Identifying Melanoma Skin Disease Using Convolutional Neural Network DenseNet-121
Priyanka Kaushik, Rachna Rathore, Aditya Kumar, Kanishka Kanishka, Ganpat Goshi, Priyanshu Sharma
Abstract
Advancements in melanoma detection are imperative for early intervention and improved patient outcomes. This study explores the transformative potential of DenseNet-121 within CNNs for melanoma identification. Leveraging a dataset from the ISIC 2019 challenge, the model achieved a remarkable 90% accuracy, surpassing traditional models. DenseNet-121 demonstrated adaptability across diverse lesion variations, showcasing its superiority. The study’s implications extend to real-time applications, collaborative efforts, and ethical considerations, heralding a significant stride in reshaping melanoma diagnostic standards.
Topics & Concepts
Convolutional neural networkComputer scienceMelanomaArtificial intelligenceArtificial neural networkPattern recognition (psychology)MedicineCancer researchCutaneous Melanoma Detection and Management