An Integrated Deep Learning Model with EfficientNet and ResNet for Accurate Multi-Class Skin Disease Classification

Diagnostics, Год журнала: 2025, Номер 15(5), С. 551 - 551

Опубликована: Фев. 25, 2025

Background: Medical diagnosis for skin diseases, including leukemia, early cancer, benign neoplasms, and alternative disorders, becomes difficult because of external variations among groups patients. A research goal is to create a fusion-level deep learning model that improves stability disease classification performance. Methods: The design merges three convolutional neural networks (CNNs): EfficientNet-B0, EfficientNet-B2, ResNet50, which operate independently under distinct branches. network uses its capability extract detailed features from multiple strong architectures reach accurate results along with tight precision. fusion mechanism completes operation by transmitting extracted dense dropout layers generalization reduced dimensionality. Analyses this utilized the 27,153-image Kaggle Skin Diseases Image Dataset, distributed testing materials into training (80%), validation (10%), (10%) portions ten disorder classes. Results: Evaluation proposed revealed 99.14% accuracy together excellent precision, recall, F1-score metrics. Conclusions: approach demonstrates potential as starting point dermatological automation since it shows promise clinical use in classification.

Язык: Английский

A skin disease classification model based on multi scale combined efficient channel attention module

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Фев. 19, 2025

Skin diseases, a significant category in the medical field, have always been challenging to diagnose and high misdiagnosis rate. Deep learning for skin disease classification has considerable value clinical diagnosis treatment. This study proposes model based on multi-scale channel attention. The network architecture of consists three main parts: an input module, four processing blocks, output module. Firstly, improved pyramid segmentation attention module extract features image entirely. Secondly, reverse residual structure is used replace backbone network, integrated into achieve better feature extraction. Finally, adaptive average pool fully connected layer, which convert aggregated global several categories generate final task. To verify performance proposed model, this two commonly datasets, ISIC2019 HAM10000, validation. experimental results showed that accuracy was 77.6 $$\%$$ series dataset 88.2 HAM10000 dataset. External validation data added evaluation validate further, comprehensive proved effectiveness paper.

Язык: Английский

Combining the Variational and Deep Learning Techniques for Classification of Video Capsule Endoscopic Images

Deleted Journal, Год журнала: 2025, Номер unknown

Опубликована: Янв. 3, 2025

Gastrointestinal tract-related cancers pose a significant health burden, with high mortality rates. In order to detect the anomalies of gastrointestinal tract that may progress cancer, video capsule endoscopy procedure is employed. The number endoscopic ( $$\mathcal {VCE}$$ ) images produced per examination enormous, which necessitates hours analysis by clinicians. Therefore, there pressing need for automated computer-aided lesion classification techniques. Computer-aided systems utilize deep learning (DL) techniques, as they can potentially enhance anomaly detection However, most DL techniques available in literature utilizes static frames purpose, uses only spatial information image. addition, perform binary classification. Thus, presented work proposes framework multi-class using dynamic images. proposed algorithm combination fractional variational model and model. captures estimating optical flow color maps. Optical maps are fed training. performs task localizes region interest maximum class score. inspired Faster RCNN approach, its backbone architecture EfficientNet B0. achieves average AUC value 0.98, mAP 0.93, 0.878 balanced accuracy value. Hence, efficient image interest.

Язык: Английский