Interpretable Deep Learning for Pediatric Pneumonia Diagnosis Through Multi-Phase Feature Learning and Activation Patterns

Electronics, Journal Year: 2025, Volume and Issue: 14(9), P. 1899 - 1899

Published: May 7, 2025

Pediatric pneumonia remains a critical global health challenge requiring accurate and interpretable diagnostic solutions. Although deep learning has shown potential for recognition on chest X-ray images, gaps persist in understanding model interpretability feature during training. We evaluated four convolutional neural network (CNN) architectures, i.e., InceptionV3, InceptionResNetV2, DenseNet201, MobileNetV2, using three approaches—standard convolution, multi-scale strided convolution—all incorporating the Mish activation function. Among tested models, with convolutions, demonstrated best performance, achieving an accuracy of 0.9718. InceptionV3 also performed well same approach, 0.9684. For DenseNet201 convolution approach was more effective, accuracies 0.9676 0.9437, respectively. Gradient-weighted class mapping (Grad-CAM) visualizations provided insights, e.g., convolutions identified diffuse viral patterns across wider lung regions, while precisely highlighted localized bacterial consolidations, aligning radiologists’ priorities. These findings establish following architectural guidelines: are suited to hierarchical CNNs, approaches optimize compact models. This research significantly advances development interpretable, high-performance systems pediatric X-rays, bridging gap between computational innovation clinical application.

Language: Английский

---

2D-CNN Architecture for Accurate Classification of COVID-19 Related Pneumonia on X-Ray Images

International Journal of Advanced Computer Science and Applications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Jan. 1, 2024

In the wake of COVID-19 pandemic, use medical imaging, particularly X-ray radiography, has become integral to rapid and accurate diagnosis pneumonia induced by virus. This research paper introduces a novel two-dimensional Convolutional Neural Network (2D-CNN) architecture specifically tailored for classification related in images. Leveraging advancements deep learning, our model is designed distinguish between viral pneumonia, typical COVID-19, other types as well healthy lung imagery. The proposed 2D-CNN characterized its depth unique layer arrangement, which optimizes feature extraction from images, thus enhancing model's diagnostic precision. We trained using substantial dataset comprising thousands annotated including those patients diagnosed with types, individuals no infection. enabled learn wide range radiographic features associated different conditions. Our demonstrated exceptional performance, achieving high accuracy, sensitivity, specificity preliminary tests. results indicate that not only outperforms existing models but also provides valuable tool healthcare professionals early detection differentiation pneumonia. capability crucial prompt appropriate treatment, potentially reducing pandemic's burden on systems. Furthermore, design allows easy integration into imaging workflows, offering practical efficient solution frontline facilities. contributes ongoing efforts combat procedures through application artificial intelligence imaging.

Language: Английский

---

Transparent and Accurate COVID-19 Diagnosis: Integrating Explainable AI with Advanced Deep Learning in CT Imaging

Computer Modeling in Engineering & Sciences, Journal Year: 2024, Volume and Issue: 139(3), P. 3101 - 3123

Published: Jan. 1, 2024

In the current landscape of COVID-19 pandemic, utilization deep learning in medical imaging, especially chest computed tomography (CT) scan analysis for virus detection, has become increasingly significant.Despite its potential, learning's "black box" nature been a major impediment to broader acceptance clinical environments, where transparency decision-making is imperative.To bridge this gap, our research integrates Explainable AI (XAI) techniques, specifically Local Interpretable Model-Agnostic Explanations (LIME) method, with advanced models.This integration forms sophisticated and transparent framework identification, enhancing capability standard Convolutional Neural Network (CNN) models through transfer data augmentation.Our approach leverages refined DenseNet201 architecture superior feature extraction employs augmentation strategies foster robust model generalization.The pivotal element methodology use LIME, which demystifies process, providing clinicians clear, interpretable insights into AI's reasoning.This unique combination an optimized Deep (DNN) LIME not only elevates precision detecting cases but also equips healthcare professionals deeper understanding diagnostic process.Our validated on SARS-COV-2 CT-Scan dataset, demonstrates exceptional accuracy, performance metrics that reinforce potential seamless modern systems.This innovative marks significant advancement creating explainable trustworthy tools decisionmaking ongoing battle against COVID-19.

Language: Английский

---

Contemporary Study for Detection of COVID-19 Using Machine Learning with Explainable AI

Computers, materials & continua/Computers, materials & continua (Print), Journal Year: 2024, Volume and Issue: 80(1), P. 1075 - 1104

Published: Jan. 1, 2024

The prompt spread of COVID-19 has emphasized the necessity for effective and precise diagnostic tools. In this article, a hybrid approach in terms datasets as well methodology by utilizing previously unexplored dataset obtained from private hospital detecting COVID-19, pneumonia, normal conditions chest X-ray images (CXIs) is proposed coupled with Explainable Artificial Intelligence (XAI). Our study leverages less preprocessing pre-trained cutting-edge models like InceptionV3, VGG16, VGG19 that excel task feature extraction. further enhanced inclusion t-SNE (t-Distributed Stochastic Neighbor Embedding) technique visualizing extracted image features Contrast Limited Adaptive Histogram Equalization (CLAHE) to improve before extraction features. Additionally, an Attention Mechanism utilized, which helps clarify how model makes decisions, builds trust artificial intelligence (AI) systems. To evaluate effectiveness approach, both benchmark permissions Jinnah Postgraduate Medical Center (JPMC) Karachi, Pakistan, are utilized. 12 experiments, showcased remarkable performance achieving 100% accuracy vs. pneumonia classification 97% distinguishing cases. Overall, across all classes, achieved 98% accuracy, demonstrating its efficiency differentiating it other disorders (Pneumonia healthy) while also providing insights into decision-making process models.

Language: Английский

---

Interpretable Deep Learning Model for Tuberculosis Detection Using X-Ray Images

Published: Jan. 1, 2024

Language: Английский

---

Privacy Preserving Federated Learning for Lung Cancer Classification

Published: Dec. 13, 2023

Lung cancer is characterized by high mortality and incidence rates, making it one of the most prevalent cancers globally. Early detection significantly improves chances survival for individuals affected this disease. The histopathological diagnosis a crucial factor in determining specific type cancer. In recent years, there has been significant increase novel computer-aided diagnostic techniques utilizing deep learning algorithms early lung However, sharing sensitive patient data restricted regulations such as HIPAA GDPR, primarily due to privacy concerns. Given current constraints, institutions face challenges effectively exchanging information enhance accuracy classification. order address issue classification, we propose federated approach. This methodology involves employing local models with an Inception-v3 backbone carry out classification images & updating global model based on weights. These have obtained from LC25000 dataset. dataset were analyzed, which consisted three distinct classes. Each class contained total 5000 images. applied achieved 99.867% categorizing into performance proposed framework demonstrated superiority over other existing methodologies. Furthermore, solution addresses concerns associated medical among different institutions.

Language: Английский

---

Optimizing Lung Condition Categorization through a Deep Learning Approach to Chest X-ray Image Analysis

BioMedInformatics, Journal Year: 2024, Volume and Issue: 4(3), P. 2002 - 2021

Published: Sept. 10, 2024

Background: Evaluating chest X-rays is a complex and high-demand task due to the intrinsic challenges associated with diagnosing wide range of pulmonary conditions. Therefore, advanced methodologies are required categorize multiple conditions from X-ray images accurately. Methods: This study introduces an optimized deep learning approach designed for multi-label categorization images, covering broad spectrum conditions, including lung opacity, normative states, COVID-19, bacterial pneumonia, viral tuberculosis. An model based on modified VGG16 architecture SE blocks was developed applied large dataset images. The evaluated against state-of-the-art techniques using metrics such as accuracy, F1-score, precision, recall, area under curve (AUC). Results: VGG16-SE demonstrated superior performance across all metrics. achieved accuracy 98.49%, F1-score 98.23%, precision 98.41%, recall 98.07% AUC 98.86%. Conclusion: provides effective categorizing X-rays. model’s high various suggests its potential integration into clinical workflows, enhancing speed disease diagnosis.

Language: Английский

---

Machine Learning Based Framework for Lung Cancer Detection and Image Feature Extraction Using VGG16 with PCA on CT-Scan Images

SN Computer Science, Journal Year: 2024, Volume and Issue: 5(8)

Published: Nov. 11, 2024

Language: Английский

---

Uma visão sobre a classificação de pneumonia viral e bacteriana por radiografias de tórax

Journal of Health Informatics, Journal Year: 2024, Volume and Issue: 16(Especial)

Published: Nov. 19, 2024

Objetivo: Este estudo apresenta uma revisão sistemática sobre o uso de Inteligência Artificial (IA), especialmente Deep Learning (DL), no diagnóstico e classificação da pneumonia por radiografias tórax (RXT). Método: O segue protocolo PRISMA conduzindo a em fases identificação, triagem análise artigos base Scopus. Resultados: A recuperou 25 relevantes entre 121 retornados identificou crescente interesse científico pelo tema, além avanços diagnóstico, com alguns estudos alcançando até 99,7% acurácia modelo proposto. Conclusão: detecção precoce é essencial para um tratamento mais eficaz, soluções que auxiliem especialistas são fundamentais. literatura mostra há evolução constante dessas soluções, embora ainda existam gargalos importantes serem resolvidos.

---

Leveraging DenseNet Features for Machine Learning Based Lung Disease Diagnosis From X-Rays

Published: Nov. 14, 2024

Language: Английский

---