Region Extraction and Classification of Skin Cancer: A Heterogeneous framework of Deep CNN Features Fusion and Reduction

被引:175
作者
Saba, Tanzila [1 ]
Khan, Muhammad Attique [2 ]
Rehman, Amjad [3 ]
Marie-Sainte, Souad Larabi [1 ]
机构
[1] Prince Sultan Univ, Coll Comp & Informat Sci, Riyadh 11586, Saudi Arabia
[2] HITEC Univ, Dept Comp Sci & Engn, Museum Rd, Taxila, Pakistan
[3] Imam Mohammad Ibn Saud Islamic Univ, Coll Comp & Informat Sci, Riyadh, Saudi Arabia
关键词
Skin cancer; Augmentation; Contrast improvement; Boundary extraction; Deep learning; Features selection; SEGMENTATION METHODS; IMAGE SEGMENTATION; DERMOSCOPY IMAGES; BORDER DETECTION; LESION DETECTION; TUMOR; DIAGNOSIS; DISEASES; REMOVAL; NETWORK;
D O I
10.1007/s10916-019-1413-3
中图分类号
R19 [保健组织与事业(卫生事业管理)];
学科分类号
摘要
Cancer is one of the leading causes of deaths in the last two decades. It is either diagnosed malignant or benign - depending upon the severity of the infection and the current stage. The conventional methods require a detailed physical inspection by an expert dermatologist, which is time-consuming and imprecise. Therefore, several computer vision methods are introduced lately, which are cost-effective and somewhat accurate. In this work, we propose a new automated approach for skin lesion detection and recognition using a deep convolutional neural network (DCNN). The proposed cascaded design incorporates three fundamental steps including; a) contrast enhancement through fast local Laplacian filtering (FlLpF) along HSV color transformation; b) lesion boundary extraction using color CNN approach by following XOR operation; c) in-depth features extraction by applying transfer learning using Inception V3 model prior to feature fusion using hamming distance (HD) approach. An entropy controlled feature selection method is also introduced for the selection of the most discriminant features. The proposed method is tested on PH2 and ISIC 2017 datasets, whereas the recognition phase is validated on PH2, ISBI 2016, and ISBI 2017 datasets. From the results, it is concluded that the proposed method outperforms several existing methods and attained accuracy 98.4% on PH2 dataset, 95.1% on ISBI dataset and 94.8% on ISBI 2017 dataset.
引用
收藏
页数:19
相关论文
共 69 条
[1]   Automatic skin tumour border detection for digital dermoscopy using a new digital image analysis scheme [J].
Abbas, Q. ;
Garcia, I. F. ;
Rashid, M. .
BRITISH JOURNAL OF BIOMEDICAL SCIENCE, 2010, 67 (04) :177-183
[2]   Microscopic skin laceration segmentation and classification: A framework of statistical normal distribution and optimal feature selection [J].
Afza, Farhat ;
Khan, Muhammad A. ;
Sharif, Muhammad ;
Rehman, Amjad .
MICROSCOPY RESEARCH AND TECHNIQUE, 2019, 82 (09) :1471-1488
[3]   Skin lesion segmentation and recognition using multichannel saliency estimation and M-SVM on selected serially fused features [J].
Akram T. ;
Khan M.A. ;
Sharif M. ;
Yasmin M. .
Journal of Ambient Intelligence and Humanized Computing, 2024, 15 (01) :1083-1102
[4]  
[Anonymous], 2018, INT C MED IM DEEP LE
[5]  
[Anonymous], IEEE J BIOMEDICAL HL
[6]  
[Anonymous], 2016, CoRR abs/1512.00567, DOI DOI 10.1109/CVPR.2016.308
[7]  
[Anonymous], 2018, ARXIV180705711
[8]  
[Anonymous], 2017, Automatic skin lesion analysis using largescale dermoscopy images and deep residual networks
[9]   Final Version of 2009 AJCC Melanoma Staging and Classification [J].
Balch, Charles M. ;
Gershenwald, Jeffrey E. ;
Soong, Seng-jaw ;
Thompson, John F. ;
Atkins, Michael B. ;
Byrd, David R. ;
Buzaid, Antonio C. ;
Cochran, Alistair J. ;
Coit, Daniel G. ;
Ding, Shouluan ;
Eggermont, Alexander M. ;
Flaherty, Keith T. ;
Gimotty, Phyllis A. ;
Kirkwood, John M. ;
McMasters, Kelly M. ;
Mihm, Martin C., Jr. ;
Morton, Donald L. ;
Ross, Merrick I. ;
Sober, Arthur J. ;
Sondak, Vernon K. .
JOURNAL OF CLINICAL ONCOLOGY, 2009, 27 (36) :6199-6206
[10]   Step-wise integration of deep class-specific learning for dermoscopic image segmentation [J].
Bi, Lei ;
Kim, Jinman ;
Ahn, Euijoon ;
Kumar, Ashnil ;
Feng, Dagan ;
Fulham, Michael .
PATTERN RECOGNITION, 2019, 85 :78-89