Deep Convolutional Neural Networks in Medical Image Analysis: A Review

Information, Год журнала: 2025, Номер 16(3), С. 195 - 195

Опубликована: Март 3, 2025

Deep convolutional neural networks (CNNs) have revolutionized medical image analysis by enabling the automated learning of hierarchical features from complex imaging datasets. This review provides a focused CNN evolution and architectures as applied to analysis, highlighting their application performance in different fields, including oncology, neurology, cardiology, pulmonology, ophthalmology, dermatology, orthopedics. The paper also explores challenges specific outlines trends future research directions. aims serve valuable resource for researchers practitioners healthcare artificial intelligence.

Язык: Английский

---

Unlocking a Promising Future: Integrating Blockchain Technology and FL-IoT in the Journey to 6G

IEEE Access, Год журнала: 2024, Номер 12, С. 115411 - 115447

Опубликована: Янв. 1, 2024

The rapid advancement of technology has set higher standards for the next generation wireless communication networks, known as 6G. These networks go beyond simple task connecting devices and aim to establish a self-sustaining system within society. One key factors in achieving this goal is integration AI services apps through Internet Things (IoT), which will be made possible with support 6G technology. artificial intelligence (AI) play crucial role enhancing protocols, architectures, operations networks. To achieve collaborative IoT applications, Federated Learning (FL) emerged popular method. FL enables training without need data sharing, ensuring privacy security. However, also faces challenges, such presence malicious risk single-point failure. address these concerns, blockchain (BCT) offers secure efficient solution. By leveraging blockchain, issues can effectively tackled, providing reliable framework implementing FL-IoT applications.

Язык: Английский

---

Efficient Federated Learning Using Dynamic Update and Adaptive Pruning with Momentum on Shared Server Data

ACM Transactions on Intelligent Systems and Technology, Год журнала: 2024, Номер unknown

Опубликована: Сен. 2, 2024

Despite achieving remarkable performance, Federated Learning (FL) encounters two important problems, i.e., low training efficiency and limited computational resources. In this paper, we propose a new FL framework, FedDUMAP, with three original contributions, to leverage the shared insensitive data on server in addition distributed edge devices so as efficiently train global model. First, simple dynamic update algorithm, which takes advantage of while dynamically adjusting steps order speed up convergence improve accuracy. Second, an adaptive optimization method algorithm exploit momentum each local device for superior Third, develop layer-adaptive model pruning carry out specific operations, is adapted diverse features layer attain excellent trade-off between effectiveness efficiency. Our proposed model, combines techniques has significantly better performance compared baseline approaches terms (up 16.9 times faster), accuracy 20.4% higher), cost 62.6% smaller).

Язык: Английский

---

Exploring privacy mechanisms and metrics in federated learning

Artificial Intelligence Review, Год журнала: 2025, Номер 58(8)

Опубликована: Май 3, 2025

Язык: Английский

---

Enhancing IoT Scalability and Interoperability Through Ontology Alignment and FedProx

Future Internet, Год журнала: 2025, Номер 17(4), С. 140 - 140

Опубликована: Март 25, 2025

The rapid expansion of IoT devices has introduced major challenges in ensuring data interoperability, enabling real-time processing, and achieving scalability, especially decentralized edge computing environments. In this paper, an advanced framework FedProx with ontology-driven standardization is proposed, which can meet such comprehensively. On the one hand, it guarantee semantic consistency across different kinds using unified ontology, so that from multiple sources could be seamlessly integrated; on other solves non-IID issues limited resources servers by FedProx. Experimental findings indicate outperforms FedAvg, a remarkable accuracy level 89.4%, having higher convergence rates, attaining 30% saving communication overhead through gradient compression. addition, ontology alignment procedure yielded 95% success rate, thereby uniform preprocessing domains, including traffic monitoring parking management. model demonstrates outstanding scalability flexibility to new devices, while maintaining high performance during evolution. These highlight its great potential for deployment smart cities, environmental monitoring, IoT-based ecosystems, creation more efficient integrated solutions these areas.

Язык: Английский

---

Trustworthy and Incentivized Federated Learning Based on Blockchain

Lecture notes in computer science, Год журнала: 2025, Номер unknown, С. 205 - 216

Опубликована: Янв. 1, 2025

Язык: Английский

---

Trustworthy federated learning: privacy, security, and beyond

Knowledge and Information Systems, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 26, 2024

Язык: Английский

---

A Framework to Design Efficent Blockchain-based Decentralized Federated Learning Architectures

IEEE Open Journal of the Computer Society, Год журнала: 2024, Номер 5, С. 705 - 723

Опубликована: Янв. 1, 2024

Язык: Английский

---