Skin Lesion Detection Research Articles

A hybrid feature fusion strategy for early fusion and majority voting for late fusion towards melanocytic skin lesion detection

AbstractA computer‐aided‐diagnostic system for diagnosing melanoma often uses distinct kinds of features for characterizing the lesions. Extracting distinct features from melanocytic images represents the various characteristics of pigmented lesions. Concatenating such features distinguishes extracted feature's information effectively while eliminating the redundant information amid them, which aids in discriminating the cancerous and noncancerous lesions. This article proposes a framework comprising segmentation, feature extraction, feature fusion, and classification to differentiate benign lesions from melanoma. The proposed framework is four‐fold: beginning with the extraction of ROI from an image, the SLICACO method is used for segmentation. Thereafter, ABCD rule‐based global and local features are extracted for effective melanoma detection. Further, we develop a new hybrid feature fusion strategy, PCAFA, leveraging the benefits of principal component analysis and factor analysis. The method performs early fusion by combining all extracted features within an individual feature vector and fed to a learning model for the prediction. While late fusion is performed using majority voting by combining the outputs of machine learning models. After that, gradient tree boosting, support vector machine, and decision tree models are trained to utilize distinct features of skin lesions for their classification as benign or malignant. The effectiveness of the designed framework is validated on the ISIC2017 benchmark skin lesion dataset based on specificity, sensitivity, and accuracy. The generalizability of the designed framework is gauged by conducting a fair comparison with conventional methods. Evaluated results reveal the potential of the proposed fused feature‐set in discriminating malignant and nonmalignant lesions with an accuracy of 96.8%.

Deep Learning and Machine Learning Techniques of Diagnosis Dermoscopy Images for Early Detection of Skin Diseases

With the increasing incidence of severe skin diseases, such as skin cancer, endoscopic medical imaging has become urgent for revealing the internal and hidden tissues under the skin. Diagnostic information to help doctors make an accurate diagnosis is provided by endoscopy devices. Nonetheless, most skin diseases have similar features, which make it challenging for dermatologists to diagnose patients accurately. Therefore, machine and deep learning techniques can have a critical role in diagnosing dermatoscopy images and in the accurate early detection of skin diseases. In this study, systems for the early detection of skin lesions were developed. The performance of the machine learning and deep learning was evaluated on two datasets (e.g., the International Skin Imaging Collaboration (ISIC 2018) and Pedro Hispano (PH2)). First, the proposed system was based on hybrid features that were extracted by three algorithms: local binary pattern (LBP), gray level co-occurrence matrix (GLCM), and wavelet transform (DWT). Such features were then integrated into a feature vector and classified using artificial neural network (ANN) and feedforward neural network (FFNN) classifiers. The FFNN and ANN classifiers achieved superior results compared to the other methods. Accuracy rates of 95.24% for diagnosing the ISIC 2018 dataset and 97.91% for diagnosing the PH2 dataset were achieved using the FFNN algorithm. Second, convolutional neural networks (CNNs) (e.g., ResNet-50 and AlexNet models) were applied to diagnose skin diseases using the transfer learning method. It was found that the ResNet-50 model fared better than AlexNet. Accuracy rates of 90% for diagnosing the ISIC 2018 dataset and 95.8% for the PH2 dataset were reached using the ResNet-50 model.

A Reinforcement Learning Algorithm for Automated Detection of Skin Lesions

Skin cancers are increasing at an alarming rate, and detection in the early stages is essential for advanced treatment. The current segmentation methods have limited labeling ability to the ground truth images due to the numerous noisy expert annotations present in the datasets. The precise boundary segmentation is essential to correctly locate and diagnose the various skin lesions. In this work, the lesion segmentation method is proposed as a Markov decision process. It is solved by training an agent to segment the region using a deep reinforcement-learning algorithm. Our method is similar to the delineation of a region of interest by the physicians. The agent follows a set of serial actions for the region delineation, and the action space is defined as a set of continuous action parameters. The segmentation model learns in continuous action space using the deep deterministic policy gradient algorithm. The proposed method enables continuous improvement in performance as we proceed from coarse segmentation results to finer results. Finally, our proposed model is evaluated on the International Skin Imaging Collaboration (ISIC) 2017 image dataset, Human against Machine (HAM10000), and PH2 dataset. On the ISIC 2017 dataset, the algorithm achieves an accuracy of 96.33% for the naevus cases, 95.39% for the melanoma cases, and 94.27% for the seborrheic keratosis cases. The other metrics are evaluated on these datasets and rank higher when compared with the current state-of-the-art lesion segmentation algorithms.

Skin Lesion Detection Algorithms in Whole Body Images

Melanoma is one of the most lethal and rapidly growing cancers, causing many deaths each year. This cancer can be treated effectively if it is detected quickly. For this reason, many algorithms and systems have been developed to support automatic or semiautomatic detection of neoplastic skin lesions based on the analysis of optical images of individual moles. Recently, full-body systems have gained attention because they enable the analysis of the patient’s entire body based on a set of photos. This paper presents a prototype of such a system, focusing mainly on assessing the effectiveness of algorithms developed for the detection and segmentation of lesions. Three detection algorithms (and their fusion) were analyzed, one implementing deep learning methods and two classic approaches, using local brightness distribution and a correlation method. For fusion of algorithms, detection sensitivity = 0.95 and precision = 0.94 were obtained. Moreover, the values of the selected geometric parameters of segmented lesions were calculated and compared for all algorithms. The obtained results showed a high accuracy of the evaluated parameters (error of area estimation <10%), especially for lesions with dimensions greater than 3 mm, which are the most suspected of being neoplastic lesions.

Deep ensemble learning for skin lesions classification with convolutional neural network

One type of skin cancer that is considered a malignant tumor is melanoma. Such a dangerous disease can cause a lot of death in the world. The early detection of skin lesions becomes an important task in cancer diagnosis, therefore the patient survival can be improved. Recently, a machine learning paradigm emerged known as deep learning (DL) utilized for skin lesions classification. However, in some previous studies by using seven class images diagnostic of skin lesions classification based on a single DL approach with CNNs architecture not produce a satisfying performance. The DL approach allows the development of a medical image analysis system that can utilize an extraordinary network for improving the performance, such as the deep convolutional neural networks (DCNNs) method. In this study, we propose an ensemble learning approach that combines three DCNNs architectures such as Inception V3, Inception ResNet V2 and DenseNet 201 for improving the performance in terms of accuracy, sensitivity, specificity, precision, and F1-score. Seven classes of dermoscopy image categories of skin lesions is utilized with 10015 dermoscopy images from well known the HAM10000 dataset. As found in the results with proposed model ensemble learning architecture produce good classification performance with 97.23% accuracy, 90.12% sensitivity, 97.73% specificity, 82.01% precision, and 85.01% F1-Score. This method gives promising results in classifying skin lesions for cancer diagnosis.

Deep ensemble learning for skin lesions classification with convolutional neural network

&lt;span lang="EN-US"&gt;One type of skin cancer that is considered a malignant tumor is melanoma. Such a dangerous disease can cause a lot of death in the world. The early detection of skin lesions becomes an important task in the diagnosis of skin cancer. Recently, a machine learning paradigm emerged known as deep learning (DL) utilized for skin lesions classification. However, in some previous studies by using seven class images diagnostic of skin lesions classification based on a single DL approach with CNNs architecture does not produce a satisfying performance. The DL approach allows the development of a medical image analysis system for improving performance, such as the deep convolutional neural networks (DCNNs) method. In this study, we propose an ensemble learning approach that combines three DCNNs architectures such as Inception V3, Inception ResNet V2 and DenseNet 201 for improving the performance in terms of accuracy, sensitivity, specificity, precision, and F1-score. Seven classes of dermoscopy image categories of skin lesions are utilized with 10015 dermoscopy images from well-known the HAM10000 dataset. The proposed model produces good classification performance with 97.23% accuracy, 90.12% sensitivity, 97.73% specificity, 82.01% precision, and 85.01% F1-Score. This method gives promising results in classifying skin lesions for cancer diagnosis.&lt;/span&gt;

Artificial intelligence techniques for enhanced skin lesion detection

Artificial intelligence techniques for enhanced skin lesion detection

A novel total body digital photography smartphone application designed to detect and monitor skin lesions: a pilot study

&lt;p class="Body"&gt;Although some smartphone applications are designed for total body photography (TBP), few offer the specificity that enables self- as well as dermatologist-, detection of new lesions, or change in lesion color or in size as little as 1mm, on an ongoing basis. The aim of this study is to assess the sensitivity of a novel TBP application in the detection of changes to color and size of simulated skin lesions. Twenty-five subjects underwent one study visit. After baseline photography, new artificial markings were made or naturally occurring pigmented lesions located in any anatomical region were enhanced/enlarged, and a second matching set of photographs was then taken. From all 25 subjects, a total of 262 skin markings were evaluable. Of these, 241 (92%) were detected by the app, which resulted in an overall sensitivity of 92%. The high sensitivity establishes the app as capable of providing reliable self-TBP that allows detection and monitoring of new skin lesions or change in both size and color. This method greatly enhances the ability to accomplish ongoing self-monitoring and yet provides quality informing images to the dermatologist to assist in decision-making with the patient. &lt;/p&gt;

Skin Lesion Detection and Classification Using Convolutional Neural Network for Deep Feature Extraction and Support Vector Machine

Pigmented skin lesion identification is essential for detecting harmful pathologies related to this large organ, especially cancer. An analysis of the different methods and projects developed to diagnose these illnesses throughout the years showed that they had become very useful tools to identify melanoma, dermatofibroma, and basal cell carcinoma, among other types of cancer, are seen through the use of new computer-aided technologies. The most common diagnosis is based on dermoscopy and the dermatologist expertise that can improve accuracy with image detection techniques and classification by computer. Therefore, this study aims to develop software models able to detect and classify skin cancer. The following work is based on the use of dermoscopy images obtained from the HAM10000 dataset, a database with 10000 images previously tested and validated for research use. The main process is divided into three relevant parts: image segmentation, feature extraction (FE) using ten different pre-trained Convolutional Neural Networks (CNNs), and Support Vector Machine (SVM) to establish a classification model. According to the results, the models of classification performed very well using the image segmentation step, showing average accuracies between 80.67% (Xception) and 90% (Alexnet). In contrast to the process without using image segmentation, where no method reached 60%. AlexNet plus SVM model showed the minor running time and presented the higher accuracy rate (90.34%) for the correct identification and classification of the seven categories of cutaneous lesions taken into account.

Skin Lesions Detection and Classification Using Deep Learning

The skin lesion is considered the most widespread malignant disease in individuals, and melanoma is the deadliest form of the disease. Early detection affects the prognosis of the disease and improves the chances of survival. Dermatologists use scientific calculation tools such as ABCD to diagnose melanoma through visual inspection of the mole. However, computer vision tools have been presented to support the quantitative scrutiny of skin lesions. Significant improvements to deep learning algorithms in image recognition tasks should be very successful in medical image examination, particularly in the classification of skin lesions used to diagnose melanoma. In this research, a deep learning simulation with 38 layers to detect and classify skin lesions was proposed. Two datasets were used for training and testing i.e., the HAM10000 dataset &amp; the ISIC2019 dataset. Experimental results show that the model outperforms on both the datasets hence making it nondependent of the dataset. 94.45% of validation top 3 accuracies are achieved on the HAM10000 dataset &amp; 93.06% of validation top 3 accuracies are achieved on the ISIC2019 dataset.

Dermo-DOCTOR: A framework for concurrent skin lesion detection and recognition using a deep convolutional neural network with end-to-end dual encoders

Background and objectiveAutomated skin lesion analysis for simultaneous detection and recognition is still challenging for inter-class homogeneity and intra-class heterogeneity, leading to low generic capability of a single convolutional neural network (CNN) with limited datasets. MethodsThis article proposes an end-to-end deep CNN-based framework for simultaneous detection and recognition of the skin lesions, named Dermo-DOCTOR, consisting of two encoders. The feature maps from two encoders are fused channel-wise, called Fused Feature Map (FFM). The FFM is utilized for decoding in the detection sub-network, concatenating each stage of two encoders’ outputs with corresponding decoder layers to retrieve the lost spatial information due to pooling in the encoders. For the recognition sub-network, the outputs of three fully connected layers, utilizing feature maps of two encoders and FFM, are aggregated to obtain a final lesion class. We train and evaluate the proposed Dermo-Doctor utilizing two publicly available benchmark datasets, such as ISIC-2016 and ISIC-2017. ResultsThe achieved segmentation results exhibit mean intersection over unions of 85.0% and 80.0% respectively for ISIC-2016 and ISIC-2017 test datasets. The proposed Dermo-DOCTOR also demonstrates praiseworthy success in lesion recognition, providing the areas under the receiver operating characteristic curves of 0.98 and 0.91 respectively for those two datasets. The experimental results show that the proposed Dermo-DOCTOR outperforms the alternative methods mentioned in the literature, designed for skin lesion detection and recognition. ConclusionAs the Dermo-DOCTOR provides better results on two different test datasets, even with limited training data, it can be an auspicious computer-aided assistive tool for dermatologists.

The development of skin lesion detection application in smart handheld devices using deep neural networks

The development of skin lesion detection application in smart handheld devices using deep neural networks

Access to a dermatoscope during dermatology courses motivates students’ towards thorough skin examination

Introduction. Dermatoscope is a tool for a skin examination, used especially in early detection of malignant skin lesions. Non-dermatologists are being trained for opportunistic melanoma detection with the usage of dermatoscopy, however, still non-satisfactory. This study was aimed to determine whether practical dermoscopy adjunct to traditional, lecture and seminar-based medical school curriculum would improve the perceived relevance of regular skin examination and basic skin lesions differentiation. Material and method. Fourth-year medical students participating in a 3-week-long dermatology course were randomly assigned to two groups: the first one called A with limited access to a dermatoscope and the second one called B, with unlimited access to dermatoscopes throughout the course. All participants answered surveys concerning their attitude towards skin examination, with a rating scale from 1 to 5, before and after the course. Also, all participants completed an image-based dermoscopy test for distinguishing benign from malignant skin lesions. Results. Students assigned to group B significantly improved their perceived importance of routine skin examination (mean scores before 4.38; after 4.57, P = 0.03). No such tendency was observed in group A — before 4.40, after 4.49 (P = 0.29). Students in the group with higher dermatoscope availability considered buying a dermatoscope more often (61%) than those without (44%) (P = 0.037). No significant score difference was observed when testing skin lesions identification, mean for extended access 7.84 vs. normal 7.64 points (P = 0.69). Conclusions. Higher availability of dermatoscopy during dermatology rotations may encourage students to use this tool in future clinical practice and improve early detection of malignant skin lesions.

A Dermoscopic Skin Lesion Classification Technique Using YOLO-CNN and Traditional Feature Model

Skin cancer is one of the most deadly diseases around the world, wherein one of the three cancers is skin cancer. Early detection of skin cancer is paramount for better treatment planning. This paper investigates a Convolutional Neural Network (CNN), specifically, You Only Look Once (YOLO), to extract features from the skin lesions. The features, obtained from the CNN, are concatenated with traditional features like texture and colour features extracted from the lesion region of the input images. Later, the concatenated features are fed to a Fully Connected Network, which is trained with the specific ground truths to achieve higher classification accuracy. The proposed method improves the detection and classification of skin lesions when compared with other models and YOLO without traditional features. The performance measures of the fusion network are able to achieve the accuracy of 94%, precision of 0.85, recall of 0.88, and area under the curve of 0.95.

Ultraviolet Fluorescence as a Field-Applicable Screening Tool for Lesions Consistent with Ophidiomycosis in Lake Erie Watersnakes (Nerodia sipedon insularum).

Ophidiomycosis, commonly called snake fungal disease, has been linked to significant morbidity of free-ranging snakes in North America and Europe. Diagnosis of ophidiomycosis currently requires detection of skin lesions via physical exam or characteristic histopathology as well as detection of the causative agent, Ophidiomyces ophidiicola, through quantitative (q)PCR or fungal culture of a skin swab or tissue sample. While reliable, these methods require specialized training, invasive procedures (e.g., biopsy), and several days or weeks to receive results. Additionally, screening entire populations can quickly become costly. A fast, easy-to-use, cost-efficient, and sensitive screening tool is needed to optimize conservation strategies and treatment intervention. Our objective was to investigate the association between skin fluorescence under long-wave ultraviolet (UV) light (365 nm) and the detection of Ophidiomyces ophidiicola DNA using qPCR. Fifty-eight Lake Erie watersnakes (Nerodia sipedon insularum) collected in June of 2018 and 2019 from islands in western Lake Erie, Ottawa County, Ohio, US were visually inspected for skin lesions, photographed under natural light and UV light, and swabbed for qPCR analysis. Fluorescence was highly associated with the presence of skin lesions, and the presence of at least one fluorescent skin lesion was 86% sensitive and 100% specific for identifying animals with apparent ophidiomycosis, with a positive predictive value of 100%. While we recommend performing standard diagnostics along with fluorescence, our study supports the use of visual UV fluorescence identification as a preliminary, affordable, noninvasive, and field-applicable method to screen populations for ophidiomycosis.

A Prior-mask-guided Few-shot Learning for Skin Lesion Segmentation

The incidence of skin cancer, which has high mortality, is growing rapidly worldwide. Early detection of skin lesions is crucial for timely diagnosis and treatment to improve the patient survival rate. Computer vision technology based on deep convolutional neural network requires a large amount of labelled data. The cost of data acquisition and annotation is relatively high, especially for skin cancer segmentation tasks. Therefore, we propose a few-shot segmentation network for skin lesion segmentation, which requires only a few pixel-level annotations. First, the co-occurrence region between the support image and query image is obtained, which is used as a prior mask to exclude irrelevant background regions. Second, the results are concatenated and sent to the inference module to predict segmentation of the query image. Third, the proposed network is retrained by reversing the support and query role, which benefits from the symmetrical structure. Extensive experiments performed on ISIC-2017, ISIC-2019, and PH2 demonstrate that our method forms a promising framework for few-shot segmentation of skin lesion.

Using deep learning for dermatologist-level detection of suspicious pigmented skin lesions from wide-field images.

A reported 96,480 people were diagnosed with melanoma in the United States in 2019, leading to 7230 reported deaths. Early-stage identification of suspicious pigmented lesions (SPLs) in primary care settings can lead to improved melanoma prognosis and a possible 20-fold reduction in treatment cost. Despite this clinical and economic value, efficient tools for SPL detection are mostly absent. To bridge this gap, we developed an SPL analysis system for wide-field images using deep convolutional neural networks (DCNNs) and applied it to a 38,283 dermatological dataset collected from 133 patients and publicly available images. These images were obtained from a variety of consumer-grade cameras (15,244 nondermoscopy) and classified by three board-certified dermatologists. Our system achieved more than 90.3% sensitivity (95% confidence interval, 90 to 90.6) and 89.9% specificity (89.6 to 90.2%) in distinguishing SPLs from nonsuspicious lesions, skin, and complex backgrounds, avoiding the need for cumbersome individual lesion imaging. We also present a new method to extract intrapatient lesion saliency (ugly duckling criteria) on the basis of DCNN features from detected lesions. This saliency ranking was validated against three board-certified dermatologists using a set of 135 individual wide-field images from 68 dermatological patients not included in the DCNN training set, exhibiting 82.96% (67.88 to 88.26%) agreement with at least one of the top three lesions in the dermatological consensus ranking. This method could allow for rapid and accurate assessments of pigmented lesion suspiciousness within a primary care visit and could enable improved patient triaging, utilization of resources, and earlier treatment of melanoma.

Extraction of optical properties from a turbid medium using fiber probe for spectral and spatial diffuse reflectance measurement

Diffuse reflectance technique is popular in the study of tissue physiology through the change in optical properties in a noninvasive manner. Diffuse reflected light intensity is commonly collected either from a single distance with spectral measurement or from a single wavelength with different spatial distances. Improving existing systems is necessary in order to obtain information from greater depths and in smaller volumes. In this paper, we propose a fast and compact fiber probe-based diffuse reflectance method for combining the spectral measurements in the range of 400–950 nm and spatial information up to 1.33 mm from the illumination source. First, we chose the most appropriate analysis model for the proposed distances between the fiber probe and tested it on solid phantoms with varying scattering and absorption components. The measurements are compared to the scattering coefficients according to Mie theory and the absorption according to spectrophotometer measurements. Next, we measured two-layer phantoms with constant scattering and absorption contrast in different layer thicknesses. We extracted the penetration depth from the measured effective absorption coefficient. In the near-IR, we were able to detect the absorption coefficient of the bottom phantom layer behind a top layer of up to 5 mm. We achieved a maximum penetration depth of 5.7 mm for 5 mm top layer thickness at 900 nm. Our fiber probe diffuse reflectance system can be used in the near future for skin lesion detection in clinical studies.

Skin Lesion Detection Using Fuzzy Approach and Classification with CNN

Skin Lesion Detection Using Fuzzy Approach and Classification with CNN

Attributes based skin lesion detection and recognition: A mask RCNN and transfer learning-based deep learning framework

Abstract Malignant melanoma is considered to be one of the deadliest types of skin cancers which is responsible for the massive number of deaths worldwide. According to the American Cancer Society (ACS), more than a million Americans are living with this melanoma. Since 2019, 192,310 new cases of melanoma are registered, where 95,380 are noninvasive, and 96,480 are invasive. The numbers of deaths due to melanoma in 2019 alone are 7,230, comprising 4,740 men and 2,490 women. Melanoma may be curable if diagnosed at the earlier stages; however, the manual diagnosis is time-consuming and also dependent on the expert dermatologist. In this work, a fully automated computerized aided diagnosis (CAD) system is proposed based on the deep learning framework. In the proposed scheme, the original dermoscopic images are initially pre-processed using the decorrelation formulation technique, which later passes the resultant images to the MASK-RCNN for the lesion segmentation. In this step, the MASK RCNN model is trained using the segmented RGB images generated from the ground truth images of ISBI2016 and ISIC2017 datasets. The resultant segmented images are later passed to the DenseNet deep model for feature extraction. Two different layers, average pool and fully connected, are used for feature extraction, which are later combined, and the resultant vector is forwarded to the feature selection block for down - sampling using proposed entropy-controlled least square SVM (LS-SVM). Three datasets are utilized for validation - ISBI2016, ISBI2017, and HAM10000 to achieve an accuracy of 96.3%, 94.8%, and 88.5% respectively. Further, the performance of MASK-RCNN is also validated on ISBI2016 and ISBI2017 to attain an accuracy of 93.6% and 92.7%. To further increase our confidence in the proposed framework, a fair comparison with other state-of-the-art is also provided.