Article
Computer Science, Artificial Intelligence
Zahra Mirikharaji, Kumar Abhishek, Alceu Bissoto, Catarina Barata, Sandra Avila, Eduardo Valle, M. Emre Celebi, Ghassan Hamarneh
Summary: Skin cancer is a major health problem that can be alleviated through computer-aided diagnosis. However, skin lesion segmentation is challenging due to various factors. This survey examines deep learning-based methods for skin lesion segmentation and analyzes their input data, model design, and evaluation aspects.
MEDICAL IMAGE ANALYSIS
(2023)
Article
Multidisciplinary Sciences
Shubham Innani, Prasad Dutande, Ujjwal Baid, Venu Pokuri, Spyridon Bakas, Sanjay Talbar, Bhakti Baheti, Sharath Chandra Guntuku
Summary: This paper introduces a framework called EGAN for skin lesion segmentation using computer-aided diagnosis tools. The EGAN framework generates accurate lesion masks through adversarial learning and achieves superior performance on the International Skin Imaging Collaboration Lesion Dataset. Additionally, a lightweight segmentation framework called MGAN is proposed, which achieves comparable performance to EGAN but with fewer training parameters, resulting in faster inference times for low compute resource settings.
SCIENTIFIC REPORTS
(2023)
Article
Mathematics
Mehwish Zafar, Javeria Amin, Muhammad Sharif, Muhammad Almas Anjum, Ghulam Ali Mallah, Seifedine Kadry
Summary: In this article, a method is proposed for the segmentation and classification of skin lesions using a pre-trained model, which achieves high classification accuracy.
Article
Computer Science, Artificial Intelligence
Qiangguo Jin, Hui Cui, Changming Sun, Zhaopeng Meng, Ran Su
Summary: This study introduces a cascade knowledge diffusion network to improve skin lesion classification and segmentation by utilizing knowledge learned from different tasks. The proposed network includes novel feature entanglement modules to transfer knowledge between classification and segmentation tasks effectively. Extensive evaluations with challenge datasets show superior performance without the need for ensemble approaches or external datasets.
APPLIED SOFT COMPUTING
(2021)
Article
Biotechnology & Applied Microbiology
Ruifeng Bai, Mingwei Zhou
Summary: In this paper, we propose a skin lesion segmentation network based on HarDNet (SL-HarDNet). By introducing cascaded fusion module (CFM), spatial channel attention module (SCAM) and feature aggregation module (FAM), automatic segmentation of skin lesions is achieved. Experimental results show that SL-HarDNet outperforms other methods and achieves state-of-the-art segmentation performance.
FRONTIERS IN BIOENGINEERING AND BIOTECHNOLOGY
(2023)
Article
Computer Science, Information Systems
Yuliang Ma, Liping Wu, Yunyuan Gao, Farong Gao, Jianhai Zhang, Zhizeng Luo
Summary: This paper proposes a ultralightweight fully asymmetric convolutional network (ULFAC-Net) for skin lesion segmentation, which addresses the challenges of irregular shape, fuzzy contours and noise interference. The ULFAC-Net utilizes parallel asymmetric convolution (PAC) module and PAC module with dual attention (Att-PAC), as well as a lightweight textual information submodule and asymmetric encoder-decoder architecture. Experimental results demonstrate that ULFAC-Net achieves competitive segmentation performance with minimal parameters and computational operations compared to other methods.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Kai Hu, Jing Lu, Dongjin Lee, Dapeng Xiong, Zhineng Chen
Summary: A novel Attention Synergy Network (AS-Net) is developed in this paper to enhance the discriminative ability for skin lesion segmentation by combining both spatial and channel attention mechanisms, and achieves better performance than other deep neural networks in experiments.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Biotechnology & Applied Microbiology
Qingqing Guo, Xianyong Fang, Linbo Wang, Enming Zhang, Zhengyi Liu
Summary: This paper proposes two novel fusion modules: Attention-based Transformer-And-CNN fusion module (ATAC) and GAting-based Multi-Scale fusion module (GAMS). ATAC combines multi-scale CNN features with global contexts using Transformer, enhancing feature representation. GAMS adaptively weights the information from multiple scales using a lightweight gating mechanism to further improve classification performance. The combination of ATAC and GAMS in an encoder-decoder framework is particularly effective for skin lesion segmentation, especially for lesions of varying sizes and shapes and low contrasts.
FRONTIERS IN BIOENGINEERING AND BIOTECHNOLOGY
(2023)
Article
Computer Science, Information Systems
Mohammad D. Alahmadi
Summary: This paper proposes a Multi-Scale Attention U-Net (MSAU-Net) for skin lesion segmentation. The method improves the typical U-Net by inserting an attention mechanism at the bottleneck of the network to model the hierarchical representation. Experimental results demonstrate that the proposed pipeline outperforms the existing alternatives.
Article
Biology
Shihan Qiu, Chengfei Li, Yue Feng, Song Zuo, Huijie Liang, Ao Xu
Summary: Automatic segmentation of skin lesions is crucial for diagnosing and treating skin diseases. This study proposes a Gated Fusion Attention Network (GFANet) to accurately segment skin lesion images. Experimental results demonstrate that GFANet achieves excellent segmentation performance and stable results on publicly available datasets.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Biology
Kaili Feng, Lili Ren, Guanglei Wang, Hongrui Wang, Yan Li
Summary: Automatic segmentation of skin lesions using the proposed skin lesion Transformer network (SLT-Net) achieved superior performance compared to state-of-the-art methods on three public datasets. The SLT-Net utilized a codec structure and multi-scale context Transformer to improve the accuracy and efficiency of skin lesion segmentation.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Wei Li, Alex Noel Joseph Raj, Tardi Tjahjadi, Zhemin Zhuang
Summary: This study proposes a deep learning approach to address hair occlusion in dermoscopic images, using hair segmentation and inpainting techniques to achieve digital hair removal for more accurate analysis of skin lesions.
PATTERN RECOGNITION
(2021)
Article
Computer Science, Interdisciplinary Applications
Zhonghua Wang, Junyan Lyu, Xiaoying Tang
Summary: Skin lesion segmentation is crucial for early diagnoses and prognoses of skin diseases, but challenging due to variability and blurry boundaries. In this paper, we propose autoSMIM, a novel superpixel-based masked image modeling method, which explores features from unlabeled dermoscopic images for segmentation.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2023)
Article
Computer Science, Information Systems
Nurullah Sahin, Nuh Alpaslan, Davut Hanbay
Summary: The text highlights the significance of early diagnosis in melanoma, the complexity of distinguishing it from other skin lesions, and the development of computer-aided diagnosis systems to assist physicians. It introduces the Bayesian optimized SegNet approach for precise skin lesion segmentation and examines its competitive results in comparison with other methods. The study demonstrates the effectiveness of Bayesian optimized SegNet in improving the segmentation performance of melanoma.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Engineering, Electrical & Electronic
Qiaokang Liang, Hai Qin, Hui Zeng, Jianyong Long, Wei Sun, Dan Zhang, Yaonan Wang
Summary: In this article, a high-precision skin lesion detection system is proposed, which adopts a modular design in the hardware part and a multimodel fusion method (ALEM) in the software part to achieve efficient and accurate skin lesion region segmentation. The experiment shows that ALEM outperforms other methods in tests with limited training data and pixel annotation. The system achieves a high average AUC on the ISIC2017 dataset and is effective in real skin.
IEEE SENSORS JOURNAL
(2023)
Article
Computer Science, Information Systems
Mehreen Kiran, Imran Ahmed, Nazish Khan, Hamood Ur Rehman, Sadia Din, Anand Paul, Alavalapati Goutham Reddy
MULTIMEDIA TOOLS AND APPLICATIONS
(2020)
Article
Computer Science, Information Systems
Nazish Khan, Imran Ahmed, Mahreen Kiran, Hamoodur Rehman, Sadia Din, Anand Paul, Alavalapati Goutham Reddy
MULTIMEDIA TOOLS AND APPLICATIONS
(2020)
Article
Computer Science, Theory & Methods
Sadia Din, Anand Paul
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2020)
Article
Computer Science, Information Systems
Faisal Saeed, Anand Paul, Won Hwa Hong, Hyuncheol Seo
MULTIMEDIA TOOLS AND APPLICATIONS
(2020)
Article
Engineering, Electrical & Electronic
Shipeng Fu, Lu Lu, Hu Li, Zhen Li, Wei Wu, Anand Paul, Gwanggil Jeon, Xiaomin Yang
CIRCUITS SYSTEMS AND SIGNAL PROCESSING
(2020)
Article
Computer Science, Software Engineering
Rathinaraja Jeyaraj, V. S. Ananthanarayana, Anand Paul
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2020)
Article
Environmental Sciences
Rizwan Ali Naqvi, Muhammad Arsalan, Abdul Rehman, Ateeq Ur Rehman, Woong-Kee Loh, Anand Paul
Article
Computer Science, Artificial Intelligence
Ateeq Ur Rehman, Aimin Jiang, Abdul Rehman, Anand Paul, Sadia Din, Muhammad Tariq Sadiq
JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING
(2020)
Article
Computer Science, Artificial Intelligence
Rathinaraja Jeyaraj, V. S. Ananthanarayana, Anand Paul
JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING
(2020)
Editorial Material
Computer Science, Artificial Intelligence
Karthigai Kumar, Anand Paul, Joy Iong-Zong Chen
JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING
(2020)
Article
Computer Science, Information Systems
Abdul Rehman, Anand Paul, Awais Ahmad, Gwanggil Jeon
COMPUTER COMMUNICATIONS
(2020)
Article
Engineering, Electrical & Electronic
Anandkumar Balasubramaniam, Malik Junaid Jami Gul, Varun G. Menon, Anand Paul
Summary: Intelligent Transportation System (ITS) is generating vast amounts of data, which can be effectively reduced using Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and Non-negative Matrix Factorization (NMF). The integration of Blockchain technology enhances data integrity and the use of smart contracts between insurance companies can streamline data collection for uncertain situations.
IETE TECHNICAL REVIEW
(2021)
Article
Computer Science, Artificial Intelligence
M. Junaid Gul, Abdul Rehman, Anand Paul, Seungmin Rho, Rabia Riaz, Jeonghong Kim
Article
Computer Science, Information Systems
Karshiev Sanjar, Olimov Bekhzod, Jaesoo Kim, Anand Paul, Jeonghong Kim
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2020)