases: A Comprehensive Review". Babylonian Journal of Machine Learning 2024 (17 settembre 2024): 149–56. http://dx.doi.org/10.58496/bjml/2024/015.
Testo completoAbstract (sommario):
Chronic kidney disease (CKD) is a prevalent and debilitating condition worldwide, characterized by progressive loss of kidney function over time. Early detection plays a crucial role in mitigating its impact on patient health and healthcare systems. In recent years, there has been a burgeoning interest in leveraging machine learning (ML) and deep learning (DL) techniques to enhance the early diagnosis of CKD. This comprehensive review explores the advancements in ML and DL models applied to CKD diagnosis, focusing on their ability to integrate diverse data sources including clinical biomarkers, imaging modalities, and patient demographics. Key ML algorithms such as Support Vector Machines (SVM), Random Forests (RF), and neural network architectures like Convolutional Neural Networks (CNNs) and Long Short-Term Memory networks (LSTMs) are examined in the context of their performance in predicting CKD progression, classifying disease stages, and identifying at-risk populations. Furthermore, the review discusses challenges such as data quality, model interpretability, and integration into clinical practice, alongside emerging trends in explainable AI, transfer learning, federated learning, and integration with electronic health records (EHRs). By synthesizing findings from recent literature, this paper aims to provide insights into current methodologies, identify gaps for future research, and underscore the transformative potential of ML and DL in revolutionizing early CKD diagnosis and management..
44
Trofimov, Yuriy V., Aleksey N. Averkin e Eugenia N. Cheremisina. "Review and Analysis of XAI Methods for Addressing Geoecological Zoning and Public Health Prevention Challenges". Geoinformatika, n. 4 (16 dicembre 2024): 93–118. https://doi.org/10.47148/1609-364x-2024-4-93-118.
Testo completoAbstract (sommario):
This study focuses on the application of Explainable Artificial Intelligence (XAI) in geoecological zoning tasks to support sustainable development and public health prevention. Amid increasing anthropogenic pressures on ecosystems and rising disease rates due to environmental degradation, emphasis is placed on early risk detection methods. The study highlights the importance of XAI in analyzing ecological data to mitigate the impacts of adverse factors on population health. Integrating XAI with Geographic Information Systems (GIS) not only provides high accuracy in geoecological forecasting but also enhances the transparency of these forecasts for experts, aiding informed decision-making in the fields of geoecology and healthcare. Special attention is given to early diagnosis of health risks, such as respiratory and oncological diseases, through the use of XAI in analyzing environmental data and medical images. Explainable AI enhances the transparency and understandability of diagnostic processes for medical professionals, fostering trust in analytical outcomes. Implementing XAI in healthcare can not only improve diagnostic accuracy but also optimize healthcare resources by reallocating them toward disease prevention. A review of existing decision support systems (DSS) demonstrates the efficacy of hybrid models combining neural networks and fuzzy logic to enhance the precision and interpretability of medical forecasts. These models open new prospects for personalized medicine, improving preventive strategies and providing individual recommendations based on comprehensive analyses of environmental and medical data. A critical aspect of the study is territorial zoning aimed at managing environmental risks and disease prevention. This approach not only reduces the burden on healthcare systems but also promotes sustainable territorial development, taking into account the influence of environmental and social factors on population health.
45
Shujaat, Sohaib. "Automated Machine Learning in Dentistry: A Narrative Review of Applications, Challenges, and Future Directions". Diagnostics 15, n. 3 (24 gennaio 2025): 273. https://doi.org/10.3390/diagnostics15030273.
Testo completoAbstract (sommario):
The adoption of automated machine learning (AutoML) in dentistry is transforming clinical practices by enabling clinicians to harness machine learning (ML) models without requiring extensive technical expertise. This narrative review aims to explore the impact of autoML in dental applications. A comprehensive search of PubMed, Scopus, and Google Scholar was conducted without time and language restrictions. Inclusion criteria focused on studies evaluating autoML applications and performance for dental tasks. Exclusion criteria included non-dental studies, single-case reports, and conference abstracts. This review highlights multiple promising applications of autoML in dentistry. Diagnostic tasks showed high accuracy, such as 95.4% precision in dental implant classification and 92% accuracy in paranasal sinus disease detection. Predictive tasks also demonstrated promise, including 84% accuracy for ICU admissions due to dental infections and 93.9% accuracy in orthodontic extraction predictions. AutoML frameworks like Google Vertex AI and H2O AutoML emerged as key tools for these applications. AutoML shows great promise in transforming dentistry by facilitating data-driven decision-making and improving patient care quality through accessible, automated solutions. Future advancements should focus on enhancing model interpretability, developing large and annotated datasets, and creating pipelines tailored to dental tasks. Educating clinicians on autoML and integrating domain-specific knowledge into automated platforms could further bridge the gap between complex ML technology and practical dental applications.
46
Luz, Ayuns, e Elizabeth Jerry. "Role of Image Segmentation and Deep Learning in Medical Imaging". International Journal of Advances in Engineering and Management 06, n. 12 (dicembre 2024): 125–35. https://doi.org/10.35629/5252-0612125135.
Testo completoAbstract (sommario):
The rapid advancements in medical imaging technologies have significantly enhanced diagnostic accuracy and clinical decision-making in modern healthcare. Image segmentation and deep learning have emerged as transformative tools among these advancements. This article explores the pivotal role of image segmentation and deep learning in medical imaging, detailing their methodologies, applications, challenges, and future directions. Deep learning, particularly Convolutional Neural Networks (CNNs), has revolutionized medical imaging by automating the analysis of complex datasets and improving diagnostic precision. Image segmentation, a fundamental component of medical imaging, allows for delineating specific structures such as organs, tissues, and pathological regions. Together, these technologies have been applied in diverse fields, including oncology, cardiology, neurology, and ophthalmology, enabling applications such as tumor detection, organ segmentation, disease progression monitoring, and treatment planning. However, despite its transformative potential, the integration of deep learning into medical imaging faces several challenges. These include data scarcity, privacy concerns, interpretability issues, and regulatory hurdles. The article discusses various strategies to address these challenges, such as data augmentation, transfer learning, and the development of explainable AI models to ensure transparency and trustworthiness. Evaluation metrics, such as accuracy, sensitivity, specificity, and Dice Similarity Coefficient (DSC), are essential for assessing model performance. Rigorous clinical validation and regulatory approval are crucial to integrating deep learning systems into clinical workflows effectively. Looking ahead, the future of deep learning in medical imaging holds immense promise. Innovations like multimodal imaging, personalized medicine, and AI-driven automation are set to further revolutionize the field, enhancing the efficiency and accuracy of diagnostics. Collaborative efforts between clinicians, researchers, and AI developers will play a vital role in overcoming current limitations and driving progress. This article concludes by emphasizing the transformative potential of deep learning and image segmentation in medical imaging, highlighting their ability to improve diagnostic accuracy, streamline clinical workflows, and ultimately, enhance patient care. By addressing current challenges and continuing to innovate, these technologies are poised to redefine the landscape of medical diagnostics and treatment in the years to come.
47
Gholi Zadeh Kharrat, Fatemeh, Christian Gagne, Alain Lesage, Geneviève Gariépy, Jean-François Pelletier, Camille Brousseau-Paradis, Louis Rochette et al. "Explainable artificial intelligence models for predicting risk of suicide using health administrative data in Quebec". PLOS ONE 19, n. 4 (3 aprile 2024): e0301117. http://dx.doi.org/10.1371/journal.pone.0301117.
Testo completoAbstract (sommario):
Suicide is a complex, multidimensional event, and a significant challenge for prevention globally. Artificial intelligence (AI) and machine learning (ML) have emerged to harness large-scale datasets to enhance risk detection. In order to trust and act upon the predictions made with ML, more intuitive user interfaces must be validated. Thus, Interpretable AI is one of the crucial directions which could allow policy and decision makers to make reasonable and data-driven decisions that can ultimately lead to better mental health services planning and suicide prevention. This research aimed to develop sex-specific ML models for predicting the population risk of suicide and to interpret the models. Data were from the Quebec Integrated Chronic Disease Surveillance System (QICDSS), covering up to 98% of the population in the province of Quebec and containing data for over 20,000 suicides between 2002 and 2019. We employed a case-control study design. Individuals were considered cases if they were aged 15+ and had died from suicide between January 1st, 2002, and December 31st, 2019 (n = 18339). Controls were a random sample of 1% of the Quebec population aged 15+ of each year, who were alive on December 31st of each year, from 2002 to 2019 (n = 1,307,370). We included 103 features, including individual, programmatic, systemic, and community factors, measured up to five years prior to the suicide events. We trained and then validated the sex-specific predictive risk model using supervised ML algorithms, including Logistic Regression (LR), Random Forest (RF), Extreme Gradient Boosting (XGBoost) and Multilayer perceptron (MLP). We computed operating characteristics, including sensitivity, specificity, and Positive Predictive Value (PPV). We then generated receiver operating characteristic (ROC) curves to predict suicides and calibration measures. For interpretability, Shapley Additive Explanations (SHAP) was used with the global explanation to determine how much the input features contribute to the models’ output and the largest absolute coefficients. The best sensitivity was 0.38 with logistic regression for males and 0.47 with MLP for females; the XGBoost Classifier with 0.25 for males and 0.19 for females had the best precision (PPV). This study demonstrated the useful potential of explainable AI models as tools for decision-making and population-level suicide prevention actions. The ML models included individual, programmatic, systemic, and community levels variables available routinely to decision makers and planners in a public managed care system. Caution shall be exercised in the interpretation of variables associated in a predictive model since they are not causal, and other designs are required to establish the value of individual treatments. The next steps are to produce an intuitive user interface for decision makers, planners and other stakeholders like clinicians or representatives of families and people with live experience of suicidal behaviors or death by suicide. For example, how variations in the quality of local area primary care programs for depression or substance use disorders or increased in regional mental health and addiction budgets would lower suicide rates.
48
Raza, Ali, Akhtar Ali, Sami Ullah, Yasir Nadeem Anjum e Basit Rehman. "Optimizing skin cancer screening with convolutional neural networks in smart healthcare systems". PLOS ONE 20, n. 3 (25 marzo 2025): e0317181. https://doi.org/10.1371/journal.pone.0317181.
Testo completoAbstract (sommario):
Skin cancer is among the most prevalent types of malignancy all over the global and is strongly associated with the patient’s prognosis and the accuracy of the initial diagnosis. Clinical examination of skin lesions is a key aspect that is important in the assessment of skin disease but comes with some drawbacks mainly with interpretational aspects, time-consuming and healthare expenditure. Skin cancer if detected early and treated in time can be controlled and its deadly impacts arrested completely. Algorithms applied in convolutional neural network (CNN) could lead to an enhanced speed of identifying and distinguishing a disease, which in turn leads to early detection and treatment. So as to eliminate these challenges, optimized CNN prediction models for cancer skin classification is studied in this researche. The objectives of this study were to develop reliable optimized CNN prediction models for skin cancer classification, to handle the severe class imbalance problem where skin cancer class was found to be much smaller than the healthy class. To evaluate model interpretability and to develop an end-to-end smart healthcare system using explainable AI (XAI) such as Grad-CAM and Grad-CAM++. In this researche new activation function namely NGNDG-AF was offered specifically to enhance the capabilities of network fitting and generalization ability, convergence rate and reduction in mathematical computational cost. A research used an optimized CNN and ResNet152V2 with the HAM10000 dataset to differentiate between the seven forms of skin cancer. Model training involved the use of two optimization functions (RMSprop and Adam) and NGNDG-AF activation functions. Cross validation technique the holdout validation is used to estimate of the model’s generalization performance for unseed data. Optimized CNN is performing well as compare to ResNet152V2 for unseen data. The efficacy of the optimized CNN method with NGNDG-AF was examined by a comparative study wirh popular CNN with various activation functions shows that better performance of NGNDG-AF, achieving the classification accuracy rates that are as high as 99% in training and 98% in the validation. The recommended system also involves the integration of the smart healthcare application as a central component to give the doctors as well as the healthcare providers diagnosing and tools that would assist in the early detection of skin cancer hence leading to better outcomes of the treatment.
49
Mehta, Varshil. "Artificial Intelligence in Medicine: Revolutionizing Healthcare for Improved Patient Outcomes". Journal of Medical Research and Innovation 7, n. 2 (3 giugno 2023): e000292. http://dx.doi.org/10.32892/jmri.292.
Testo completoAbstract (sommario):
Introduction: Artificial intelligence (AI) has emerged as a groundbreaking technology with the potential to transform various sectors, and the field of medicine is no exception. With its ability to process vast amounts of data and perform complex tasks, AI has begun to revolutionize healthcare, offering promising avenues for diagnosis, treatment, and patient care. In this editorial article, we will explore the significant impact of AI in medicine, highlighting its potential benefits and the challenges that lie ahead.
AI-Driven Diagnosis
One of the most remarkable applications of AI in medicine is its capacity to assist in accurate and efficient diagnosis. By leveraging machine learning algorithms, AI systems can analyze medical imaging, such as X-rays, MRIs, and CT scans, with a level of precision that rivals human experts. Studies have demonstrated the effectiveness of AI in detecting various conditions, including lung cancer, cardiovascular diseases, and neurological disorders, leading to earlier and more accurate diagnoses.
For instance, a study published in Nature Medicine by McKinney et al. revealed that an AI model trained on a large dataset of mammograms outperformed radiologists in breast cancer detection. The AI system achieved a lower false-negative rate and reduced the number of false positives, thereby potentially reducing unnecessary biopsies [1]. Similarly, a study by Esteva et al., showed that a deep learning algorithm outperformed dermatologists in diagnosing skin cancer based on images [2]. Such advancements in AI-driven diagnosis hold immense promise for improving patient outcomes and reducing healthcare costs.
Personalized Treatment and Precision Medicine
AI has also opened doors to personalized treatment strategies, enabling healthcare professionals to tailor therapies to individual patients. By analyzing vast amounts of patient data, including genetic information, medical history, and treatment outcomes, AI algorithms can identify patterns, predict responses to specific treatments, and recommend personalized interventions. This approach, known as precision medicine, has the potential to revolutionize disease management.
An example of AI's impact on precision medicine is showcased in the work of Poplin et al. The study demonstrated how a deep learning algorithm could predict the onset of cardiovascular events by analyzing electronic health records. The algorithm outperformed traditional risk models by incorporating a broader range of patient data, allowing for more accurate and timely interventions to prevent adverse events [3]. Similarly, Obermeyer et al., demonstrated that an AI model outperformed traditional methods in predicting acute kidney injury in hospitalized patients [4] while a study by Che et al., demonstrated the effectiveness of an AI model in predicting sepsis, allowing for early intervention and improved patient outcomes [5].
Enhanced Clinical Decision-Making and Workflow
AI has the capacity to enhance clinical decision-making by assisting healthcare providers in analyzing complex data and generating evidence-based recommendations. AI systems can process and interpret vast amounts of medical literature, patient records, and clinical guidelines, providing healthcare professionals with timely insights and decision support. This augmentation of human expertise can lead to more accurate diagnoses, improved treatment plans, and enhanced patient care.
A notable example is the work of Rajkomar et al., published in The New England Journal of Medicine. The authors developed an AI algorithm capable of predicting patient deterioration within the next few hours, based on electronic health record data. By alerting healthcare providers in advance, this AI system helped to prevent adverse events and facilitated proactive interventions [6].
Drug Discovery and Clinical Research
The drug discovery and development process is notoriously expensive and time-consuming. AI has the potential to accelerate this process by analyzing vast amounts of biomedical literature, genomic data, and clinical trial outcomes. Machine learning models can identify potential drug targets, predict drug toxicity, and optimize drug formulations. In fact, a study by Aliper et al., demonstrated that an AI system outperformed human researchers in designing new drugs to target age-related diseases [7].
Virtual Assistants and Telemedicine
AI-powered virtual assistants and chatbots are transforming the way patients interact with healthcare providers. These virtual assistants can provide instant medical advice, answer queries, and triage patients based on their symptoms. Furthermore, telemedicine platforms integrated with AI algorithms can enhance remote patient monitoring, enabling healthcare professionals to monitor patients' vital signs and provide timely interventions [8,9].
Challenges and Ethical Considerations
While the potential benefits of AI in medicine are substantial, it is important to address the challenges and ethical considerations associated with its implementation. Privacy and data security remain critical concerns when handling vast amounts of patient data. Maintaining patient confidentiality and ensuring secure data sharing frameworks must be prioritized to protect patient privacy.
Moreover, the need for transparency and interpretability of AI algorithms is vital to build trust between healthcare professionals and AI systems. Understanding how AI arrives at its recommendations or diagnoses is crucial for healthcare providers to make informed decisions and ensure accountability.
Conclusion: Artificial intelligence holds tremendous potential to revolutionize healthcare and improve patient outcomes. From enhancing diagnostic accuracy to enabling personalized treatment strategies and augmenting clinical decision-making, AI is transforming the field of medicine. However, to fully realize the benefits, it is essential to address the challenges surrounding privacy, data security, and algorithm transparency. By leveraging the power of AI responsibly, healthcare providers can usher in a new era of precision medicine, advancing the quality and effectiveness of patient care.
50
Temitope Oluwatosin Fatunmbi. "Integrating quantum neural networks with machine learning algorithms for optimizing healthcare diagnostics and treatment outcomes". World Journal of Advanced Research and Reviews 17, n. 3 (30 marzo 2023): 1059–77. https://doi.org/10.30574/wjarr.2023.17.3.0306.
Testo completoAbstract (sommario):
The rapid advancements in artificial intelligence (AI) and quantum computing have catalyzed an unprecedented shift in the methodologies utilized for healthcare diagnostics and treatment optimization. This research paper explores the integration of quantum neural networks (QNNs) with classical machine learning (ML) algorithms to enhance diagnostic accuracy, facilitate personalized treatment plans, and predict patient outcomes with a higher degree of precision. Quantum neural networks, leveraging the principles of quantum mechanics such as superposition, entanglement, and quantum parallelism, have demonstrated the potential to perform complex computations more efficiently than classical counterparts. When coupled with established machine learning algorithms, QNNs can overcome traditional limitations in data processing, enabling more sophisticated models capable of uncovering intricate patterns in large and high-dimensional datasets. Machine learning, with its vast applications in the medical field, has long been instrumental in improving diagnostics and tailoring treatment regimens to patient-specific characteristics. However, despite significant advancements, classical ML approaches face substantial challenges, particularly in terms of computational complexity and the ability to process large-scale, multi-modal healthcare data effectively. Quantum neural networks address these challenges by introducing quantum computational paradigms that facilitate exponentially faster processing, allowing for real-time analysis of vast and complex datasets. The synergy between QNNs and ML algorithms introduces novel approaches that are poised to revolutionize predictive analytics in healthcare, optimizing patient outcomes and enabling highly personalized treatment plans. A key aspect of integrating quantum neural networks into machine learning frameworks is the potential for improved precision in diagnostic systems. Traditional diagnostic procedures often rely on predefined models that may overlook nuanced correlations within patient data. Quantum neural networks, with their ability to represent and process data in a quantum space, provide a more robust framework that can adaptively learn from intricate relationships in patient information. For instance, QNNs can significantly enhance the efficacy of disease detection algorithms, such as those used for identifying early-stage cancers or predicting the onset of chronic conditions like diabetes and heart disease, by offering superior pattern recognition capabilities. Furthermore, QNNs combined with classical machine learning architectures facilitate the creation of hybrid models that harness the strengths of both approaches, leading to diagnostic tools that are not only more precise but also more adaptive to varied data sources. The integration of QNNs with machine learning extends beyond diagnostics to personalized treatment optimization. Traditional treatment planning methodologies, including rule-based and data-driven ML models, often face difficulties in accounting for the multifaceted nature of patient data and individual variability. Quantum neural networks enhance this process by leveraging quantum algorithms that provide an efficient search space for complex treatment optimization problems, allowing for a more detailed understanding of patient responses and potential treatment outcomes. The ability of QNNs to perform parallel processing enables the assessment of a wide range of treatment scenarios simultaneously, leading to more accurate predictions regarding patient reactions to specific drugs, therapies, or medical interventions. This facilitates an adaptive approach that can recommend personalized treatment regimens based on comprehensive patient profiles, ultimately enhancing patient outcomes and reducing the likelihood of adverse drug reactions. In addition to enhancing diagnostics and treatment recommendations, quantum neural networks show promise in forecasting patient outcomes by offering a more robust analysis of longitudinal patient data. Forecasting models that leverage the combined power of quantum and classical algorithms can process historical data more rapidly, allowing healthcare providers to anticipate potential health issues and intervene earlier. For example, predictive models utilizing QNNs can anticipate patient deterioration in critical care settings, facilitating timely interventions that mitigate risks and improve survival rates. Such predictive models can be instrumental in managing chronic diseases, monitoring recovery trajectories, and optimizing resource allocation within healthcare systems, thus contributing to overall efficiency and better resource management. Despite the promising capabilities of integrating QNNs with ML algorithms, there are notable challenges that need to be addressed to fully realize their potential. The practical implementation of quantum algorithms in a healthcare context faces hurdles related to hardware limitations, the need for high fidelity in quantum states, and the scalability of quantum systems to handle real-world clinical data. Additionally, the hybrid nature of combining classical and quantum approaches requires sophisticated algorithms that can bridge the gap between quantum computation and classical data processing pipelines. Solutions to these challenges may include advancements in quantum hardware, such as the development of more stable qubits and noise reduction techniques, as well as the optimization of hybrid algorithms that leverage both classical machine learning and quantum computing capabilities effectively. The exploration of quantum neural networks for healthcare applications also necessitates rigorous ethical considerations, particularly in ensuring data privacy and security. The incorporation of quantum computing must comply with healthcare data protection regulations, and quantum algorithms must be designed to maintain patient confidentiality while processing sensitive health data. Moreover, the interpretability of quantum models poses challenges that could hinder their acceptance in clinical practice. Advances in explainable AI and quantum algorithm transparency are crucial to foster trust among healthcare professionals and patients alike. Integration of quantum neural networks with classical machine learning models represents a transformative approach that could significantly advance healthcare diagnostics, personalized treatment strategies, and patient outcome prediction. By harnessing the computational advantages offered by quantum systems and the flexibility of machine learning algorithms, healthcare applications can achieve a new level of precision and adaptability. Despite current challenges, continued research into quantum algorithms, quantum hardware development, and hybrid computational models promises substantial strides in overcoming these limitations. The synergy between QNNs and ML algorithms could ultimately lead to more effective, personalized, and efficient healthcare solutions, ushering in a new era of data-driven medical care characterized by increased diagnostic accuracy and improved treatment outcomes. As the field evolves, interdisciplinary collaboration between quantum physicists, computer scientists, and healthcare professionals will be vital to unlock the full potential of these innovative computational techniques and bring them to mainstream clinical use.