Download PDFOpen PDF in browser

An AutomaticSkin Melanoma Detection Based on Convolution Neural Network

EasyChair Preprint no. 8350

10 pagesDate: June 22, 2022


Cancer that develops in the tissues of the skin is referred to as skin cancer. This form of cancer can damage the tissues that are nearby, lead to disability, and even result in death. After cervical and breast cancer, the most common type of cancer seen in Indonesians is skin cancer, which ranks third overall. The detrimental effects of skin cancer can be minimized and brought under control if an accurate diagnosis and prompt, appropriate treatment are given. As a result of the comparable appearance of the disease among malignancy and benign tumor lesions, medical professionals spend significantly more time trying to diagnose these types of lesions. Utilizing the Convolutional Neural Network, the system that was constructed for this study had the capability of automatically identifying skin cancer as well as benign tumor lesions (CNN). The model that has been presented has three hidden layers, each of which has an output channel that ranges from 16 to 32 to 64. With a learning rate of 0.001, the model that has been suggested makes use of a number of optimizers including Adam, and Nadam. The Adam optimizer achieves the greatest results, with an accuracy value of 93.9 percent, when it comes to classifying the skin lesions from the ISIC dataset into the following four categories: dermatofibroma, nevus pigmentosus, squamous cell carcinoma, and melanomas. The accuracy of the existing technique for classifying skin cancer has been outperformed by the findings that were obtained.

Keyphrases: CNN algorithm, deep learning, machine learning, Melanoma detection, Skin Melanoma

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
  author = {Ashish Meshram and Anup Gade and Abhimanyu Dutonde},
  title = {An AutomaticSkin Melanoma Detection Based on Convolution Neural Network},
  howpublished = {EasyChair Preprint no. 8350},

  year = {EasyChair, 2022}}
Download PDFOpen PDF in browser