Attention-based deep learning for tire defect detection: Fusing local and global features in an industrial case study

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 126473 - 126473

Published: Jan. 1, 2025

Language: Английский

---

Enhanced Diagnosis of Skin Cancer from Dermoscopic Images Using Alignment Optimized Convolutional Neural Networks and Grey Wolf Optimization

Journal of Computing Theories and Applications, Journal Year: 2025, Volume and Issue: 2(3)

Published: Jan. 15, 2025

Skin cancer (SC) is a highly serious kind of that, if not addressed swiftly, might result in the patient’s demise. Early detection this condition allows for more effective therapy and prevents disease development. Deep Learning (DL) approaches may be used as an efficient tool SC (SCD). Several DL-based algorithms automated SCD have been reported. However, models are needed to improve accuracy. As result, paper introduces new strategy based on Grey Wolf optimization (GWO) methodologies CNN. The proposed methodology has four stages: preprocessing, segmentation, feature extraction, classification. method utilizes Convolutional Neural Network (CNN) extract features from Regions Interest (ROIs). CNN employed categorization, whereas GWO approach enhances accuracy by refining edge segmentation. This technique probabilistic model accelerate convergence algorithm. Employing optimize structure weight vectors CNNs can enhance diagnostic minimum 5%, evaluation outcomes. application its performance comparison with other methods indicate that predicted average 95.11% without Accuracy 92.66%, respectively, enhancing 2.5% when we train our GWO.

Language: Английский

---

Personalized recommendation system to handle skin cancer at early stage based on hybrid model

Network Computation in Neural Systems, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 40

Published: Jan. 15, 2025

Skin cancer is one of the most prevalent and harmful forms cancer, with early detection being crucial for successful treatment outcomes. However, current skin methods often suffer from limitations such as reliance on manual inspection by clinicians, inconsistency in diagnostic accuracy, a lack personalized recommendations based patient-specific data. In our work, we presented Personalized Recommendation System to handle Cancer at an stage Hybrid Model (PRSSCHM). Preprocessing, improved deep joint segmentation, feature extraction, classification are major steps identify stages cancer. The input image first preprocessed using Gaussian filtering method. Improved segmentation employed segment image. A set features including Median Binary Pattern (MBP), Gray Level Co-occurrence Matrix (GLCM), Local Direction Texture (ILDTP) extracted next step. Finally, hybrid includes Bi-directional Long Short-Term Memory (Bi-LSTM) Deep Belief Network (DBN) used process, where training will be carried out Integrated Bald Eagle Average Subtraction Optimizer (IBEASO) algorithm via optimizing weights models.

Language: Английский

---

An intelligent framework for skin cancer detection and classification using fusion of Squeeze-Excitation-DenseNet with Metaheuristic-driven ensemble deep learning models

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 3, 2025

Skin cancer is the most dominant and critical method of cancer, which arises all over world. Its damaging effects can range from disfigurement to major medical expenditures even death if not analyzed preserved timely. Conventional models skin recognition require a complete physical examination by specialist, time-wasting in few cases. Computer-aided medicinal analytical methods have gained massive popularity due their efficiency effectiveness. This model assist dermatologists initial significant for early diagnosis. An automatic classification utilizing deep learning (DL) help doctors perceive kind lesion improve patient's health. The one hot topics research field, along with development DL structure. manuscript designs develops Detection Cancer Using an Ensemble Deep Learning Model Gray Wolf Optimization (DSC-EDLMGWO) method. proposed DSC-EDLMGWO relies on biomedical imaging. presented initially involves image preprocessing stage at two levels: contract enhancement using CLAHE noise removal wiener filter (WF) model. Furthermore, utilizes SE-DenseNet method, fusion squeeze-and-excitation (SE) module DenseNet extract features. For process, ensemble models, namely long short-term memory (LSTM) technique, extreme machine (ELM) model, stacked sparse denoising autoencoder (SSDA) employed. Finally, gray wolf optimization (GWO) optimally adjusts models' hyperparameter values, resulting more excellent performance. effectiveness approach evaluated benchmark database, outcomes measured across various performance metrics. experimental validation portrayed superior accuracy value 98.38% 98.17% under HAM10000 ISIC datasets other techniques.

Language: Английский

---

Pedestrian trajectory prediction via physical-guided position association learning

Engineering Science and Technology an International Journal, Journal Year: 2025, Volume and Issue: 64, P. 102008 - 102008

Published: March 8, 2025

Language: Английский

---

Modified Whale Optimization Algorithm for Multiclass Skin Cancer Classification

Mathematics, Journal Year: 2025, Volume and Issue: 13(6), P. 929 - 929

Published: March 11, 2025

Skin cancer is a major global health concern and one of the deadliest forms cancer. Early accurate detection significantly increases chances survival. However, traditional visual inspection methods are time-consuming prone to errors due artifacts noise in dermoscopic images. To address these challenges, this paper proposes an innovative deep learning-based framework that integrates ensemble two pre-trained convolutional neural networks (CNNs), SqueezeNet InceptionResNet-V2, combined with improved Whale Optimization Algorithm (WOA) for feature selection. The features extracted from both models fused create comprehensive set, which then optimized using proposed enhanced WOA employs quadratic decay function dynamic parameter tuning advanced mutation mechanism prevent premature convergence. fed into machine learning classifiers achieve robust classification performance. effectiveness evaluated on benchmark datasets, PH2 Med-Node, achieving state-of-the-art accuracies 95.48% 98.59%, respectively. Comparative analysis existing optimization algorithms skin approaches demonstrates superiority method terms accuracy, robustness, computational efficiency. Our outperforms genetic algorithm (GA), Particle Swarm (PSO), slime mould (SMA), as well models, have reported 87% 94% previous studies. A more effective selection methodology improves accuracy reduces overhead while maintaining technique can improve early-stage diagnosis, shown by data.

Language: Английский

---

Enhancing Early Detection of Skin Cancer in Clinical Practice with Hybrid Deep Learning Models

Engineering Technology & Applied Science Research, Journal Year: 2025, Volume and Issue: 15(2), P. 20927 - 20933

Published: April 3, 2025

Skin cancer is a significant global health issue where early detection essential to improve outcomes. This study evaluates hybrid deep learning models that combine CNN architectures (MobileNetV2, ResNet-18, EfficientNet-B0, and others) with metadata (age, lesion localization) for classification using the SLICE-3D subset of ISIC 2024 dataset. MobileNetV2 achieved recall 99.2% an accuracy 97.7%, while EfficientNet-B0 demonstrated 98.5% 97.2%, making them ideal telemedicine in resource-limited settings due their low computational demands. ResNet-18 DenseNet-121, recalls 99.0% 98.7%, respectively, excelled clinical applications but required greater resources. These show great potential as accessible accurate tools improving skin detection. Future work should validate these findings on diverse datasets optimize preprocessing further enhance sensitivity diagnostic accuracy.

Language: Английский

---

Enhancing skin lesion classification: a CNN approach with human baseline comparison

PeerJ Computer Science, Journal Year: 2025, Volume and Issue: 11, P. e2795 - e2795

Published: April 15, 2025

This study presents an augmented hybrid approach for improving the diagnosis of malignant skin lesions by combining convolutional neural network (CNN) predictions with selective human interventions based on prediction confidence. The algorithm retains high-confidence CNN while replacing low-confidence outputs expert assessments to enhance diagnostic accuracy. A model utilizing EfficientNetB3 backbone is trained datasets from ISIC-2019 and ISIC-2020 SIIM-ISIC melanoma classification challenges evaluated a 150-image test set. model’s are compared against 69 experienced medical professionals. Performance assessed using receiver operating characteristic (ROC) curves area under curve (AUC) metrics, alongside analysis resource costs. baseline achieves AUC 0.822, slightly below performance experts. However, improves true positive rate 0.782 reduces false 0.182, delivering better minimal involvement. offers scalable, resource-efficient solution address variability in image analysis, effectively harnessing complementary strengths humans CNNs.

Language: Английский

---