article · Journal Of Big Data
Automated skin lesion classification is often complicated by visual similarities between benign marks and melanoma. A residual deep convolutional neural network has been developed to improve the accuracy of multiclass skin lesion identification. The technique employs multi-layer feature extraction and cross-channel correlation using sliding dot product filters rather than conventional horizontal sliding filters. To tackle dataset imbalance, image and label data are converted into vectors of images and associated weights. The method was evaluated using the ISIC-2019 and ISIC-2020 benchmark datasets. Experimental testing showed that the approach outperforms existing deep convolutional networks in classifying diverse skin lesion categories across these challenging visual datasets.
Skin cancer detection relies heavily on visual screening, but distinguishing benign lesions from melanomas can be difficult. Enhancing computer-aided diagnostic tools with improved feature extraction and better handling of unbalanced medical datasets helps artificial intelligence models recognise complex lesion types more reliably, potentially supporting earlier and more accurate clinical assessments.
The primary application lies in diagnostic software for dermatologists and clinical screening providers. Because the method has only been validated on standard benchmark research datasets (ISIC-2019 and ISIC-2020), it represents an applied and tested algorithmic stage that requires integration into clinical workflows, medical device software, and prospective validation before real-world diagnostic use.
AI-generated from the published abstract. Always read the original work before citing.
Abstract Human Skin cancer is commonly detected visually through clinical screening followed by a dermoscopic examination. However, automated skin lesion classification remains challenging due to the visual similarities between benign and melanoma lesions. In this work, the authors proposed a new Artificial Intelligence-Based method to classify skin lesions. In this method, we used Residual Deep Convolution Neural Network. We implemented several convolution filters for multi-layer feature extraction and cross-channel correlation by sliding dot product filters instead of sliding filters along the horizontal axis. The proposed method overcomes the imbalanced dataset problem by converting the dataset from image and label to vector of image and weight. The proposed method is tested and evaluated using the challenging datasets ISIC-2019 & ISIC-2020. It outperformed the existing deep convolutional networks in the multiclass classification of skin lesions. Graphical Abstract
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1186/s40537-023-00769-6
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
