COMPLEMENTARY NETWORK WITH ADAPTIVE RECEPTIVE FIELDS FOR MELANOMA SEGMENTATION

Research output: Chapters, Conference Papers, Creative and Literary Works (RGC: 12, 32, 41, 45)32_Refereed conference paper (with ISBN/ISSN)peer-review

8 Scopus Citations
View graph of relations

Related Research Unit(s)

Detail(s)

Original languageEnglish
Title of host publicationIEEE ISBI 2020 International Symposium on Biomedical Imaging
Subtitle of host publication2020 SYMPOSIUM PROCEEDINGS
PublisherInstitute of Electrical and Electronics Engineers
Pages2010-2013
ISBN (Electronic)978-1-5386-9330-8
ISBN (Print)978-1-5386-9331-5
Publication statusPublished - Apr 2020

Publication series

NameProceedings - International Symposium on Biomedical Imaging
ISSN (Print)1945-7928
ISSN (Electronic)1945-8452

Conference

Title17th International Symposium on Biomedical Imaging, IEEE ISBI 2020
LocationVirtual
PlaceUnited States
CityIowa City
Period3 - 7 April 2020

Abstract

Automatic melanoma segmentation in dermoscopic images is essential in computer-aided diagnosis of skin cancer. Existing methods may suffer from the hole and shrink problems with limited segmentation performance. To tackle these issues, we propose a novel complementary network with adaptive receptive filed learning. Instead of regarding the segmentation task independently, we introduce a foreground network to detect melanoma lesions and a background network to mask non-melanoma regions. Moreover, we propose adaptive atrous convolution (AAC) and knowledge aggregation module (KAM) to fill holes and alleviate the shrink problems. AAC explicitly controls the receptive field at multiple scales and KAM convolves shallow feature maps by dilated convolutions with adaptive receptive fields, which are adjusted according to deep feature maps. In addition, a novel mutual loss is proposed to utilize the dependency between the foreground and background networks, thereby enabling the reciprocally influence within these two networks. Consequently, this mutual training strategy enables the semi-supervised learning and improve the boundary-sensitivity. Training with Skin Imaging Collaboration (ISIC) 2018 skin lesion segmentation dataset, our method achieves a dice coefficient of 86.4% and shows better performance compared with state-of-the-art melanoma segmentation methods1.

Research Area(s)

  • adaptive receptive fields, Melanoma segmentation, semi-supervised learning

Citation Format(s)

COMPLEMENTARY NETWORK WITH ADAPTIVE RECEPTIVE FIELDS FOR MELANOMA SEGMENTATION. / Guo, Xiaoqing; Chen, Zhen; Yuan, Yixuan.

IEEE ISBI 2020 International Symposium on Biomedical Imaging: 2020 SYMPOSIUM PROCEEDINGS. Institute of Electrical and Electronics Engineers, 2020. p. 2010-2013 9098417 (Proceedings - International Symposium on Biomedical Imaging).

Research output: Chapters, Conference Papers, Creative and Literary Works (RGC: 12, 32, 41, 45)32_Refereed conference paper (with ISBN/ISSN)peer-review