An interpretable deep neural network for colorectal polyp diagnosis under colonoscopy

Knowledge-Based Systems, Год журнала: 2021, Номер 234, С. 107568 - 107568

Опубликована: Окт. 6, 2021

Язык: Английский

---

Deep Learning Based Classification of Dermatological Disorders

Biomedical Engineering and Computational Biology, Год журнала: 2023, Номер 14

Опубликована: Янв. 1, 2023

Automated medical diagnosis has become crucial and significantly supports doctors. Thus, there is a demand for inventing deep learning (DL) convolutional networks analyzing images. Dermatology, in particular, one of the domains that was recently targeted by AI specialists to introduce new DL algorithms or enhance neural network (CNN) architectures. A high proportion studies field are concerned with skin cancer, whereas other dermatological disorders still limited. In this work, we examined performance 6 CNN architectures named VGG16, EfficientNet, InceptionV3, MobileNet, NasNet, ResNet50 top 3 frequently appear Middle East. An Image filtering denoising were imposed work image quality increase architecture performance. Experimental results revealed MobileNet achieved highest accuracy among can classify disorder (95.7% accuracy). Future scope will focus more on proposing methodology deep-based classification. addition, expand dataset images consider variations.

Язык: Английский

---

Smart IoMT-based segmentation of coronavirus infections using lung CT scans

Alexandria Engineering Journal, Год журнала: 2023, Номер 69, С. 571 - 583

Опубликована: Фев. 17, 2023

Computed Tomography (CT) is one of the biomedical imaging modalities which are used to confirm COVID-19 cases and/or identify infected areas in lung. Therefore, this article aims at assisting crucial radiological task by proposing squeeze-and-excitation networks (SENets) within Internet medical things (IoMT) framework for automated segmentation infections lung CT images. The proposed SE block has been directly integrated with deep residual form Seresnets based on U-Net and LinkNet models. Extensive tests were conducted a public dataset including 20 1800 + annotated slices evaluate results our method. Seresnet models showed good performance Dice score 0.73, structure similarity index 0.98, enhanced alignment measure mean absolute error 0.06. This study demonstrated new advanced tool radiologists achieve automatic using scans. main prospect research work deploying IoMT diagnosis routine positive patients.

Язык: Английский

---

Linear-exponential loss incorporated deep learning for imbalanced classification

ISA Transactions, Год журнала: 2023, Номер 140, С. 279 - 292

Опубликована: Июнь 21, 2023

Язык: Английский

---

Enhancing multimodal medical image analysis with Slice-Fusion: A novel fusion approach to address modality imbalance

Computer Methods and Programs in Biomedicine, Год журнала: 2025, Номер 261, С. 108615 - 108615

Опубликована: Янв. 29, 2025

Язык: Английский

---

IoT‐Based Secured Biomedical Device to Remote Monitoring to the Patient

Опубликована: Фев. 20, 2025

Язык: Английский

---

Path‐enhanced chunking approach with residual attention for medical image segmentation

Medical Physics, Год журнала: 2025, Номер unknown

Опубликована: Март 8, 2025

Medical image segmentation is an essential component of computer-aided diagnosis. While U-Net has been widely used in this field, its performance can be limited by incomplete feature information transfer and the imbalance between foreground background pixel classes medical images. To improve utilization address challenges, such as missing target regions insufficient edge detail preservation, study proposes a method that integrates path enhancement, residual attention, zone-based chunking training. The proposed introduces enhancement structure consisting bottom-up aggregation branch (PAB) multilevel fusion complementary (FEB). PAB aims to transmission semantic positional information, while FEB provides richer representation for mask prediction. Additionally, block with directional frontier support combinatorial attention designed focus on important content units boundary features. further refine segmentation, strategy employed enhance extraction fine-grained details through localized processing. was evaluated extensive ablation experiments, demonstrating consistent across multiple trials. When applied lung nodule computed tomography (CT) images, showed reduction mis-segmented regions. experimental results suggest approach accuracy stability compared baseline methods. Overall, shows promise tasks, particularly applications requiring precise delineation complex structures.

Язык: Английский

---

SEMLCC: A Stacked Ensemble Model with Transfer Learning for High-Accuracy Lung Cancer Classification from CT Images

Procedia Computer Science, Год журнала: 2025, Номер 258, С. 2584 - 2596

Опубликована: Янв. 1, 2025

Язык: Английский

---

When liver disease diagnosis encounters deep learning: Analysis, challenges, and prospects

iLiver, Год журнала: 2023, Номер 2(1), С. 73 - 87

Опубликована: Март 1, 2023

The liver is the second-largest organ in human body and essential for digesting food removing toxic substances. Viruses, obesity, alcohol use, other factors can damage cause disease. diagnosis of disease used to depend on clinical experience doctors, which made it subjective, difficult, time-consuming. Deep learning has breakthroughs various fields; thus, there a growing interest using deep methods solve problems research assist doctors treatment. In this paper, we provide an overview 139 papers from last 5 years. We also show relationship between data modalities, topics, applications Sankey diagrams summarize each topic, addition relations trends these methods. Finally, discuss challenges expectations research.

Язык: Английский

---

Omni-dimensional dynamic convolution feature coordinate attention network for pneumonia classification

Visual Computing for Industry Biomedicine and Art, Год журнала: 2024, Номер 7(1)

Опубликована: Июль 8, 2024

Abstract Pneumonia is a serious disease that can be fatal, particularly among children and the elderly. The accuracy of pneumonia diagnosis improved by combining artificial-intelligence technology with X-ray imaging. This study proposes X-ODFCANet, which addresses issues low excessive parameters in existing deep-learning-based pneumonia-classification methods. network incorporates feature coordination attention module an omni-dimensional dynamic convolution (ODConv) module, leveraging residual for extraction from images. utilizes two one-dimensional encoding processes to aggregate information different spatial directions. Additionally, ODConv extracts fuses four dimensions: dimension kernel, input output channel quantities, kernel quantity. experimental results demonstrate proposed method effectively improve classification, 3.77% higher than ResNet18. model are 4.45M, was reduced approximately 2.5 times. code available at https://github.com/limuni/X-ODFCANET .

Язык: Английский

---