ADEGUNADEGUNADEGUNADEGUNADEGUNADEGUNADEGUN.pdf

Received July 27, 2020, accepted August 5, 2020, date of publication August 14, 2020, date of current version August 26, 2020.
Digital Object Identifier 10.1109/ACCESS.2020.3016651
FCN-Based DenseNet Framework for Automated
Detection and Classification of Skin
Lesions in Dermoscopy Images
ADEKANMI A. ADEGUN AND SERESTINA VIRIRI , (Member, IEEE)
School of Mathematics, Statistics and Computer Science, University of KwaZulu-Natal, Durban 4000, South Africa
Corresponding author: Serestina Viriri (viriris@ukzn.ac.za)
This work was supported by the University of KwaZulu-Natal, South Africa.
ABSTRACT Skin Lesion detection and classification are very critical in diagnosing skin malignancy.
Existing Deep learning-based Computer-aided diagnosis (CAD) methods still perform poorly on challenging
skin lesions with complex features such as fuzzy boundaries, artifacts presence, low contrast with the
background and, limited training datasets. They also rely heavily on a suitable turning of millions of
parameters which often leads to over-fitting, poor generalization, and heavy consumption of computing
resources. This study proposes a new framework that performs both segmentation and classification of
skin lesions for automated detection of skin cancer. The proposed framework consists of two stages: the
first stage leverages on an encoder-decoder Fully Convolutional Network (FCN) to learn the complex and
inhomogeneous skin lesion features with the encoder stage learning the coarse appearance and the decoder
learning the lesion borders details. Our FCN is designed with the sub-networks connected through a series
of skip pathways that incorporate long skip and short-cut connections unlike, the only long skip connections
commonly used in the traditional FCN, for residual learning strategy and effective training. The network also
integrates the Conditional Random Field (CRF) module which employs a linear combination of Gaussian
kernels for its pairwise edge potentials for contour refinement and lesion boundaries localization. The second
stage proposes a novel FCN-based DenseNet framework that is composed of dense blocks that are merged
and connected via the concatenation strategy and transition layer. The system also employs hyper-parameters
optimization techniques to reduce network complexity and improve computing efficiency. This approach
encourages feature reuse and thus requires a small number of parameters and effective with limited data. The
proposed model was evaluated on publicly available HAM10000 dataset of over 10000 images consisting
of 7 different categories of diseases with 98% accuracy, 98.5% recall, and 99% of AUC score respectively.
INDEX TERMS Skin lesion, deep leraning, CAD, classification, FCN, CRF, DenseNet, encoder- decoder,
hyper-parameter, skin cancer.
I. INTRODUCTION
A Malignant tumor is a disorder in the human body in which
unusual cells divide uncontrollably and destroy body tis-
sue [1]. One of the prevailing malignancies in humans today
is skin cancer [2] and this has been stated to be widespread
in some parts of the world [3]–[6]. Among various categories
of skin cancer [7]–[9], melanoma is the most deadly and dan-
gerous form of cancer [3]. Timely identification and diagnosis
of skin cancer can cure nearly 95% of cases [10]. Primarily,
this disease is diagnosed visually via clinical screening and
The associate editor coordinating the review of this manuscript and
approving it for publication was Jiachen Yang .
analysis of dermoscopic, biopsy, and histopathological
images [2], [10]. However, accurate diagnosis of skin lesions
using these techniques is difficult, time-consuming, and
error-prone even for experienced radiologists; considering the
heterogeneous appearances, irregular shapes, and boundaries
of the skin lesion lesions [11] as shown in Fig.1. These
traditional approaches to skin lesions detection are highly
intensive and laborious. They also require magnified and
well-illuminated skin images for clear identification of the
lesions [12], [13].
Rule-based techniques for detecting the type of skin lesions
mostly employ rules such as ABCD-rule, 3-point checklist,
7-point checklist, and Menzies-rule [14], [15]. These rules
VOLUME 8, 2020
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
150377

A. A. Adegun, S. Viriri: FCN-Based DenseNet Framework for Automated Detection and Classification of Skin Lesions
FIGURE 1. Challenging skin lesions samples: a) hair artefact, b) ruler
mark artefact, c) low contrast, d) color illumination, e) bubbles,
(f) irregular boundaries, (g) blood vessels, (h) frame artefact.
have always been the foundation for diagnosis and detection
by dermatologists [16], [17]. In the ABCDrule, the ABCD
represents asymmetry, border structure, color variation,
and diameter respectively, and asymmetry means that the
two sides are unequal while symmetry means that they
match. This assists in distinguishing between the benign from
the malignant skin lesions. For example, the color composi-
tion is always single for Benign but can be two or more for
malignant. The diameter of the general structure of the benign
is always very small like a fraction of an inch but bigger and
wider in malignant [17]. This dermoscopy imaging procedure
is error-prone and requires years of experience in difficult
situations.
Conventional methods for detecting skin lesions include
thresholding methods, clustering methods, edge-based, and
region-based techniques [18]. Various machine learning
based-CADe systems have been designed in assisting the
medicals in automated detection of skin cancer [19]. Tradi-
tional machine-learning algorithms such as gradient boosting,
support vector machine (SVM) [20], artificial neural network
(ANN) [21], etc have been employed by researchers for the
diagnosis of skin lesions. For instance, Hameed et al. [22]
extracted gray-level co-occurrence matrix features from skin
lesions and utilized SVM to perform features classification.
Murugan et al. [23] utilized Gaussian filters to extract lesion
features and employed SVM to classify the extracted features.
Seeja and Suresh [24] employed a Convolutional Neural
Network (CNN) based U-net algorithm for segmentation of
skin lesion and utilized a set of features extraction meth-
ods such as Local Binary Pattern ( LBP), Edge Histogram
(EH), Histogram of Oriented Gradients (HOG) and Gabor
methods to extract color, texture, and shape features from the
segmented image. The extracted features were sent into the
K-Nearest Neighbor (KNN), Naïve Bayes (NB), SVM, and
Random Forests (RF) classifiers to categorize them into either
melanoma or benign lesions. However, since skin lesions vary
in shape, size, and border features, the low-level hand-crafted
methods utilized in these conventional CAD, methods pos-
sess limited discriminative capability due to their intrinsic
naivety and locality. They also have other drawbacks, such as
lack of adaptability in which the methods are not transferable
for solving new problems [25].
In recent times, deep learning architectures have been
utilized to develop computerized automated systems for
detection, classification, and diagnosis of several diseases via
medical image analysis [26]. They have produced promising
results most especially in the detection and classification of
skin lesions cancers. They have been proven to outperform
both human and existing Computer-Aided Diagnostic sys-
tems. The performance of the deep learning-based system on
skin lesion detection has been evaluated against dermatolo-
gists and the conventional machine learning techniques in the
recent past. Heckler et al. [27] explored the possibility and
the advantages of using artificial intelligence for skin cancer
classification against dermatologists. They established that
CNN outperforms humans in the task of skin cancer classifi-
cation. They employed 112 dermatologists from 13 German
university hospitals and an independently well-trained CNN
to classify a set of 300 biopsy-verified skin lesions into five
diagnostic categories. Esteva et al. [11] performed classifica-
tion of skin lesions using a single CNN that was trained end-
to-end using only images’ pixels and disease labels of skin
lesions as inputs. The performance of their system was tested
against 21 board-certified dermatologists on biopsy-proven
clinical images. According to Brinker et al. [28], CNN pos-
sess the ability to classify images of skin cancer on par
with dermatologists and can as well enable life-saving and
quick diagnoses, through the installation of apps on mobile
devices most especially outside the hospital. Guha et al. [29]
performed experiments to compare the performance of deep
learning-based techniques with traditional machine learning
techniques such as SVM in the detection and classification of
skin lesions. They utilized three techniques: SVM, VGGNet,
and Inception-ResNet-v2, for the classification of seven cat-
egories of skin diseases.
Although existing deep learning techniques are gener-
ally more powerful than traditional methods most especially
in the ability to learn highly discriminative features, their
performance is still limited due to the following reasons:
(1) Training deep learning methods with limited labeled data
can lead to over-fitting and poor generalization. (2) Most deep
learning methods require higher memory and computational
resources with heavy reliant on millions of parameters tuning
to perform efficiently. (3) The deep learning approach also
needs to be able to process multi-scale and multi-resolution
features since the skin lesion images are always acquired
with different devices with varying imaging resolution.
(4) Automated detection of the skin lesion is also challenging
due to the heterogeneous visual attributes of skin lesions
images and fine-grained contrast in the appearance of skin
lesions [19].
This study proposes a new deep learning framework for
automated detection and classification of skin lesion images.
The proposed CAD framework consists of two main steps:
the first step is detection and segmentation of skin lesions by
a multi-stage encoder-decoder network and the refinement of
the detected lesion border with post-processing CRF modules
for better classification into various disease categories, and
the second step is the classification of detected lesions with
an FCN-based DenseNet system. In the first step, an encoder-
decoder network was constructed to detect and segment skin
150378 VOLUME 8, 2020

lesion of different scales and resolution, in which the encoder
network connects with the decoder sub-network via series
of skip pathway which is designed to integrate high-level
semantic information with lower-level feature maps for effi-
cient detection. This overcomes the problem by learning the
complex and inhomogeneous skin lesion features. The system
leverages on the skip pathway which is the combination of
both long and short skip connections, using the short skip
connections to build very deep FCNs with residual learning
strategy while the long skip connections in the upsampling
stage reuse the residual features to recover spatial infor-
mation lost during downsampling. Specifically, in addition
to the extraction of semantic features from skin lesions,
our multi-stage encoder-decoder network also integrates a
CRF module to further refine the extracted features for a
well-defined boundary. The CRF module exploits a linear
combination of Gaussian kernels for its pairwise edge poten-
tials and efficient mean-field inference. This ensures contour
refinement and lesion boundaries localization in boosting
the detection performance of the classifier. In the second
step, we devised an FCN-based DenseNet framework which
utilized a concatenation strategy in which, the output feature
maps are concatenated with the incoming feature maps to pro-
duce a large number of feature maps with a small number of
convolution layers; solving the problem of a limited dataset.
Also, we introduced a regularization strategy with hyperpa-
rameter optimization to train the images, which can enhance
the network performance, reduce the network complexity,
and improve computing efficiency for better classification
performance.
The performance of the proposed model was evaluated on
publicly available and standard HAM10000 dataset, which
contains samplings from seven typical skin lesion categories:
Melanoma (MEL), Melanocytic-Nevi (NV), Basal-Cell Car-
cinoma (BCC), Actinic-Keratoses and Intra-epithelial Carci-
noma (AKIEC), Benign-Keratosis (BKL), Dermato-fibroma
(DF), and Vascular (VASC) lesions. Standard evaluation met-
rics such as Accuracy, F1-Score, AUC, and Recall (Sensi-
tivity) were used to measure the performance of the system.
The results of 98% accuracy, 98.5% recall, and 99% of AUC
scores respectively were obtained. Each unit of the proposed
system functions independently, so we utilize the classifi-
cation unit to also classify some samples of un-segmented
skin lesion images from HAM10000 and PH2, and the
segmented skin lesion images from ISBI 2017; the perfor-
mance were compared (in the two scenarios; segmented and
non-segmented lesion images).
II. LITERATURE BACKGROUND AND RELATED WORKS
A. BACKGROUND
Improving the performance of deep learning techniques for
the analysis of skin lesion images requires a robust frame-
work. This research examines three major factors that limit
the performance of deep learning techniques in the analysis of
skin lesion images: Firstly, the performance of deep learning
methods is reliant on the appropriate tuning of a large number
of parameters. Most deep learning frameworks are composed
of millions of parameters which directly increases the system
complexity and the required computational resources [30].
Secondly, skin lesion images analysis is challenging because
of the coarse visual appearances of these images which makes
detection difficult [31]. These images are intricate with com-
plex features such as fuzzy boundaries, low contrast with the
background, inhomogeneous textures, or contain artifacts.
Lastly, the performance of deep learning methods is primarily
leveraged on large labeled datasets [32] to hierarchically learn
the features that correspond to the appearance and the seman-
tics of the skin lesion images [31]. They generally require
large training data set to build efficient models and utilizing
limited labeled data in a situation with skin cancer analy-
sis can result in over-fitting and poor generalization [31].
Training deep learning methods with limited data can also
lead to the generation of the coarse region of interest (ROI)
detections and poor boundary definitions [30].
B. APPROACHES AND RELATED WORKS
Lately, deep learning-based methods have been developed for
the detection and classification of skin lesions into various
categories of skin cancer. Various approaches and techniques
of deep learning systems have been employed in the past to
tackle this problem. These include methods such as trans-
fer learning, unsupervised learning, supervised, and hybrid
approaches. These approaches are however with each of them
having its pros and cons:
1) UNSUPERVISED LEARNING
Unsupervised fully automatic approaches have been
employed in the past to tackle the problem with the scarcity of
annotated medical training datasets in the analysis, segmenta-
tion, and classification of skin lesions images. Unsupervised
deep learning approaches utilize strategies that derive infer-
ences directly from a dataset which can be further used for
decision making [33]. These methods generally rely on tech-
niques such as iterative or statistical region merging, thresh-
olding, and energy functions application [31], [34], [35].
They also utilize a probabilistic generative model with the
capacity to learn the hierarchy level of features and the
probability distribution over any given input space for image
classification tasks [33], [36]. They, therefore, do not require
large training datasets and are not in any way limited by the
scarcity of annotated medical training dataset. However, they
are limited in performance by the inhomogeneous appear-
ance of medical images such as skin lesion images with the
intensity distribution of the lesion containing multiple peaks.
They also have a limited capacity to accurately segment
challenging skin lesions, such as lesions that touch the image
boundary and those with artifacts. Recently some of these
methods which have been applied for medical images anal-
ysis include Restricted Boltzmann machines (RBM), Deep
belief networks (DBN), Deep Boltzmann machine (DBM),
Generative adversarial network(GAN) auto-encoders and its
several variants [33], [34].
VOLUME 8, 2020 150379

Semantic segmentation of medical images via the unsu-
pervised approaches is thus challenging in producing accept-
able accuracy in a life-related diagnosis. For example,
Pereira et al. [37] developed a deep learning model that
utilized the Restricted Boltzmann Machine for unsupervised
feature learning of brain lesion images. They also used a Ran-
dom Forest classifier for the segmentation of brain lesions.
The system achieved a dice coefficient accuracy of 74% on
brain MRI image datasets. Also, Akhavan Aghdam et al. [38]
developed a deep learning algorithm using DBN for the pre-
diction of autism. The algorithm was evaluated on Autism
Brain Imaging Data Exchange I and II datasets with an
average accuracy of 65.56%. They combined some series
of unsupervised models comprising of rs-fMRI, GM, and
WM for DBN. A Deep Neural Network (DNN) model based
on Restricted Boltzmann Machine model was proposed by
Al Nahid et al. [39] for the classification of Histopathological
breast-cancer images. The system achieved an overall accu-
racy of 88.7% when evaluated on the breast-cancer image
dataset.
Zhu et al. [40] presented an unsupervised classification
model for MRI-based prostate cancer detection. The sys-
tem achieved an averaged Section-based evaluation accu-
racy of 89.90% when evaluated on 21 real patient’s dataset.
An unsupervised model that utilized a bag of adversarial
features (BAF) for the identification of mild traumatic brain
injury (MTBI) in patients using their diffusion magnetic res-
onance images (MRI) was proposed by Minaee et al. [41].
The system was evaluated on a dataset of 227 samples that
include 109 MTBI patients, and 118 age and sex-matched
healthy controls with the mean values of over 80% accu-
racy on brain MRI images. Also in similar research,
Vergara et al. [42] employed a resting-state functional net-
work connectivity (rsFNC) model for MTBI identification.
They then used a linear Support Vector Machine for the image
classification. The system achieved a classification accuracy
of 84.1% on extracted rsFNC features. Lastly, an experiment
was performed to compare the performance of the supervised
deep learning-based approach with the unsupervised deep
learning-based approach on skin lesion images segmentation
by Ali et al. [43]. It was discovered in the experiment that
even though the unsupervised approach can detect the fine
structures of skin lesions in some occasions, the supervised
approach still produced much higher accuracy in terms of dice
coefficient and Jaccard index with the supervised approach
achieving a 77.7% dice coefficient score as against 40% dice
coefficient score achieved by the unsupervised approach.
2) HYBRID LEARNING
Recently, models that employed the combination of
supervised and unsupervised approaches yielded an improved
performance in medical image analysis. Minaee et al. [44]
carried out the general survey of various supervised and
unsupervised methods for both semantic and instance-level
segmentation. Feng et al. [45] for example proposed a Reti-
nal vessel segmentation (RVS) based on a cross-connected
convolutional neural network (CcNet) for the automatic
segmentation of retinal vessel trees. The system explored
cross-training for model training and prediction of the pixel
classes. The system was evaluated on two publicly avail-
able datasets of DRIVE and STARE with performance
results of 0.7625 and 0.9528 sensitivity and accuracy scores
on Drive datasets and 0.7709 and 0.9633 sensitivity and
accuracy scores on the Stare dataset. A High-Resolution
Network (HRNet) model which maintains high-resolution
representations throughout the process of image analysis was
proposed by Wang et al. [46] for general object detection.
The proposed system has been applied in a wide range
of applications, including human pose estimation, semantic
segmentation, and object detection with an average 85%
detection accuracy. Multi-task Framework for Skin Lesion
Detection and segmentation that utilized the combination
unsupervised and supervised models have been proposed by
Vesal et al. [47]. A Multi-Class Multi-Level (MCML) classi-
fication model based on an unsupervised divide and conquer
rule was developed by Hameed et al. [48] for medical image
classification. The model explored both traditional machine
learning and advanced deep learning approaches. The model
was evaluated on 3672 images with a diagnostic accuracy
of 96.47%. Ali et al. [49] proposed a model that combined
the Gaussian Bayes ensemble with Convolutional Neural
Network for the tasks of feature extraction and automatic
detection of border irregularity from skin lesion images. The
system achieved accuracy, sensitivity, specificity, and F-score
results of 93.6%, 100%, 92.5%, and 96.1%, respectively
when evaluated on skin lesion images dataset. Vesal et al. [47]
proposed a faster region-based CNN (Faster-RCNN) for
skin lesion images analysis. The system was composed of
an unsupervised region proposal network (RPN) model for
generating bounding boxes or region proposals for lesion
localization in imaging. A supervised modified UNET model,
SkinNet, which employed a softmax classifier was then used
for the semantic segmentation of the images. The system
achieved 93% for the Dice coefficient and 96% accuracy
performance when trained and evaluated on ISBI 2017 and
the PH2 datasets. From this literature, inferences can be
made that the unsupervised approaches are still limited in
medical image analysis most especially in the analysis of
skin lesion images. They require millions of parameters for
their architectures and thereby requiring a large number of
computational resources.
3) TRANSFER LEARNING
Transfer learning approaches have been utilized in training
supervised deep learning models for medical image analysis.
This has been employed to overcome the challenges with
limited training labeled dataset. Transfer learning approaches
are generally effective but are suboptimal on medical images
analysis due to the large discrepancy that exists with the
target data in this context. This can be seen from the visual
appearance of images and class labels, which may cause the
feature extraction process to be biased to the source data and
150380 VOLUME 8, 2020

eventually generalize less well on the target data [50]. This is
because the models are originally pre-trained on images that
are different from medical images. Some of these images may
include images such as animals, automobiles, equipment, etc.
which have different forms from medical images that usually
possess characteristics such as fuzzy boundaries, fine-grained
variability, and heterogeneous appearance. Systems based on
this approach are also heavy-weight and require millions of
parameters and a large number of computational resources.
These challenges have limited the performance of these mod-
els on medical image analysis.
The performance evaluation of these models on medi-
cal images shows that they are still yet to outperform the
state-of-the-art. For example, El-Khatib et al. [51] applied
the transfer learning approach on CNN models which were
already pre-trained on ImageNet and Places365 datasets.
They also used other pre-trained models such as GoogleNet,
ResNet-101, and NasNet-Large. These models were then
fine-tuned on skin lesions datasets via the transfer learn-
ing approach for skin lesion images detection. The models
were integrated and evaluated on skin lesion images with
the accuracy scores of 88.33% 88.24% 88.46% 86.79% for
Accuracy, Specificity, Sensitivity, and Dice coefficient
respectively. Also, an intelligent diagnosis scheme was
proposed for multi-class skin lesion classification by
Hammed et al. [52] using a hybrid approach of deep convolu-
tion neural network and SVM based error-correcting output
codes (ECOC). A pre-trained CNN model, AlexNET, was
utilized for feature extraction. The system achieved an overall
accuracy of 86.21% when evaluated on skin lesion image
datasets. Another CNN model pre-trained on Imagenet was
utilized by Almaraz-Damian et al. [53] for the extraction and
segmentation of both handcraft and deep learning features.
The system achieved similar results of 87% accuracy with the
models developed by El-Khatib et al.
Kalouche et al. [54] utilized three different models: logistic
regression, a deep neural network, and a pre-trained CNN
VGG-16 model for skin lesion images classification. The
system achieved a 78% classification accuracy on skin lesion
images containing melanoma cancer. A segmentation recom-
mender based on transfer learning and crowdsourcing algo-
rithm was proposed by Soudani and Walid [55]. The system
utilized two pre-trained CNN models based on VGG16 and
ResNet50 for features extraction and classification of skin
lesion images. The system achieved 78.6% accuracy with
the two models when evaluated on ISIC 2017 skin lesion
dataset. An automatic skin lesions classification system that
employed the transfer learning approach was presented by
Hosny et al. [56]. The proposed system was based on a
pre-trained CNN model based on Alex-net architecture. The
architecture weight was then fine-tuned on the ISIC skin
lesion dataset. The system achieved 95.91% accuracy when
evaluated on ISIC 2017 skin lesion dataset. Akram et al. [57]
proposed another classification system based on three
pre-trained CNN models: DenseNet 201, Inception-
ResNet-v2, and Inception-V3. These models were integrated
and fused with an entropy-controlled neighborhood compo-
nent analysis (ECNCA) algorithm for feature selection and
classification of skin lesion images. The system also achieved
95.9% when evaluated on ISBI 2017 skin lesion dataset.
Ahmad et al. [58] performed discriminative analysis and
classification of features from skin disease images using the
CNN model based on two pre-trained models: ResNet152 and
InceptionResNet-V2. They achieved an average accuracy
of 84.91% and 87.42% on ResNet152 and InceptionResNet-
V2 respectively. An integrated diagnostic system that utilized
segmentation techniques for optimization to improve the
classification performance of deep learning models for skin
lesion classification was proposed by Al-Masini et al. [59].
The system was based on four pre-trained CNN architec-
tures: Inception-v3, ResNet-50, Inception-ResNet-v2, and
DenseNet-201. These were integrated and evaluated on both
ISIC 2016 and ISIC 2017 skin lesion datasets. The sys-
tem achieved the prediction accuracies of 77.04% on ISIC
2016, and 81.29% on ISIC 2017 dataset. Finally, an in-
depth analysis of several deep learning-based techniques such
as a fully convolution neural network, pre-trained model,
ensemble, and handcrafted methods for skin lesion analysis
and melanoma detection was carried out by Naeem et al. [60].
They concluded that by performing fine-tuning of hyper-
parameters, overfitting can be reduced and the performance
of a deep learning system can be improved greatly for the
analysis and diagnosis of skin lesion images.
4) SUPERVISED LEARNING
Lastly, we also reviewed the supervised learning approaches
that have been utilized for skin lesion analysis and detec-
tion; Esteva et al. [11] devised a deep learning-based
method using CNN for automated classification and detec-
tion of skin lesions. They utilized a CNN model that was
trained in an end-to-end approach from images’ pixels and
disease labels serving as inputs to achieve the classification
of skin lesions. They performed two binary classifications
with keratinocyte-carcinomas versus benign seborrheic-
keratosis, and malignant melanomas versus benign nevi.
Gessert et al. [61] utilized a multi-resolution ensemble of
CNNs comprising of EfficientNets, SENet, and ResNeXt
WSL for the detection of skin lesions. They achieved sat-
isfactory performance on a much smaller dataset of HAM
10000 and ISIC 2018. Khalid et al. [56] also performed
an automatic skin lesions classification system using the
approach of transfer learning and the pre-trained deep neu-
ral network. The transfer learning was applied on Alex-net
and the architecture’s weight was fine-tuned. The system
was able to detect and classify segmented color image
lesions into either melanoma and nevus or melanoma, seb-
orrheic keratosis, and nevus. Three popular skin lesion
datasets; MED-NODE, Derm-IS, and Derm-Quest and ISIC
were utilized for both training and testing. They obtained
classification-accuracy of 96.86%, 97.70%, and 95.91% on
the datasets respectively.
VOLUME 8, 2020 150381

A segmentation methodology, FRCN, was developed for
the segmentation of skin lesions by first learning the full
resolution features of individual image’ pixel of the input
skin lesion images. The system was assessed on two publicly
accessible datasets; ISBI 2017 and PH2 datasets. The pro-
posed system attained a segmentation accuracy of 95.62%
for some representative of clinical benign cases, 90.78% of
melanoma cases, and 91.29% of seborrheic-keratosis cases
in the ISBI 2017 dataset [62]. Ratul et al. [16] devised
a deep learning model with dilated convolution based on
transfer learning from four standard architectures; VGG16,
VGG19, MobileNet, and Inception-V3. They utilized the
HAM10000 dataset that comprises a total of 10015 dermo-
scopic images of seven skin lesion categories with large
class imbalances for training, validating, and testing. They
achieved a classification accuracy of 87.42%, 85.02%,
88.22%, and 89.81%, with VGG16, VGG19, MobileNet, and
InceptionV3 respectively.
Shimizu et al. [63] proposed a method that is suitable for
both melanocytic skin lesions (MSLs) and non-melanocytic
skin lesions (NoMSLs). They devised a method to identify
Melanomas, Nevi, BCCs, and Seborrheic-keratosis using fea-
tures such as color, sub-region, and texture. They utilized both
layered model and flat models to function as baselines for
evaluating performance. Their method was tested on 964 der-
moscopy images: 105 melanomas, 692 nevi, 69 BCCs, and
98 SKs with the layered model outperforming the flat models
and achieved an accuracy of 90.48%, 82.51%, 82.61%, and
80.61% for melanomas, nevi, BCCs, and SKs, respectively.
Alqudah et al. [64] employed both GoogleNet and AlexNet
with transfer learning and optimization gradient descent
adaptive momentum learning rate (ADAM) for the classifi-
cation of skin lesion images. The methods were applied on
the ISBI 2018 database to perform classification of images
into three main categories; benign, melanoma, seborrheic
keratosis under two schemes: classification of segmented
and non-segmented lesion images. The overall classification
accuracy of 92.2% was obtained for the segmented dataset
and 89.8% was obtained for the non-segmented dataset.
Preprocessing steps such as lesion image enhance-
ment, filtering, and segmentation were utilized on lesion
images to acquire the Region-of-Interest (ROI) by
Almaraz-Damian et al. [53]. Both handcraft features and
deep learning features were extracted. ABCD rule was used
to extract features such as shape, color, and texture while
CNN was used to further extract the deep learning features.
The CNN architecture used was first pre-trained on Image-
net. MI measurement metrics were used as fusion rules for
collecting vital details from both the handcraft and deep
learning features. Kawahara et al. [65] utilized a linear
classifier that was trained on extracted features from CNN.
The CNN was pre-trained on natural images to differenti-
ate between ten skin lesions. The approach also utilized a
fully convolutional network for the extraction of multi-scale
features via the pooling-over of augmented feature space.
The proposed approach achieved an accuracy of 85.8% over
a 5-class dataset of 1300 images. Finally, a deeply super-
vised multi-scale network [66] was utilized for the detection
and segmentation of skin cancer from skin lesion images.
They utilized the side output layers of the architecture to
accumulate information from both shallow and deep layers
to design a multi-scale connection block that can process
various changes in cancer size. Generally, the supervised
approaches perform better than the other approaches in the
analysis of skin lesion images.
C. OUR CONTRIBUTIONS
In this research, we devised a fully automatic system for skin
lesion detection and classification on an FCN-based densenet
framework. We propose FCN for the system optimization to
achieve the following: a) to reduce the computational cost and
weight size by integrating compressed convolutional blocks
(via the encoder-decoder and the skip pathway approach)
that are light-weight into the densenet framework; b) the
encoder-decoder and the skip pathway of the FCN will also
allow the system to efficiently extract skin lesion features
even with the limited training dataset.
The main components of the framework that serve as our
contribution include the following:
1) ENCODER-DECODER SEGMENTATION APPROACH
We proposed an efficient pre-processing and segmentation
of skin lesions for effective features extraction by utilizing
an encoder-decoder network in which the sub-networks can
learn and extract the complex features of the skin lesion
with the encoder stage learning the coarse appearance and
localization information while the decoder learns the region
based global features of the lesion. The encoder provides
low-resolution features mapping and the decoder restores
the features into full-resolution and further improves the
boundaries delineation. This mechanism also achieves better
detection and extraction of multi-scale lesion features in a
limited dataset.
2) RESIDUAL LEARNING STRATEGY WITH SKIP PATHWAYS
The skip pathways introduce both long skip and short-cut
connections unlike the only long skip connections commonly
used in the standard FCN. The system leverages the short skip
connections to build very deep FCNs and as a residual learn-
ing strategy for extracting features. The long skip connections
in the up-sampling stage reuse the features to recover spatial
information lost during downsampling. The skip pathways
can hierarchically merge both the down-sampling features
with the up-sampling features and bring the semantic level
of the encoder feature maps closer to that of the decoder to
reliably detect lesions with flexible sizes and scales.
3) INTEGRATION WITH CRF
We employed parallel integration of dense CRFs and fast
mean-field inference which exploits the linear combination
of Gaussian kernels for its pairwise edge potentials. This is to
150382 VOLUME 8, 2020

FIGURE 2. Schematic layout diagram of our proposed deep learning framework for skin lesion images segmentation and classification where
C1...C7 represents the 7-class (AKIEC, BCC, BKL, DF, MEL, NV, VASC).
ensure contour refinement and lesion boundaries localization
to boost classification performance.
4) DENSENET FRAMEWORK
To develop an efficient classification system that eschews the
learning of redundant feature maps to improve the classifi-
cation accuracy, we devised a novel FCN-based DenseNets
framework. DenseNet needs fewer parameters than a coun-
terpart conventional CNN since it does not require learning
redundant feature maps. Our proposed framework can pro-
duce selective features in a data-driven approach that can effi-
ciently process the fine-grained unevenness in the appearance
of skin lesions with a reduced computation cost.
5) CONCATENATION STRATEGY WITH TRANSITION LAYER
The output feature maps are concatenated with the incoming
feature maps to produce a large number of feature maps with
very little convolution. This enables us to use fewer param-
eters to produce a large number of feature maps thereby,
overcoming the limitation with heavy reliance on a large
number of parameters and datasets. The transition layers
utilize a 1×1 convolution layer between the two contiguous
dense blocks for easy information transfer.
6) REGULARIZER STRATEGY AND HYPER-PARAMETERS
OPTIMIZATION
The proposed system employs a regularization strategy and
utilizes dropout modules in between the dense blocks. The
system also performs an experimental tuning of the Hyper-
parameters to enhance the network performance, reduce the
network complexity, and improve computing efficiency.
III. METHODS
The methodology consists of two main components; the first
component is an encoder-decoder network integrated with
a fully connected CRF for lesion contour and boundaries
refinement to produce highly accurate, soft segmentation
maps; the second component is an FCN-based Densenet
framework composing of six consecutive dense blocks with a
fixed feature maps size connected with a transition layer for
effective classification process. The methodology framework
of this research is described and illustrated in Figure 2 and
Figure 4, and discussed within the components stated below:
FIGURE 3. Architectural Diagram for Deep Convolutional
Encoder-Decoder Network.
A. MULTI-SCALE FEATURE LEARNING, DETECTION AND
EXTRACTION
An enhanced encoder-decoder network which is deeply
supervised is employed for the task of feature learning, detec-
tion and extraction of multi-scale and multi-size skin lesion
features. The composition of this network is described below:
1) FEATURE EXTRACTION WITH ENCODER-DECODER
NETWORK
The network is made up of encoder and decoder sections [67]
with each of the sections composed of five consecutive stages
VOLUME 8, 2020 150383

FIGURE 4. The Proposed framework and flow diagram of the Deep
Convolutional Encoder-Decoder Network Integrated with CRFs.
as illustrated in Figure 3. Each of the stages is made up of
a convolution layer with a kernel size of 3 x 3, a ReLU
activation layer, a series of skip pathways, and a concate-
nation layer. The number of convolutional filters increases
from 64 in the first stage to 1024 in the last stage. We have
replaced the usual short skip connection with a series of skip
pathways which is made up of both long and short skip con-
nections. The ReLU activation module is utilized to introduce
nonlinearity which results in faster training for the network.
The encoder section, in addition, utilizes max-pooling mod-
ules for down-sampling tasks. Features vectors are extracted
via the convolution layers from the input images, these are
then down-sampled by half using the max-pooling modules
and the pooling indexes are passed to the corresponding
upsampling layer in the decoder section. This is illustrated
in equation 1.
Yi = U(F(I : r) : d) (1)
where Yi is the final output, F is the downsampled feature
map, r is the RELU activation function, d is the downsam-
pling module and U is the upsampling module
The decoder section then utilizes up-sampling layer to
upsample the feature vectors from the previous layers with
a multiplier factor of 2. These are then concatenated with the
corresponding output feature maps of the matched encoder
section to achieve enriched information, avoid vanishing gra-
dient and restore the lost feature information. The last part
of the decoder section is made up of a convolutional layer
with 1 x 1 kernel and softmax module to perform mapping of
each pixel to a particular category of skin lesion. The softmax
classifier then predicts the class for each pixel with the output
in an N-channel image of probabilities and the predicted
segmentation corresponded to the class with the maximum
probability of each pixel. This is illustrated in equation 2.
P(y = i|x) =
exT wi
P
n=1
exT wn
(2)
where x is the feature map, w is the kernel operator and n
represents the number of classes.
The encoder section achieves a low-resolution feature vec-
tors and can also learn the coarse appearance and the local-
ization details of the skin lesion while the decoder achieves
restored full-resolution feature vectors and can also learn the
lesion boundaries’ features. The system is also able to process
efficiently multi-scale skin lesion images using a scalable
framework that is adaptable and easy to modify.
2) SERIES OF SKIP PATHWAYS FOR CONNECTION
From the diagram in Figure 3, the skip pathway utilizes both
long skip and short-cut connection and the system lever-
ages on the short skip connections to build very deep FCNs
and also as a residual learning strategy for efficient features
extraction. The short-cut connections are made up of 2 x 2
convolution layers and they facilitate features extraction and
learning. The system utilizes the series of skip pathway to
hierarchically merge both the down-sampling features with
the up-sampling features and bring the semantic level of the
encoder feature maps closer to that of the decoder in order
to reliably detect lesions with flexible sizes and scales. The
long skip connections in the up-sampling stage reuse the
extracted features to recover spatial information lost during
downsampling.
B. FULLY CONNECTED CRF FOR POST PROCESSING
Fully connected dense CRFs with an efficient mean-field
approximation and probabilistic inference are integrated into
the Encoder-Decoder networks. The final output of the
encoder-decoder network is then sent into the CRF module
for refinement and enhancement of lesion contour, to produce
the final predicted feature map and mask.
1) GAUSSIAN KERNEL EXPLORATION FOR PAIRWISE EDGE
POTENTIALS IN FULLY CONNECTED CRF
The input image X: x1::::xN and the corresponding labelling
mask Y: y1:::yN are taken into the CRF model in an end-to-
end trainable fashion. CRF utilizes Gibbs distribution [68],
a probabilistic inference model to model P(y|x) for prediction
as follows in equation 3.
P(y|x) =
1
Z(x)
exp[−E(y, x)] (3)
where X : x1 . . . .xN are the input features, Y : y1 . . . yN,
as label mask, E(x|y) is the cost of assigning label to pixel
also known as energy and Z is the constant known as partition
function.
The CRF presents a probabilistic graphical model where
each node represents a pixel in an image, I, and each edge
represents relation between pixels. These then produce the
unary and pairwise terms [69]. The unary term measures
the cost of assigning label y to pixel x and it represents
per-pixel classifications while the pairwise terms shows rela-
tionship between neighbouring pixels and it presents a set of
smoothness constraints. The energy function is represented
by E(x) represents the parameters used by unary and pair-
wise networks as illustrated in equation 4. The Unary poten-
tial encodes local information about a given pixel with the
likelihood of a pixels to belong to a certain class such as
foreground or background. The pairwise potential encode the
150384 VOLUME 8, 2020

neighbourhood information between two neighbouring pixels
and ensures smooth edges and annotations. Unary potential
functions on nodes while the pairwise potentials function on
edges. Assigning the most probable label to each pixel will
give lower energy which implies lower cost, and thus, higher
accuracy.
E(x) =
X
i
ψ(xi) +
X
i<j
ψ(xi, xj) (4)
The values of i and j in the above formula range from 1 to N.
where X : x1 . . . .xN represents input image, Y : y1 . . . yN
represents labelling mask, 9(xi) represents the unary poten-
tials and 9(xi, xj) represents the pair-wise potentials.
Introducing Gaussian Kernel:
We utilized Gaussian kernel function for the mean field
update of all variables in the fully connected CRF model [69].
This enables the CRF model to optimize the probability map
via the exploitation of local similarity of the neighbourhood
pixels. The individual pixels in the unary potentials of the
probability map are propagated according to their neighbour-
hood pixels via the pairwise potentials. A Gaussian kernel
is applied to finally smoothen the boundary and to further
improve the appearance kernel and smoothness kernel. The
Gaussian kernel function is represented in equation 5 as:
k(m)
(fi, fj) = exp(−
1
2
(fi − fj)T
3m
(fi − fj)) (5)
where k(m)(fi, fj) is the Gaussian Kernel function where vec-
tors fi and fj are feature vectors for pixels i and j in an
arbitrary feature space 3m is a symmetric positive-definite
precision matrix. The pairwise potential is defined as a
linear combination of Gaussian kernel in arbitrary feature
space. The pairwise potentials in the model is represented in
equation 6 as:
ψ(xi, xj) = µ(xi, xj)
K
X
m=1
w(m)
k(m)
(fi, fj) (6)
A multi-class image segmentation with color vectors Ii and
Ij is represented in equation 7 as:
k(fi, fj) = w(1)
exp(−
|pi − pj|2
2θ2
−
|Ii − Ij|2
2θ2
) + m (7)
m = w(2)
exp(−
|pi − pj|2
2θ2
(8)
2) EFFICIENT INFERENCE- MEAN FIELD APPROXIMATION
For an efficient inference in fully connected CRFs, the CRF
distribution is approximated by the mean field [69]. Approx-
imate inference program which is based on mean-field
approximation is applied to minimise variational free energy.
This computes a distribution Q(x) instead of the exact distri-
bution P(x) i.e Distribution Q(x) minimises KL-divergence
D(Q||P) and is expressed in equation 9 as:
Q(x) = NiQi(Xi) (9)
This represents the products of independent marginals Qi
and Xi respectively and i ranges from 1 to N. Performing
sequential updates of Qi will guarantee converge. The model
proposes an approach to guarantee convergence with any
shape of the pairwise potentials and with parallel updating
using convolution mean fields together with Gaussian poten-
tials derived from the unary and pairwise potentials.
C. CLASSIFICATION BASED ON DenseNet SCHEME
A novel FCN based Dense-Net framework is utilized for
the classification task of skin lesions into 7 categories. The
structure of this framework is described below:
1) DENSE BLOCKS
An efficient classification system is developed by utiliz-
ing some combination of dense blocks. These dense blocks
exploit DenseNets CNN architecture which does not require
learning of redundant feature maps unlike the traditional
CNN that learns from redundant feature maps. The input
images are first sent into 2 convolution layers with 128 and
256 output channels respectively to boost feature extrac-
tion and learning process before being sent into the dense
blocks. The convolution layers have the kernel size of 2 x 2
and each side of the inputs is zero-padded by one pixel
to keep the feature map constant and reduce the network
parameters size. The architecture is composed of six Dense
Blocks with an equal number of layers; all layers with the
same feature-map sizes and are connected directly with one
another. The first three layers possess an output channel
of 512 each and the remaining three blocks have an output
channel of 1024 each. This is to ensure the utmost information
flow between layers. Each of the dense blocks also consists
of a dense layer, a ReLu activation function and a flatten layer
to downsample the feature maps. In the model, the dense
connections within the dense blocks employ sum operation
for the feature merging inside the dense block to reduce
the computing cost of the dense blocks. The dense blocks
constructed can produce selective features in a data-driven
manner to solve the problem of fine-grained variability in the
appearance of skin lesions. The generated feature maps are
finally processed by a 7 channel dense layer to classify the
merged feature map into 7 categories of skin lesion using the
sigmoid classifier. The DenseNets framework is illustrated
in Figure 5.
2) CONCATENATION STRATEGY
The concatenation strategy is employed to reduce extremely
the number of network parameters in our proposed architec-
ture. The layers in the dense blocks are connected to each
other in a feed-forward pattern and the input feature map for
each layer is concatenated with the feature maps of the pre-
ceding layers. In order to reduce the computing cost, the fea-
tures for all the inner layers of the dense blocks are merged
by sum operation illustrated in equation 11 while the feature
maps form the input and output layer only are concatenated.
The concatenation operation generates an increased number
VOLUME 8, 2020 150385

FIGURE 5. The framework diagram for the dense blocks where C1...C7 represents the 7-classes (AKIEC, BCC, BKL, DF, MEL, NV, VASC).
of feature maps with very little convolution layers which
are then repeatedly used. This also reduces the number of
parameters employed as it eliminates the amount of redundant
feature maps learned through encouraging feature reuse. The
last (nth) layer obtains the feature vectors of all previous
layers x0 . . . .xn−1 as illustrated in equation 10.
xi = k(x1+, . . . . . . ., xi−1) (10)
where xi represents the sum operation for the feature merging
within the dense block and k is the merging function.
y = Cn([x0, x1, . . . . . . ., xn−1]) (11)
where x0 . . . .xn−1 denotes the concatenation of the input
feature-maps with the concantenation function Cn
3) TRANSITION LAYER AND HYPERPARAMETER
OPTIMIZATION
A training strategy was devised in the framework that exploits
both transition layer procedure and hyper-parameter opti-
mization technique. The transition layer is composed of a
convolution layer with a kernel size of 1 x 1, a ReLU acti-
vation function, and a dropout module. This is utilized in
between two neighboring dense blocks for smooth features
transition. Convolutional operation is exploited to prevent
vanishing gradient i.e protecting feature information from
vanishing and also make the parameters of the whole frame-
work effectively learnable. The dropout module performs a
stochastic transformation on the input dimensions to avoid
over-fitting.
Hyper-parameter optimization is introduced to fine-tune
network parameters to optimize the system performance. The
aim is to train the model faster, reduce overfitting, and make
better predictions with the model. Three major optimization
algorithms which include Adaptive Moment Optimization
(Adam), Stochastic Gradient Descent algorithm (SGD) and
Root Mean Square Propagation (RMSprop) were explored
and deployed. Major hyper-parameters in the network such as
learning rate, decay constant and the number of dense layers
were also varied and tuned. The network was finally opti-
mized using Adam optimizer algorithm with the following
parameters set as: (Adam optimizer = 0.0001, batch size =
128, weight decay = 0.001, drop out rate = 0.5). Experiment
results are presented in Table 6 which shows the impact of
varying these hyper-parameters on the system performance.
IV. EXPERIMENTS AND RESULTS
In this section, various experiments were performed to
evaluate the performance of each of the stages of our pro-
posed framework. The segmentation stage was first evalu-
ated, the classification stage was then evaluated and the whole
system was finally evaluated. Publicly available skin lesion
datasets were utilized to demonstrate the performance of each
section of the system and the whole system entirely. The
performance was evaluated and compared with the existing
state-of-the-arts.
A. DATASETS
The datasets used in this work can be categorized into train-
ing, validation and testing datasets:
Our training data contains 10030 images and 1 ground truth
response CSV file was taken from HAM10000 (‘‘Human
Against Machine with 10000 training images’’) [70] dataset.
It is made up of dermatoscopic images collected from differ-
ent populations under different procedures. It is a composi-
tion of important skin lesion diagnostic categories: Actinic
keratoses and intraepithelial carcinoma(akiec), basal cell
carcinoma (bcc), benign keratosis-like lesions (bkl), der-
matofibroma (df), melanoma (mel), melanocytic nevi (nv)
and vascular lesions (vasc). The system was validated
on the validation data that also contains 10030 images
from HAM10000 dataset and also containing skin lesion
diagnostic categories: Actinic keratoses and intraepithe-
lial carcinoma(akiec), basal cell carcinoma (bcc), benign
keratosis-like lesions (bkl), dermatofibroma (df), melanoma
(mel), melanocytic nevi (nv) and vascular lesions (vasc). The
test data are taken from both ISBI 2018 [71] and PH2 [72].
The PH2 dataset is made up of 8-bit RGB color images with a
resolution of 768×560 pixels containing a total of 200 dermo-
scopic images of melanocytic lesions. These include 80 com-
mon nevi, 80 atypical nevi, and 40 melanomas. For the
150386 VOLUME 8, 2020

segmentation section, we utilized ISBI 2017 [73] dataset
which is composed of 2000 images and ground truth labels
respectively for the segmentation model training.
1) DATA AUGMENTATION
We performed on the fly data augmentation on 2000 images
for segmentation and the 10030 images for classification in
the training dataset for both segmentation and classification
process. This process was performed by applying settings
such as flipping, rotation, scaling, and shear on the dataset
as stated below:
1) Rescaling=1./255,
2) Shear range=0.2,
3) Zoom range=0.2,
4) Horizontal flip=True,
5) Rotation = random
B. PERFORMANCE EVALUATION METRICS
The following standard metrics have been employed in this
research to measure the performance of the proposed system
at different stages. They are defined as stated below:
Dice Similarity Coefficient: It is the measures of similarity
between the ground truth and predicted outcomes.
DSC =
2TP
FP + 2TP + FN
(12)
Recall (Sensitivity): This is the proportion of actual posi-
tives that are identified correctly.
Sensitivity =
TP
TP + FN
(13)
Precision: This is the proportion of correctly predicted pos-
itive observations to the total predicted positive observations.
Precision =
TP
TP + FP
(14)
F1 Score: This is the weighted average of Precision and
Recall.
F1Score =
2 ∗ (Recall ∗ Precision)
Recall + Precision
(15)
Specificity: This is the proportion of actual negatives that
are identified correctly.
Specificity =
TN
TN + FP
(16)
Accuracy: This is the proportion of correctly predicted
observation(both true positives and true negatives) to the total
observations.
Accuracy =
TP + TN
TP + TN + FP + FN
(17)
ROC curve: An ROC curve (receiver operating character-
istic curve) is a graph that shows the performance of a clas-
sification model. It is curve plot of Recall vs False Positive
Rate.
AUC(Area Under the ROC Curve): It represents the com-
plete two-dimensional area within the entire ROC curve from
origin (0,0) to point (1,1).
Where FP is the amount of false-positive outcome, FN is
the amount of false-negative outcome, TP is the amount of
true-positive outcome and TN is the amount of true-negative
outcome.
C. RESULTS AND DISCUSSION
In this section, both the automated segmentation and clas-
sification performance of our proposed frameworks were
evaluated and the results compared with the performance of
the state-of-the-art methods. The performance of the segmen-
tation unit was conducted in two phases: In the first phase,
the performance of the multi-scale detection encoder-decoder
network was only evaluated. This performance was evaluated
against existing methods as shown in Table 1 and Table 2.
In the second phase, the encoder-decoder network was inte-
grated with the CRF modules after which the performance
was again evaluated and the result was compared with the
performance of the encoder-decoder network only. The seg-
mented image outputs were compared in Figure 7 and per-
formance metrics results were compared through the chart
in Figure 6. The output was then sent into the classifica-
tion network for further processing. The performance of the
classification unit was also evaluated in two phases: In the
first phase, the performance of the classification system was
evaluated separately on un-segmented images. The results
were compared with the state-of-the-art classification meth-
ods as shown in Table 3. Also, Figure 8 and Figure 9 show
the accuracy performance curve and dice-coefficient curve of
the system respectively. The classification performance of the
TABLE 1. Performance evaluation(%) of the proposed model as against
existing methods for segmentation process on ISBI 2017 Datatest.
TABLE 2. Performance evaluation(%) of the proposed model as against
some recent semantic segmentation models for medical image analysis.
TABLE 3. Performance Evaluation of the Proposed model Compared with
Existing Methods on HAM10000 DataSet.
VOLUME 8, 2020 150387

FIGURE 6. The figure compares the general performance of the
encoder-decoder network using Dice-coefficient, accuracy, senitivity and
specificity when used with CRF and when used without CRF for
segmentation of skin lesion images.
FIGURE 7. The figure shows the segmentation outcome of the
encoder-decoder network when used with CRF and when used without
CRF on ISBI 2017 dataset: The first row shows the Input images;
the second rows shows the Ground truth labels; the third row shows
segmented outcome on encoder-decoder network only; the last row
shows the segmented outcome of encoder-decoder network + CRF.
FIGURE 8. Accuracy and loss performance curves of the proposed
classification model for both validation and training on HAM10000.
system on the 7 categories of skin lesion are presented with
confusion matrix, ROC curve, image classification output
and classification reports in Figure10, Figure 11, Figure 12,
Table 4, Table 5, and Figure 13 respectively. In the sec-
ond phase, the classification system was evaluated on the
TABLE 4. Classification Performance Reports (%) of the proposed model
on HAM10000 Datatest.
TABLE 5. Confusion Matrix of the proposed model on
HAM10000 DataSet.
FIGURE 9. Dice-coefficient performance curves of the proposed
classification model for both validation and training on HAM10000.
segmented skin lesion images from the CRF-based encoder-
decoder network. The classification output of the segmented
images is shown in Figure 14. Finally, the performance of the
classification system with segmented skin lesion images was
compared with its performance with un-segmented images
as shown in the chart in Figure 16. The system was also
evaluated on sample images from PH2 Dataset to test the
generalization ability of the system as shown in Figure 15.
Basically, the evaluation approach adopted focused on evalu-
ation of each stage of the system as stated below:
1) SEGMENTATION AND DETECTION RESULTS
The segmentation model was trained and evaluated on aug-
mented ISBI 2017-challenge dataset containing 2000 images
and 600 images for both training and testing tasks
respectively. In the first section of the experiment, we
evaluate the performance of our multi-scale detection
encoder- decoder network and compare its performance with
state-of-the-arts among which are FrCN, CDNN, FCN, and
mFCNPI methods. The evaluation was carried out on the
ISBI 2017 dataset using metrics such as segmentation accu-
racy, dice-coefficient, sensitivity, and specificity respectively,
150388 VOLUME 8, 2020

FIGURE 10. The figure shows the Confusion matrix evaluation of the
proposed classification model on HAM10000.
and the corresponding results are summarized in Table 1.
As shown in Table 1, we achieved the Accuracy of 95.5%,
Dice Coefficient of 92.1%, Sensitivity of 97.5%, and Speci-
ficity of 96.5% on the ISBI 2017 dataset. This outperformed
FIGURE 11. ROC performance curves of the proposed classification
model for the seven categories of skin lesion in HAM10000.
some existing methods in Table 1. The performance of the
proposed model was also evaluated against some recent
semantic segmentation models for medical image analysis
such as CC-Net, ExFuse, and Multi-class multi-level classi-
fication algorithms as shown in Table 2. The result shows the
highest recall(sensitivity) and dice score of 97.5% and 92.1%
FIGURE 12. The figure shows the classification results of the proposed classification method on some sample unsegmented images from
HAM10000.
VOLUME 8, 2020 150389

when compared with other techniques. This result shows that
the proposed segmentation system can detect and differen-
tiate correctly diseased lesions from the healthy tissues on
ISBI 2017 dataset as shown in Figure 6.
The encoder-decoder network integrated with the CRF
model yields better performance compared with the
encoder-decoder network only from the chart in Figure 6 with
the overall accuracy was improved by 0.5 (96 vs. 95.5),
dice-coefficient was improved by 0.9 (93 vs. 92.1), sensitiv-
ity by 0.5(98 vs. 97.5) and specificity by 0.5(97 vs. 96.5)
respectively when tested on ISBI 2017. Figure 7 shows the
segmentation outputs of both encoder-decoder network with
CRF and without CRF. In Figure 7, the first row shows the
input images, the second row shows the ground truth labels
for the images, the third row shows the segmented output of
the encoder-decoder network without CRF and the last row
shows the segmented output of the encoder-decoder network
when combined with the CRF module. The output result
from Figure 7 shows that; the CRF-based encoder-decoder
network approach gives better detection and segmentation
performance results than only the encoder-decoder approach
on all groups of skin lesions. Both perform better than the
traditional FCN method and other existing methods. The
reason for the better performance is that the CRF method
facilitates feature learning of fine-grained lesions which gives
a well-defined lesion boundary as seen in Figure 7 with
some improvements in accuracy and dice-coefficient score as
shown in the chart in Figure 6. The CRF-based approach out-
performs the encoder-decoder network only, which validates
the effectiveness of the localization. This also highlights that
the combination with the probabilistic graphical CRF model
produces segmentation output with more precise borders as
shown in the fourth row.
2) CLASSIFICATION RESULTS
The classification model was trained and validated on the
HAM10000 which is composed of 10030 skin images
with corresponding class labels. The dataset is composed
of 7 important diagnostic categories of skin lesions which
are represented by AKIEC, BCC, BKL, DF, MEL, NV and
VASC. Sample skin lesion images from PH2 and ISBI 2017
were also used to test the classification model. For a general
evaluation of the classification system, we first evaluated the
performance using metrics such as Accuracy, loss, and dice-
coefficient. Figure 8 and Figure 9 show the accuracy-loss
curve and the dice-coefficient curve of the classification sys-
tem. From Figure 8, we got the overall accuracy of 98.3%
and training loss of 0.6%, and from Figure 9 we got an
overall dice-coefficient of 92%. The classification system
was evaluated using performance metrics such as accuracy,
precision, recall, and F1-score and the results were compared
against existing methods such as Deep Convolutional net-
work with transfer learning, Multi-level Densenet, Dilated
VGG16 and Dilated InceptionV3 as summarised in Table 3.
The classification system outperforms the existing methods
with the overall accuracy, precision, recall, and F1-Score
of 98.3%, 98%, 98.5%, and 98.0% respectively when eval-
uated on HAM10000 as shown in Table 3. The classifica-
tion results from Table 3 can be analyzed as follows: First,
our FCN-based DenseNet classification system obtained the
highest overall accuracy of 98.3% when compared with
recent deep learning methods (i.e., Multi-level Densenet,
Deep Convolution Network with Transfer Learning, Dilated
VGG16 and Dilated InceptionV3 ), indicating better learn-
ing ability which is beneficial for skin lesion classification.
Second, the FCN-based DenseNet network consistently out-
performed the other six deep learning methods in FI-Score,
which implies its ability for effective analysis of discrimina-
tive features for automatic skin lesion classification. Third,
the FCN-based DenseNet network yielded the best perfor-
mance in all other metrics such as Precision and Recall show-
ing its ability in effective identification of relevant instances
and higher measure of completeness and exactness.
The detailed results and experiments focus on comparison
of the performance of the classification model on the 7-class
(AKIEC, BCC, BKL, DF, MEL, NV, VASC). In order to
achieve this, we evaluated the performance using confusion
matrix, ROC curve, image classification output, and classi-
fication reports in Figure 10, Figure 11, Figure 12, Table 4,
Table 5 and Figure 13 respectively. The confusion matrix was
reported across all classes for better evaluation of the perfor-
mance per class and following this, the results of our 7-class
predictions were also reported using the ROC curve. The
results of the performance analyses were presented through
the confusion matrix to get explanatory insights into the
results as shown in Figure 10 and Table 5. The following
analyses were carried out from the confusion matrix table
and reported: 330 AKIEC images were utilized for the exper-
iment; the prediction of 319 images of AKIEC were cor-
rectly classified as AKIEC with a classification accuracy
of 96.66%. Also, there was a prediction of 2 images of
AKIEC FIGURE 8. Accuracy and loss performance curves
of the proposed classification model for both validation and
training on HAM10000 classified incorrectly as BCC, pre-
diction of 3 images of AKIEC were incorrectly classified as
BKL and prediction of 3 images of AKIEC were incorrectly
classified as melanoma. 514 BCC images were also used for
the experiment; the prediction of 495 images of BCC were
correctly classified as BCC with a classification accuracy
of 96.30%. Also, there was prediction of 12 images of BCC
classified incorrectly as AKIEC, and prediction of 7 images
of BCC were incorrectly classified as BKL. 1099 BKL
images were also utilized; the prediction of 1095 images of
BKL was correctly classified as BKL with a classification
accuracy of 99.63%. Also, there was a prediction of 1 image
of a BKL classified incorrectly as a Mel, and prediction
of 3 images of BKL incorrectly classified as NV. 115 DF
images were utilized; the prediction of 111 images of a DF
were correctly classified as DF with classification accuracy
of 96.52%. Also, there was a prediction of 2 images of a DF
classified incorrectly as AKIEC, and prediction of 2 images
of a DF were incorrectly classified as BKL. 1112 MEL
150390 VOLUME 8, 2020

images were utilized for the experiment; the prediction
of 1048 images of MEL was correctly classified as MEL with
a classification accuracy of 94.24%. Also, there was a predic-
tion of 1 image of an AKIEC classified incorrectly as a MEL,
prediction of 18 images of a MEL were incorrectly classi-
fied as a BKL and prediction of 45 images of a MEL were
incorrectly classified as NV. The total number of 6722 NV
images were utilized; the prediction of 6654 images of NV
were correctly classified as NV with classification accuracy
of 98.98%. Also, there was a prediction of 2 images of
an NV classified incorrectly as an AKIEC, prediction of 1
image of an NV incorrectly classified as a BCC, prediction
of 53 images of NV were incorrectly classified as BKL, and
prediction of 12 images of NV were incorrectly classified as
MEL. The total number of 142 VASC images were used; the
prediction of 141 images of a VASC were correctly classified
as VASC with a classification accuracy of 99.29%. Also, there
was a prediction of 1 image of VASC classified incorrectly as
an NV.
The results of ROC analysis are illustrated for the 7 classes
in Figure 11. Here, we represented AKIEC, BCC, BKL,
DF, MEL, NV, VASC by class 0, class 1, class 2, class
3, class 4, class 4, class 5, and class 6 respectively. Look-
ing at the ROC curves from Figure11, the classification
system achieved the highest classification performance on
the identification of class 2 and class 5 which are BCC
and NV with AUC 98%. However, the minimum AUC
achieved was 92% for class 3. The overall AUC of this
model was 98% with the macro average AUC of 98% and
micro-average AUC of 96%. This indicates better detection of
BCC and NV classes with the poorest detection performance
of BKL which was as a result of the size of their training
dataset.
For more insight into the classification performance,
Tables 4–5 shows the classification report of the classification
results by the 7 classes, respectively.
For more insight into the classification performance,
Tables 4–5 shows the classification report of the classification
results by the 7 classes, respectively.The classification result
and output of the sample images used for testing the proposed
classification system are shown in Figure 12. The system
was able to identify most images accurately as illustrated
in Figure 12 except in only two cases highlighted with the
red line: the case of a BCC image classified incorrectly as
BKL in the first row and another case of a MEL image
classified incorrectly as BKL in the second row. This is
further collaborated by the classification performance reports
in Table 4 showing the precision, recall, and F1-Score for
the 7 diagnostic categories of pigmented skin lesions with
MEL and BCC having the lowest recall score of 94% and
96% respectively and also BKL having the lowest precision
of 93%.
For the purpose of this work, over the entire 7-class
dataset was benchmarked to compare the performance of
the state-of-the-arts with the proposed classification method
as shown in Figure 12. The proposed method achieves the
FIGURE 13. The figure shows the comparison of the performance of the
proposed classification method and other methods on classification of
the seven categories of skin lesion.
best performance in all the classes when compared with the
state-of-the-arts.
Classification of the segmented images: The effect of the
segmentation process was examined by applying the clas-
sification networks to classify samples of well-segmented
images from the ISBI 2017 Dataset. The classification result
of segmented images in Figure 14 shows that the skin lesion
images were all correctly classified as either NV or MEL.
This shows the effect of segmentation on the performance of
the proposed classification model. There was improvement
suggesting that, better performance of the classifier on the
segmented images than its performance on the unsegmented
images. This leverages on the effect of CRF-based multi-
scale encoder-decoder network to effectively pre-process
the images and improve the detection accuracy, especially
for the skin lesions with complex features. The FCN-based
Densenet network optimized with CRF-based multi-scale
encoder-decoder network yielded better performance com-
pared with only the FCN-based Densenet network and the
existing state-of-the-art. For example, the overall accuracy
was improved by 0.6 (98.90 vs. 98.30), precision improved
by 0.5 (98.5 vs. 98.0), recall improved by 0.5 (99 vs. 98.5)
and F1-score improved by 0.5 (98.5 vs. 98.0) respec-
tively as shown in Figure 16. It indicates that incorpo-
rating the CRF-based multi-scale encoder-decoder network
could effectively refine the feature learning to improve the
lesion detection accuracy. This is also discovered in Fig-
ure 14 where all the 10 images were detected correctly with
98.9% accuracy.
Generalization Effects: We also tested the generaliza-
tion performance of the FCN-based Densenet network on
the classification of sample skin lesion images from another
skin lesion dataset, PH2, as shown in Figure 15. All the
images were correctly classified except the first image that
was wrongly classified as BKL whereas the image is MEL.
This also shows the effect of segmentation because the
sample images used were not preprocessed and segmented.
Finally, the effect of segmentation also reflects on the
dice-coefficient curve in Figure 9 where the images used were
un-segmented.
VOLUME 8, 2020 150391

FIGURE 14. The figure shows the classification results of the proposed classification method on some sample sample segmented skin lesion
images.
FIGURE 15. The figure shows the classification result output of the proposed classification method on some sample PH2 skin lesion images.
3) EFFECTS OF HYPER-PARAMETERS TUNING
We performed extensive experiments for hyper-parameters
tuning to achieve optimal performance for the proposed
system. Some sets of experiments were conducted to show
the effects of tuning of these hyper-parameters as shown
in Table 6. Major hyper-parameters in the network such as
learning rate, optimizer, decay constant, and the number of
dense layers as were varied and tuned. Three major opti-
mization algorithms were explored and tuned which include
Adaptive Moment Optimization (Adam), stochastic gradient
descent algorithm (SGD) and Root Mean Square Propaga-
tion (RMSprop) optimizers. The aim is to reduce overfitting
and make better predictions with the model. Experiment
results are presented in Table 6 which shows the impact of
varying these hyper-parameters on the system performance.
From the results, the first row presents the best result in
terms of accuracy. The experiment results show that the
accuracy of the model performance has a significant improve-
ment by using hyper-parameter optimization algorithms.
In order to achieve optimal performance, the major hyper-
parameters were explored. These parameters were varied as
follows:
FIGURE 16. The figure compares the general performance (AUC) of the
proposed classification method on segmented and un-segmented lesion
images.
1) Optimizers: We considered and varied three most
general optimization algorithms: ADAM, SGD and
RMSPROP. The optimization algorithms generally
affect the training speed and the final predictive
performance of deep learning models. We per-
formed an experiment in which all the optimization
150392 VOLUME 8, 2020

TABLE 6. Table showing the impact of varying the Hyper-Parameters on the Performance of the Proposed System.
hyper-parameters were varied for each optimizer. The
performance of these optimization algorithms were
compared as shown in Table 6 after tuning their respec-
tive hyper-parameters with the Adam optimizer pro-
duced the best performance followed by SGD and then
RMSprop.
2) Learning Rate: We experimented between three
values of learning rates: 0.01, 0.001, and 0.0001.
We achieved the best system performance with a small
learning rate value of 0.0001. Model training tends
to diverge when the learning rates become too large.
A decreased learning rate yielded an improved gener-
alization accuracy for the proposed model.
3) Weight Decay Constant: We experimented between
three values of weight decay constants: 0.01, 0.001 and
0.0001. The weight decay value of 0.01 eventually
VOLUME 8, 2020 150393

produced the best system performance with Adam opti-
mizer. This can be due to the nature of our dataset and
the architecture which are not too complex.
4) Dropout rate: We also utilized dropout as a regulariza-
tion technique to avoid over-fitting and increase the val-
idation accuracy and thus increasing the generalizing
power. We experimented between 0.5 and 0.25 values.
The value 1.0 means no dropout, and 0.0 means no
outputs from the layer.
5) Number of Dense Layers: In order to reduce the com-
plexity of the architecture, we experimented between
dense layers number 4 and 6 number of dense layers
for the system architecture. We achieved optimal per-
formance at level 6.
6) Batch size: Recent empirical research [81] has shown
that increasing the batch size also affects the training
speed. In this research, the batch size of 128 produced
the best results.
Finalizing Hyper-parameters Setting: The system
achieved the best performance by setting the hyper-parameters
as stated below and in Table 6:
1) Optimizers: Adam
2) Learning Rate: 0.0001
3) Weight Decay Values: 0.001
4) Dense Layers Level: 6
With these settings, we were able to achieve the optimal
performance for the proposed system.
Lastly, it can be stated that the proposed light-weight clas-
sification system achieved better performance with reduced
computing resources that can meet up with the requirement
in the real-time clinical practice. The system performed better
than most existing methods and can meet up with real time
medical diagnosis task in diagnosing skin cancer with the
processing time for each dermoscopy image at averagely 8s.
The performance evaluation of our model was done under the
same hardware conditions and the same dataset with some
state-of-the-arts as shown in Table 7 and the comparison
result shows that the proposed system outperforms these
existing techniques in the computational speed during both
the training and testing phases.
TABLE 7. Measurements in seconds of the training time per epoch and
test time per skin lesion image.
V. CONCLUSION
This work provides some novel approaches using deep learn-
ing techniques in the segmentation and classification method-
ologies of skin lesion images towards detection and diagnosis
of skin cancer. A deep learning-based CAD framework that
is composed of a multi-scale encoder-decoder segmentation
network and an FCN-based DenseNet classification network,
has been proposed for the detection and classification of skin
lesion images to diagnose skin cancer diseases. The proposed
method was evaluated on publicly available database of HAM
10000 that is made up of 7 important diagnostic categories of
skin lesion and it has shown superior performance than the
existing state-of-the-art methods in most of the classification
performance evaluation; and most especially in both the seg-
mentation and classification accuracies. The system includes
a segmentation stage which employs a novel encoder-decoder
network that is integrated into CRF module for accurate and
refined lesion border detection. It also includes an FCN-based
DenseNet framework for an efficient classification of skin
lesions. This was shown to outperform existing state-of-the-
arts classification techniques. It was established that introduc-
ing the multi-scale encoder-decoder segmentation network
into the classification system will improve the classification
accuracy of the entire system. The classification system was
evaluated separately on unsegmented images to show the
effect of the segmentation network. It can be concluded from
our results that application of efficient pre-processing and
segmentation techniques on skin lesion images before clas-
sification can lead to better detection performance of deep
learning-based classification system. The proposed system
has been able to overcome the challenges of dealing with the
complex features of skin lesion images and heavy parameter
tuning of the traditional CNN.
REFERENCES
[1] A. R. D. Delbridge, L. J. Valente, and A. Strasser, ‘‘The role of the apoptotic
machinery in tumor suppression,’’ Cold Spring Harbor Perspect. Biol.,
vol. 4, no. 11, Nov. 2012, Art. no. a008789
[2] M. A. Albahar, ‘‘Skin lesion classification using convolutional neural
network with novel regularizer,’’ IEEE Access, vol. 7, pp. 38306–38313,
2019.
[3] R. Marks, ‘‘An overview of skin cancers,’’ Cancer, vol. 75, no. S2,
pp. 607–612, Jan. 1995.
[4] H. W. Rogers, M. A. Weinstock, S. R. Feldman, and B. M. Coldiron, ‘‘Inci-
dence estimate of nonmelanoma skin cancer (Keratinocyte Carcinomas) in
the US population, 2012,’’ JAMA Dermatology, vol. 151, no. 10, p. 1081,
Oct. 2015.
[5] Cancer Facts and Figures 2017. Accessed: Jan. 20, 2020. [Online].
Available: https://www.cancer.org/research/cancer-facts-statistics/all-
cancer-facts-figures.html
[6] R. L. Siegel, K. D. Miller, and A. Jemal, ‘‘Cancer statistics,’’ CA, A cancer
J. Clinicians, vol. 65, no. 1, pp. 5–29, 2015.
[7] S. A. Gandhi and J. Kampp, ‘‘Skin cancer epidemiology, detection, and
management,’’ Med. Clinics North Amer., vol. 99, no. 6, pp. 1323–1335,
Nov. 2015.
[8] W. E. Damsky and M. Bosenberg, ‘‘Melanocytic nevi and melanoma:
Unraveling a complex relationship,’’ Oncogene, vol. 36, no. 42,
pp. 5771–5792, Oct. 2017.
[9] N. Eisemann, A. Waldmann, A. C. Geller, M. A. Weinstock, B. Volkmer,
R. Greinert, E. W. Breitbart, and A. Katalinic, ‘‘Non-melanoma skin cancer
incidence and impact of skin cancer screening on incidence,’’ J. Investiga-
tive Dermatology, vol. 134, no. 1, pp. 43–50, Jan. 2014.
[10] M. Thorn, F. Ponte, R. Bergstrom, P. Sparen, and H.-O. Adami, ‘‘Clini-
cal and histopathologic predictors of survival in patients with malignant
melanoma: A population-based study in sweden,’’ JNCI J. Nat. Cancer
Inst., vol. 86, no. 10, pp. 761–769, May 1994.
[11] A. Esteva, B. Kuprel, R. A. Novoa, J. Ko, S. M. Swetter, H. M. Blau,
and S. Thrun, ‘‘Dermatologist-level classification of skin cancer with deep
neural networks,’’ Nature, vol. 542, no. 7639, pp. 115–118, Feb. 2017.
[12] H. Kittler, H. Pehamberger, K. Wolff, and M. Binder, ‘‘Diagnostic accuracy
of dermoscopy,’’ Lancet Oncol., vol. 3, no. 3, pp. 159–165, Mar. 2002.
150394 VOLUME 8, 2020

[13] G. Argenziano, H. P. Soyer, S. Chimenti, R. Talamini, R. Corona, F. Sera,
M. Binder, ‘‘Dermoscopy of pigmented skin lesions: Results of a con-
sensus meeting via the Internet,’’ J. Amer. Acad. Dermatology, vol. 5,
pp. 679–693, May 2003.
[14] M. Binder, M. Schwarz, A. Winkler, A. Steiner, A. Kaider, K. Wolff,
and H. Pehamberger, ‘‘Epiluminescence microscopy. A useful tool for the
diagnosis of pigmented skin lesions for formally trained dermatologists,’’
Arch. Dermatology, vol. 131, no. 3, pp. 286–291, Mar. 1995.
[15] J. Gachon, P. Beaulieu, J. F. Sei, J. Gouvernet, J. P. Claudel, M. Lemaitre,
M. A. Richard, and J. J. Grob, ‘‘First prospective study of the recognition
process of melanoma in dermatological practice,’’ Arch. Dermatology,
vol. 141, no. 4, pp. 434–438, Apr. 2005.
[16] A. R. Ratul, M. H. Mozaffari, W.-S. Lee, and E. Parimbelli, ‘‘Skin lesions
classification using deep learning based on dilated convolution,’’ BioRxiv,
Jan. 2019, Art. no. 860700.
[17] V. Revathi and A. Chithra, ‘‘A review on segmentation techniques in
skin lesion images,’’ Intl Res. J. Eng. Tech. (IRJET), vol. 2, no. 9,
pp. 2598–2603, 2015.
[18] Q. Abbas, I. F. Garcia, M. Emre Celebi, W. Ahmad, and Q. Mushtaq,
‘‘A perceptually oriented method for contrast enhancement and seg-
mentation of dermoscopy images,’’ Skin Res. Technol., vol. 19, no. 1,
pp. e490–e497, Feb. 2013.
[19] A. A. Adegun and S. Viriri, ‘‘Deep learning-based system for automatic
melanoma detection,’’ IEEE Access, vol. 8, pp. 7160–7172, 2020.
[20] H. Alquran, I. A. Qasmieh, A. M. Alqudah, S. Alhammouri, E. Alawneh,
A. Abughazaleh, and F. Hasayen, ‘‘The melanoma skin cancer detection
and classification using support vector machine,’’ in Proc. IEEE Jordan
Conf. Appl. Electr. Eng. Comput. Technol. (AEECT), Oct. 2017, pp. 1–5.
[21] K. Hoffmann et al., ‘‘Diagnostic and neural analysis of skin cancer
(DANAOS). A multicentre study for collection and computer-aided anal-
ysis of data from pigmented skin lesions using digital dermoscopy,’’ Brit.
J. Dermatology, vol. 149, no. 4, pp. 801–809, Oct. 2003.
[22] N. Hameed, F. Hameed, A. Shabut, S. Khan, S. Cirstea, and A. Hossain,
‘‘An intelligent computer-aided scheme for classifying multiple skin
lesions,’’ Computers, vol. 8, no. 3, p. 62, Aug. 2019.
[23] A. Murugan, S. A. H. Nair, and K. P. S. Kumar, ‘‘Detection of skin cancer
using SVM, random forest and kNN classifiers,’’ J. Med. Syst., vol. 43,
no. 8, p. 269, Aug. 2019.
[24] R. D. Seeja and A. Suresh, ‘‘Deep learning based skin lesion segmentation
and classification of melanoma using support vector machine (SVM),’’
Asian Pacific J. Cancer Prevention, vol. 20, no. 5, pp. 1555–1561,
May 2019.
[25] Y. Li and L. Shen, ‘‘Skin lesion analysis towards melanoma detection using
deep learning network,’’ Sensors, vol. 18, no. 2, p. 556, Feb. 2018.
[26] S. M. Rajpara, A. P. Botello, J. Townend, and A. D. Ormerod, ‘‘System-
atic review of dermoscopy and digital dermoscopy/ artificial intelligence
for the diagnosis of melanoma,’’ Brit. J. Dermatology, vol. 161, no. 3,
pp. 591–604, Sep. 2009.
[27] A. Hekler, J. S. Utikal, A. H. Enk, A. Hauschild, M. Weichenthal,
R. C. Maron, C. Berking, S. Haferkamp, J. Klode, D. Schadendorf,
B. Schilling, T. Holland-Letz, B. Izar, C. Von Kalle, S. Fröhling, and
T. J. Brinker, ‘‘Superior skin cancer classification by the combination of
human and artificial intelligence,’’ Eur. J. Cancer, vol. 120, pp. 114–121,
Oct. 2019.
[28] T. J. Brinker, A. Hekler, J. S. Utikal, N. Grabe, D. Schadendorf, J. Klode,
C. Berking, T. Steeb, A. H. Enk, and C. von Kalle, ‘‘Skin cancer classifi-
cation using convolutional neural networks: Systematic review,’’ J. Med.
Internet Res., vol. 20, no. 10, Oct. 2018, Art. no. e11936.
[29] S. R. Guha and S. R. Haque, ‘‘Performance comparison of machine
learning-based classification of skin diseases from skin lesion images,’’ in
Proc. Int. Conf. Commun., Comput. Electron. Syst., Singapore: Springer,
2020, pp. 15–25.
[30] L. Bi, D. Feng, and J. Kim, ‘‘Dual-path adversarial learning for fully
convolutional network (FCN)-based medical image segmentation,’’ Vis.
Comput., vol. 34, nos. 6–8, pp. 1043–1052, Jun. 2018.
[31] L. Bi, J. Kim, E. Ahn, A. Kumar, M. Fulham, and D. Feng, ‘‘Dermoscopic
image segmentation via multistage fully convolutional networks,’’ IEEE
Trans. Biomed. Eng., vol. 64, no. 9, pp. 2065–2074, Sep. 2017.
[32] B. Abdollahi, N. Tomita, and S. Hassanpour, ‘‘Data Augmentation in
Training Deep Learning Models for Medical Image Analysis,’’ in Deep
Learners and Deep Learner Descriptors for Medical Applications. Cham,
Switzerland: Springer, 2020, pp. 167–180.
[33] K. Raza and N. Kumar Singh, ‘‘A tour of unsupervised deep learning for
medical image analysis,’’ 2018, arXiv:1812.07715. [Online]. Available:
http://arxiv.org/abs/1812.07715
[34] A. Voulodimos, N. Doulamis, A. Doulamis, and E. Protopapadakis, ‘‘Deep
learning for computer vision: A brief review,’’ Comput. Intell. Neurosci.,
vol. 2018, pp. 1–13, Feb. 2018.
[35] L. Bi, J. Kim, E. Ahn, A. Kumar, D. Feng, and M. Fulham, ‘‘Step-
wise integration of deep class-specific learning for dermoscopic image
segmentation,’’ Pattern Recognit., vol. 85, pp. 78–89, Jan. 2019.
[36] H. Lee, R. Grosse, R. Ranganath, and A. Y. Ng, ‘‘Unsupervised learning
of hierarchical representations with convolutional deep belief networks,’’
Commun. ACM, vol. 54, no. 10, pp. 95–103, Oct. 2011.
[37] S. Pereira, R. Meier, R. McKinley, R. Wiest, V. Alves, C. A. Silva, and
M. Reyes, ‘‘Enhancing interpretability of automatically extracted machine
learning features: Application to a RBM-random forest system on brain
lesion segmentation,’’ Med. Image Anal., vol. 44, pp. 228–244, Feb. 2018.
[38] M. Akhavan Aghdam, A. Sharifi, and M. M. Pedram, ‘‘Combination
of rs-fMRI and sMRI data to discriminate autism spectrum disorders in
young children using deep belief network,’’ J. Digit. Imag., vol. 31, no. 6,
pp. 895–903, Dec. 2018.
[39] A. Al Nahid, A. Mikaelian, and Y. Kong, ‘‘Histopathological breast-image
classification with restricted Boltzmann machine along with backpropaga-
tion,’’ Biomed. Res., vol. 29, no. 10, pp. 2068–2077, 2018.
[40] Y. Zhu, L. Wang, M. Liu, C. Qian, A. Yousuf, A. Oto, and D. Shen,
‘‘MRI-based prostate cancer detection with high-level representation and
hierarchical classification,’’ Med. Phys., vol. 44, no. 3, pp. 1028–1039,
Mar. 2017.
[41] S. Minaee, Y. Wang, A. Aygar, S. Chung, X. Wang, Y. W. Lui, E. Fieremans,
S. Flanagan, and J. Rath, ‘‘MTBI identification from diffusion MR images
using bag of adversarial visual features,’’ IEEE Trans. Med. Imag., vol. 38,
no. 11, pp. 2545–2555, Nov. 2019.
[42] V. M. Vergara, A. R. Mayer, E. Damaraju, K. A. Kiehl, and V. Calhoun,
‘‘Detection of mild traumatic brain injury by machine learning clas-
sification using resting state functional network connectivity and frac-
tional anisotropy,’’ J. Neurotrauma, vol. 34, no. 5, pp. 1045–1053,
Mar. 2017.
[43] A.-R. Ali, J. Li, and T. Trappenberg, ‘‘Supervised versus unsupervised
deep learning based methods for skin lesion segmentation in dermoscopy
images,’’ in Proc. Can. Conf. Artif. Intell., Cham, Switzerland: Springer,
2019, pp. 373–379.
[44] S. Minaee, Y. Boykov, F. Porikli, A. Plaza, N. Kehtarnavaz, and
D. Terzopoulos, ‘‘Image segmentation using deep learning: A survey,’’
2020, arXiv:2001.05566. [Online]. Available: http://arxiv.org/abs/2001.
05566
[45] S. Feng, Z. Zhuo, D. Pan, and Q. Tian, ‘‘CcNet: A cross-connected convo-
lutional network for segmenting retinal vessels using multi-scale features,’’
Neurocomputing, vol. 392, pp. 268–276, Jun. 2020.
[46] J. Wang, K. Sun, T. Cheng, B. Jiang, C. Deng, Y. Zhao, D. Liu, Y. Mu,
M. Tan, X. Wang, W. Liu, and B. Xiao, ‘‘Deep high-resolution represen-
tation learning for visual recognition,’’ IEEE Trans. Pattern Anal. Mach.
Intell., early access, Apr. 1, 2020, doi: 10.1109/TPAMI.2020.2983686.
[47] S. Vesal, S. M. Patil, N. Ravikumar, and A. K. Maier, ‘‘A multi-task
framework for skin lesion detection and segmentation,’’ in OR 2.0 Context-
Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clin-
ical Image-Based Procedures, and Skin Image Analysis. Cham, Switzer-
land: Springer, 2018, pp. 285–293.
[48] N. Hameed, A. M. Shabut, M. K. Ghosh, and M. A. Hossain, ‘‘Multi-class
multi-level classification algorithm for skin lesions classification using
machine learning techniques,’’ Expert Syst. Appl., vol. 141, Mar. 2020,
Art. no. 112961.
[49] A.-R. Ali, J. Li, G. Yang, and S. J. O’Shea, ‘‘A machine learning approach
to automatic detection of irregularity in skin lesion border using dermo-
scopic images,’’ PeerJ Comput. Sci., vol. 6, Jun. 2020, Art. no. e268.
[50] X. He, X. Yang, S. Zhang, J. Zhao, Y. Zhang, E. Xing, and P. Xie, ‘‘Sample-
efficient deep learning for COVID-19 diagnosis based on CT scans,’’
MedRxiv, vol. 7, Jan. 2020.
[51] H. El-Khatib, D. Popescu, and L. Ichim, ‘‘Deep learning–based methods
for automatic diagnosis of skin lesions,’’ Sensors, vol. 20, no. 6, p. 1753,
Mar. 2020.
[52] N. Hameed, A. M. Shabut, and M. A. Hossain, ‘‘Multi-class skin diseases
classification using deep convolutional neural network and support vector
machine,’’ in Proc. 12th Int. Conf. Softw., Knowl., Inf. Manage. Appl.
(SKIMA), Dec. 2018, pp. 1–7.
VOLUME 8, 2020 150395

ADEGUNADEGUNADEGUNADEGUNADEGUNADEGUNADEGUN.pdf

ADEGUNADEGUNADEGUNADEGUNADEGUNADEGUNADEGUN.pdf

Related slideshows

More Related Content

Similar to ADEGUNADEGUNADEGUNADEGUNADEGUNADEGUNADEGUN.pdf

Similar to ADEGUNADEGUNADEGUNADEGUNADEGUNADEGUNADEGUN.pdf (20)

Recently uploaded

Recently uploaded (16)

ADEGUNADEGUNADEGUNADEGUNADEGUNADEGUNADEGUN.pdf