Deep Attention Network for Melanoma Detection Improved by Color Constancy

Ma, Ze; Yin, Shiqun

doi:10.1109/itme.2018.00037

Cited by 6 publications

(6 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, the attention mechanism has also been successfully used for the automatic classification of skin images. Ma et al 22 conceived an attention‐based module that improves the recognition architecture by determining the real color of the skin lesion independently of the data collection conditions. Gessert et al 23 proposed an attention‐based approach for full exploitation of high‐resolution skin images.…”

Section: Related Workmentioning

confidence: 99%

Att2ResNet: A deep attention‐based approach for melanoma skin cancer classification

Boulahia

Benatia

Bouzar

2021

Int J Imaging Syst Tech

View full text Add to dashboard Cite

This paper presents an in‐depth study on the contribution and integration of attention mechanisms into deep learning architectures for melanoma classification. Indeed, the concept of attention helps guide the learning process to focus on some parts of the input image deemed to be the most significant. Nevertheless, to the best of our knowledge, a study on how and where to integrate such mechanisms has never been conducted in the context of melanoma classification. Consequently, we propose such a study in three main stages. First, we propose an improvement in the transfer learning process for the specific case of melanoma classification. Then, we develop a basic variant of our architecture allowing to integrate an additive attention mechanism in an adapted ResNet network. We finally investigate the impact of the number of mechanisms and their location in the melanoma classification architecture. Better results are reported on two standard datasets.

show abstract

Section: Related Workmentioning

confidence: 99%

Att2ResNet: A deep attention‐based approach for melanoma skin cancer classification

Boulahia

Benatia

Bouzar

2021

Int J Imaging Syst Tech

View full text Add to dashboard Cite

show abstract

“…The network proposed in the corresponding paper simultaneously addressed the lesion segmentation and the classification task, where two deep FCRN were used along with two different training sets for this paper. Deep attention network optimised using attention mechanism and fisher criteria is explained by Ma and Yin [34] for melanoma detection. This paper tackles the problem of insufficient training data by using the attention module to learn features of the dermoscopic images.…”

Section: Related Workmentioning

confidence: 99%

Diagnosis of melanoma from dermoscopic images using a deep depthwise separable residual convolutional network

Sarkar¹,

Chatterjee²,

Hazra³

2019

IET image process

View full text Add to dashboard Cite

Melanoma is one of the four major types of skin cancers caused by malignant growth in the melanocyte cells. It is the rarest one, accounting to only 1% of all skin cancer cases. However, it is the deadliest among all the skin cancer types. Owing to its rarity, efficient diagnosis of the disease becomes rather difficult. Here, a deep depthwise separable residual convolutional algorithm is introduced to perform binary melanoma classification on a dermoscopic skin lesion image dataset. Prior to training the model with the dataset noise removal from the images using non‐local means filter is performed followed by enhancement using contrast‐limited adaptive histogram equilisation over discrete wavelet transform algorithm. Images are fed to the model as multi‐channel image matrices with channels chosen across multiple color spaces based on their ability to optimize the performance of the model. Proper lesion detection and classification ability of the model are tested by monitoring the gradient weighted class activation maps and saliency maps, respectively. Dynamic effectiveness of the model is shown through its performance in multiple skin lesion image datasets. The proposed model achieved an ACC of 99.50% on international skin imaging collaboration (ISIC), 96.77% on PH2, 94.44% on DermIS and 95.23% on MED‐NODE datasets.

show abstract

“…Due to the inefficiency of manual clinical inspections of dermoscopic images, the existence of an automated system for a faster and more accurate diagnosis of skin lesions is essential. An automated method can help the dermatologist in decision making and provide the patient with a more effective treatment [8]. Using computers to recognize melanoma is still a challenge.…”

Section: Introductionmentioning

confidence: 99%

“…The segmentation process separates the image into parts to identify regions of interest and, from these regions, obtain relevant information. The automatic diagnosis methods proposed in recent years have shown better results to solve skin lesion segmentation problems use deep learning techniques [8]. However, a disadvantage in using such techniques is related to training, as they require many images, and acquiring dermoscopic images and labeling them by a specialist is not a simple task.…”

Section: Introductionmentioning

confidence: 99%

U-Net_based_Network_Applied_to_Skin_Lesion_Segmentation_An_Ablation_Study

Araújo

2022

CLEIej

View full text Add to dashboard Cite

O câncer de pele é um dos cânceres que mais agrava os problemas de saúde pública. Dentre os tipos de câncer, o melanoma é o mais agressivo. Seu diagnóstico precoce é fundamental para aumentar a possibilidade de tratamento adequado, visando reduzir a taxa de mortalidade. Os dermatologistas geralmente usam métodos manuais para diagnosticar lesões de pele. Além de demorados por serem realizados manualmente, esses métodos podem apresentar resultados diferentes para a mesma lesão quando analisados ??por diferentes especialistas. Portanto, um diagnóstico automatizado pode ser necessário para lidar com esse problema e evitar testes invasivos. Segmentar a lesão cutânea na imagem dermatoscópica pode ser fundamental, pois se trata de umatarefa essencial no processo de análise de imagens. No presente trabalho, um modelo de Rede Neural Convolucional (CNN), baseado na U-Net, é proposto para segmentar a lesão em imagens dermatoscópicas . Este modelo obteve um Jaccard de 0,833 para a base ISIC 2017 e 0,850 para a base ISIC 2018. O modelo proposto possui uma arquitetura mais simples, exigindo menos recursos computacionais. Para verificar a eficácia do método proposto, realizamos um estudo de ablação e experimentos comparativos com outros métodos.

show abstract

Deep Attention Network for Melanoma Detection Improved by Color Constancy

Cited by 6 publications

References 18 publications

Att2ResNet: A deep attention‐based approach for melanoma skin cancer classification

Att2ResNet: A deep attention‐based approach for melanoma skin cancer classification

Diagnosis of melanoma from dermoscopic images using a deep depthwise separable residual convolutional network

U-Net_based_Network_Applied_to_Skin_Lesion_Segmentation_An_Ablation_Study

Contact Info

Product

Resources

About