Dense Fully Convolutional Network for Skin Lesion Segmentation.
Year: 2017
Type: preprint
Abstract: Skin cancer is a deadly disease and is on the rise in the world. Computerized diagnosis of skin cancer can accelerate the detection of this type of cancer that is a key point in increasing the survival rate of patients. Lesion segmentation in skin images is an important step in computerized detection of the skin cancer. Existing methods for this aim usually lack accuracy especially in fuzzy borders of lesions. In this paper, we propose a new class of fully convolutional networks with novel dense pooling layers for segmentation of lesion regions in non-dermoscopic images. Unlike other existing convolutional networks, the proposed dense pooling layers are designed to preserve all of the input features. This has led to highly accurate segmentation of lesions. Our proposed method produces dice score of 91.6% which outperforms all state-of-the-art algorithms in segmentation of skin lesions based on the Dermquest dataset.
Source: arXiv (Cornell University)
Language: English
Subfield: Oncology
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W2779734280