Automated diagnosis of pigmented skin lesions

Machine Learning and Deep Learning based AI Tools for Development of Diagnostic Tools

Artificial intelligence (AI) systems exhibit human-like intelligence. Human intelligence is converted to machines or computer technologies using AI algorithms. Machine learning (ML) is a subset of AI that can learn from extracted data and models to perform a task whereas deep learning (DL) is a subset of ML that imitates the human brain functioning to solve real-world problems in almost all fields. AI caused a paradigm shift in healthcare that can be employed for decision support and forecasting. Medical diagnostic tools developed using AI, perform disease diagnosis based on the symbolic models of disease and provide therapy recommendations. The key AI applications employed with medical diagnosis are characterized as learning systems and expert systems. Diagnostic tools, developed using Expert systems utilize facts, implications, and knowledge processing techniques for disease diagnosis, whereas a learning system utilizes statistical pattern recognition, ML, and neural networks. In March 2020, an infectious disease caused by the severe acute respiratory syndrome-Coronavirus-2 (SARS-CoV-2) virus, Coronavirus disease-2019 (COVID-19) was declared a pandemic by the World Health Organization. Recent research studies have shown that AI, ML, and DL can be leveraged to combat COVID-19 having objectives of disease diagnosis, to forecast epidemic and sustainable development, and so on. DL algorithms are implemented on image data, more specifically on chest X-rays and computed tomography scans, for developing diagnostic tools. In this chapter, various ML and DL-based AI tools for the development of diagnostic tools have been discussed.

ZooME: Efficient Melanoma Detection Using Zoom-in Attention and Metadata Embedding Deep Neural Network

Melanoma detection is a crucial yet hard task for both dermatologists and computer-aided diagnosis (CAD). Many traditional machine learning algorithms including deep learning-based methods are employed for melanoma classification. However, more and more complex network architectures do not harvest a leap in model performance. In this paper, we aim to enhance the credibility of CAD approach for melanoma by paying more attention to clinically important information. We propose a Zoom-in Attention and Metadata Embedding (ZooME) melanoma detection network by: 1) introducing a Zoom-in Attention model to better extract and utilize unique pathological information of dermoscopy images; 2) embedding patients' demographic information including age, gender, and anatomic body site, to provide well-rounded information for better prediction. We apply a ten-fold cross-validation on the latest ISIC-2020 dataset with 33,126 dermoscopy images. The proposed ZooME achieved state-of-the-art results with 92.23% in AUC score, 84.59% in accuracy, 85.95% in sensitivity, and 84.63% in specialty, respectively.

Machine Learning and Deep Learning Methods for Skin Lesion Classification and Diagnosis: A Systematic Review

Computer-aided systems for skin lesion diagnosis is a growing area of research. Recently, researchers have shown an increasing interest in developing computer-aided diagnosis systems. This paper aims to review, synthesize and evaluate the quality of evidence for the diagnostic accuracy of computer-aided systems. This study discusses the papers published in the last five years in ScienceDirect, IEEE, and SpringerLink databases. It includes 53 articles using traditional machine learning methods and 49 articles using deep learning methods. The studies are compared based on their contributions, the methods used and the achieved results. The work identified the main challenges of evaluating skin lesion segmentation and classification methods such as small datasets, ad hoc image selection and racial bias.

Analysis of lesional color to differentiate infantile hemangiomas from port-wine birthmarks in infants less than 3 months old: A pilot study

BACKGROUND

In their early phase, infantile hemangiomas (IH) can sometimes be difficult to differentiate from port-wine birthmarks (PWB). Until recently, inexpensive diagnostic tools have not been readily available.

OBJECTIVE

To determine the diagnostic utility of widely available colorimetric technology when differentiating PWB from IH in photographs of infants less than 3 months old.

METHODS

Multi-center, retrospective analysis of RGB (red, green, and blue) and HSL (hue, saturation, lightness) values collected using electronic colorimeters from images of clinically confirmed untreated IH or PWB. Subgroup analysis of flat vascular birthmarks was subsequently performed.

RESULTS

Images of 119 IH (specifically, 45 flat IH) and 59 PWB were identified. PWB had significantly (P < .001) higher RGB values of all primary colors, most notably for blue and green (mean difference: >50), irrespective of thickness. RGB or RGB with HSL values had an excellent accuracy (90%), sensitivity (92%), specificity (98%), and positive predictive value (98%) when discriminating PWB from flat IH. IH could be distinctly clustered from PWB when combining their RGB with HSL values.

CONCLUSION

Electronic colorimeters with emphasis on blue and green values, are able to differentiate PWB from IH, irrespective of thickness, with a high degree of accuracy. A larger scale evaluation is now required.

An Integrated Platform for Skin Cancer Heterogenous and Multilayered Data Management

Electronic health record (EHR) systems improve health care services by allowing the combination of health data with clinical decision support features and clinical image analyses. This study presents a modular and distributed platform that is able to integrate and accommodate heterogeneous, multidimensional (omics, histological images and clinical) data for the multi-angled portrayal and management of skin cancer patients. The proposed design offers a layered analytical framework as an expansion of current EHR systems, which can integrate high-volume molecular -omics data, imaging data, as well as relevant clinical observations. We present a case study in the field of dermatology, where we attempt to combine the multilayered information for the early detection and characterization of melanoma. The specific architecture aspires to lower the barrier for the introduction of personalized therapeutic approaches, towards precision medicine. The paper describes the technical issues of implementation, along with an initial evaluation of the system and discussion.

Effective Melanoma Recognition Using Deep Convolutional Neural Network with Covariance Discriminant Loss

Melanoma recognition is challenging due to data imbalance and high intra-class variations and large inter-class similarity. Aiming at the issues, we propose a melanoma recognition method using deep convolutional neural network with covariance discriminant loss in dermoscopy images. Deep convolutional neural network is trained under the joint supervision of cross entropy loss and covariance discriminant loss, rectifying the model outputs and the extracted features simultaneously. Specifically, we design an embedding loss, namely covariance discriminant loss, which takes the first and second distance into account simultaneously for providing more constraints. By constraining the distance between hard samples and minority class center, the deep features of melanoma and non-melanoma can be separated effectively. To mine the hard samples, we also design the corresponding algorithm. Further, we analyze the relationship between the proposed loss and other losses. On the International Symposium on Biomedical Imaging (ISBI) 2018 Skin Lesion Analysis dataset, the two schemes in the proposed method can yield a sensitivity of 0.942 and 0.917, respectively. The comprehensive results have demonstrated the efficacy of the designed embedding loss and the proposed methodology.

A comparison of imaging software and conventional cell counting in determining melanocyte density in photodamaged control sample and melanoma in situ biopsies

BACKGROUND

Objective methods for distinguishing melanoma in situ (MIS) from photodamaged skin (PS) are needed to guide treatment in patients with melanocytic proliferations. Melanocyte density (MD) could serve as an objective histopathological criterion in difficult cases. Calculating MD via manual cell counts (MCC) with immunohistochemical (IHC)-stained slides has been previously published. However, the clinical application of this method is questionable, as quantification of MD via MCC on difficult cases is time consuming, especially in high volume practices.

METHODS

ImageJ is an image processing software that uses scanned slide images to determine cell count. In this study, we compared MCC to ImageJ calculated MD in microphthalmia transcription factor-IHC stained MIS biopsies and control PS acquired from the same patients.

RESULTS

We found a statistically significant difference in MD between PS and MIS as measured by both MCC and ImageJ software (P < 0.01). Additionally, no statistically significant difference was found when comparing MD measurements recorded by ImageJ vs those determined by the MCC method.

CONCLUSION

MD as determined by ImageJ strongly correlates with the MD calculated by MCC. We propose the use of ImageJ as a time-efficient, objective, and reproducible tool to assess MD.

Past and present of computer-assisted dermoscopic diagnosis: performance of a conventional image analyser versus a convolutional neural network in a prospective data set of 1,981 skin lesions

BACKGROUND

Convolutional neural networks (CNNs) have shown a dermatologist-level performance in the classification of skin lesions. We aimed to deliver a head-to-head comparison of a conventional image analyser (CIA), which depends on segmentation and weighting of handcrafted features, to a CNN trained by deep learning.

METHODS

Cross-sectional study using a real-world, prospectively acquired, dermoscopic dataset of 1981 skin lesions to compare the diagnostic performance of a market-approved CNN (Moleanalyzer-Pro™, developed in 2018) to a CIA (Moleanalyzer-3™/Dynamole™; developed in 2004, all FotoFinder Systems Inc, Germany). As a reference standard, we used histopathological diagnoses (n = 785) or, in non-excised benign lesions (n = 1196), expert consensus plus an uneventful follow-up by sequential digital dermoscopy for at least 2 years.

RESULTS

A total of 281 malignant lesions and 1700 benign lesions from 435 patients (62.2% male, mean age: 52 years) were prospectively imaged. The CNN showed a sensitivity of 77.6% (95% confidence interval [CI]: [72.4%-82.1%]), specificity of 95.3% (95% CI: [94.2%-96.2%]), and receiver operating characteristic (ROC)-area under the curve (AUC) of 0.945 (95% CI: [0.930-0.961]). In contrast, the CIA achieved a sensitivity of 53.4% (95% CI: [47.5%-59.1%]), specificity of 86.6% (95% CI: [84.9%-88.1%]) and ROC-AUC of 0.738 (95% CI: [0.701-0.774]). The data set included melanomas originally diagnosed by dynamic changes during sequential digital dermoscopy (52 of 201, 20.6%), which reduced the sensitivities of both classifiers. Pairwise comparisons of sensitivities, specificities, and ROC-AUCs indicated a clear outperformance by the CNN (all p < 0.001).

CONCLUSIONS

The superior diagnostic performance of the CNN argues against a continued application of former CIAs as an aide to physicians' clinical management decisions.

Computational texture features of dermoscopic images and their link to the descriptive terminology: A survey

Computer-extracted texture features are relevant to diagnose cutaneous lesions such as melanomas. Our goal is to set a relationship between a well-established descriptive terminology, which describes the attributes of dermoscopic structures based on their aspect rather than their underlying causes, and the computational methods to extract texture-based features. By tackling this problem, we can ascertain what indicators used by dermatologists are reflected in the extracted texture features. We first review the state-of-the-art models for texture extraction in dermoscopic images. By comparing the methods' performance and goals, we conclude that (I) a single color space does not seem to give performances as good as using several ones, thus the latter is reasonable (II) the optimal number of extracted features seems to vary depending on the method's goal, and extracting a large number of features can lead to a loss of models robustness (III) methods such as GLCM, Sobel or Law energy filters are mainly used to capture local properties to detect specific dermoscopic structures (IV) methods that extract local and global features, like Gabor wavelets or SPT, tend to be used to analyze the presence of certain patterns of dermoscopic structures, e.g. globular, reticular, etc.

Expert-Level Diagnosis of Nonpigmented Skin Cancer by Combined Convolutional Neural Networks

Importance

Convolutional neural networks (CNNs) achieve expert-level accuracy in the diagnosis of pigmented melanocytic lesions. However, the most common types of skin cancer are nonpigmented and nonmelanocytic, and are more difficult to diagnose.

Objective

To compare the accuracy of a CNN-based classifier with that of physicians with different levels of experience.

Design, Setting, and Participants

A CNN-based classification model was trained on 7895 dermoscopic and 5829 close-up images of lesions excised at a primary skin cancer clinic between January 1, 2008, and July 13, 2017, for a combined evaluation of both imaging methods. The combined CNN (cCNN) was tested on a set of 2072 unknown cases and compared with results from 95 human raters who were medical personnel, including 62 board-certified dermatologists, with different experience in dermoscopy.

Main Outcomes and Measures

The proportions of correct specific diagnoses and the accuracy to differentiate between benign and malignant lesions measured as an area under the receiver operating characteristic curve served as main outcome measures.

Results

Among 95 human raters (51.6% female; mean age, 43.4 years; 95% CI, 41.0-45.7 years), the participants were divided into 3 groups (according to years of experience with dermoscopy): beginner raters (<3 years), intermediate raters (3-10 years), or expert raters (>10 years). The area under the receiver operating characteristic curve of the trained cCNN was higher than human ratings (0.742; 95% CI, 0.729-0.755 vs 0.695; 95% CI, 0.676-0.713; P < .001). The specificity was fixed at the mean level of human raters (51.3%), and therefore the sensitivity of the cCNN (80.5%; 95% CI, 79.0%-82.1%) was higher than that of human raters (77.6%; 95% CI, 74.7%-80.5%). The cCNN achieved a higher percentage of correct specific diagnoses compared with human raters (37.6%; 95% CI, 36.6%-38.4% vs 33.5%; 95% CI, 31.5%-35.6%; P = .001) but not compared with experts (37.3%; 95% CI, 35.7%-38.8% vs 40.0%; 95% CI, 37.0%-43.0%; P = .18).

Conclusions and Relevance

Neural networks are able to classify dermoscopic and close-up images of nonpigmented lesions as accurately as human experts in an experimental setting.

Computer-assisted diagnosis techniques (dermoscopy and spectroscopy-based) for diagnosing skin cancer in adults

BACKGROUND

Early accurate detection of all skin cancer types is essential to guide appropriate management and to improve morbidity and survival. Melanoma and cutaneous squamous cell carcinoma (cSCC) are high-risk skin cancers which have the potential to metastasise and ultimately lead to death, whereas basal cell carcinoma (BCC) is usually localised with potential to infiltrate and damage surrounding tissue. Anxiety around missing early curable cases needs to be balanced against inappropriate referral and unnecessary excision of benign lesions. Computer-assisted diagnosis (CAD) systems use artificial intelligence to analyse lesion data and arrive at a diagnosis of skin cancer. When used in unreferred settings ('primary care'), CAD may assist general practitioners (GPs) or other clinicians to more appropriately triage high-risk lesions to secondary care. Used alongside clinical and dermoscopic suspicion of malignancy, CAD may reduce unnecessary excisions without missing melanoma cases.

OBJECTIVES

To determine the accuracy of CAD systems for diagnosing cutaneous invasive melanoma and atypical intraepidermal melanocytic variants, BCC or cSCC in adults, and to compare its accuracy with that of dermoscopy.

SEARCH METHODS

We undertook a comprehensive search of the following databases from inception up to August 2016: Cochrane Central Register of Controlled Trials (CENTRAL); MEDLINE; Embase; CINAHL; CPCI; Zetoc; Science Citation Index; US National Institutes of Health Ongoing Trials Register; NIHR Clinical Research Network Portfolio Database; and the World Health Organization International Clinical Trials Registry Platform. We studied reference lists and published systematic review articles.

SELECTION CRITERIA

Studies of any design that evaluated CAD alone, or in comparison with dermoscopy, in adults with lesions suspicious for melanoma or BCC or cSCC, and compared with a reference standard of either histological confirmation or clinical follow-up.

DATA COLLECTION AND ANALYSIS

Two review authors independently extracted all data using a standardised data extraction and quality assessment form (based on QUADAS-2). We contacted authors of included studies where information related to the target condition or diagnostic threshold were missing. We estimated summary sensitivities and specificities separately by type of CAD system, using the bivariate hierarchical model. We compared CAD with dermoscopy using (a) all available CAD data (indirect comparisons), and (b) studies providing paired data for both tests (direct comparisons). We tested the contribution of human decision-making to the accuracy of CAD diagnoses in a sensitivity analysis by removing studies that gave CAD results to clinicians to guide diagnostic decision-making.

MAIN RESULTS

We included 42 studies, 24 evaluating digital dermoscopy-based CAD systems (Derm-CAD) in 23 study cohorts with 9602 lesions (1220 melanomas, at least 83 BCCs, 9 cSCCs), providing 32 datasets for Derm-CAD and seven for dermoscopy. Eighteen studies evaluated spectroscopy-based CAD (Spectro-CAD) in 16 study cohorts with 6336 lesions (934 melanomas, 163 BCC, 49 cSCCs), providing 32 datasets for Spectro-CAD and six for dermoscopy. These consisted of 15 studies using multispectral imaging (MSI), two studies using electrical impedance spectroscopy (EIS) and one study using diffuse-reflectance spectroscopy. Studies were incompletely reported and at unclear to high risk of bias across all domains. Included studies inadequately address the review question, due to an abundance of low-quality studies, poor reporting, and recruitment of highly selected groups of participants.Across all CAD systems, we found considerable variation in the hardware and software technologies used, the types of classification algorithm employed, methods used to train the algorithms, and which lesion morphological features were extracted and analysed across all CAD systems, and even between studies evaluating CAD systems. Meta-analysis found CAD systems had high sensitivity for correct identification of cutaneous invasive melanoma and atypical intraepidermal melanocytic variants in highly selected populations, but with low and very variable specificity, particularly for Spectro-CAD systems. Pooled data from 22 studies estimated the sensitivity of Derm-CAD for the detection of melanoma as 90.1% (95% confidence interval (CI) 84.0% to 94.0%) and specificity as 74.3% (95% CI 63.6% to 82.7%). Pooled data from eight studies estimated the sensitivity of multispectral imaging CAD (MSI-CAD) as 92.9% (95% CI 83.7% to 97.1%) and specificity as 43.6% (95% CI 24.8% to 64.5%). When applied to a hypothetical population of 1000 lesions at the mean observed melanoma prevalence of 20%, Derm-CAD would miss 20 melanomas and would lead to 206 false-positive results for melanoma. MSI-CAD would miss 14 melanomas and would lead to 451 false diagnoses for melanoma. Preliminary findings suggest CAD systems are at least as sensitive as assessment of dermoscopic images for the diagnosis of invasive melanoma and atypical intraepidermal melanocytic variants. We are unable to make summary statements about the use of CAD in unreferred populations, or its accuracy in detecting keratinocyte cancers, or its use in any setting as a diagnostic aid, because of the paucity of studies.

AUTHORS' CONCLUSIONS

In highly selected patient populations all CAD types demonstrate high sensitivity, and could prove useful as a back-up for specialist diagnosis to assist in minimising the risk of missing melanomas. However, the evidence base is currently too poor to understand whether CAD system outputs translate to different clinical decision-making in practice. Insufficient data are available on the use of CAD in community settings, or for the detection of keratinocyte cancers. The evidence base for individual systems is too limited to draw conclusions on which might be preferred for practice. Prospective comparative studies are required that evaluate the use of already evaluated CAD systems as diagnostic aids, by comparison to face-to-face dermoscopy, and in participant populations that are representative of those in which the test would be used in practice.

Artificial intelligence in healthcare

Artificial intelligence (AI) is gradually changing medical practice. With recent progress in digitized data acquisition, machine learning and computing infrastructure, AI applications are expanding into areas that were previously thought to be only the province of human experts. In this Review Article, we outline recent breakthroughs in AI technologies and their biomedical applications, identify the challenges for further progress in medical AI systems, and summarize the economic, legal and social implications of AI in healthcare.

Medical Image Classification Based on Deep Features Extracted by Deep Model and Statistic Feature Fusion with Multilayer Perceptron‬

Medical image classification is a key technique of Computer-Aided Diagnosis (CAD) systems. Traditional methods rely mainly on the shape, color, and/or texture features as well as their combinations, most of which are problem-specific and have shown to be complementary in medical images, which leads to a system that lacks the ability to make representations of high-level problem domain concepts and that has poor model generalization ability. Recent deep learning methods provide an effective way to construct an end-to-end model that can compute final classification labels with the raw pixels of medical images. However, due to the high resolution of the medical images and the small dataset size, deep learning models suffer from high computational costs and limitations in the model layers and channels. To solve these problems, in this paper, we propose a deep learning model that integrates Coding Network with Multilayer Perceptron (CNMP), which combines high-level features that are extracted from a deep convolutional neural network and some selected traditional features. The construction of the proposed model includes the following steps. First, we train a deep convolutional neural network as a coding network in a supervised manner, and the result is that it can code the raw pixels of medical images into feature vectors that represent high-level concepts for classification. Second, we extract a set of selected traditional features based on background knowledge of medical images. Finally, we design an efficient model that is based on neural networks to fuse the different feature groups obtained in the first and second step. We evaluate the proposed approach on two benchmark medical image datasets: HIS2828 and ISIC2017. We achieve an overall classification accuracy of 90.1% and 90.2%, respectively, which are higher than the current successful methods.

Recognition of pigment network pattern in dermoscopy images based on fuzzy classification of pixels

BACKGROUND AND OBJECTIVE

One of the most relevant dermoscopic patterns is the pigment network. An innovative method of pattern recognition is presented for its detection in dermoscopy images.

METHODS

It consists of two steps. In the first one, by means of a supervised machine learning process and after performing the extraction of different colour and texture features, a fuzzy classification of pixels into the three categories present in the pattern's definition ("net", "hole" and "other") is carried out. This enables the three corresponding fuzzy sets to be created and, as a result, the three probability images that map them out are generated. In the second step, the pigment network pattern is characterised from a parameterisation process -derived from the system specification- and the subsequent extraction of different features calculated from the combinations of image masks extracted from the probability images, corresponding to the alpha-cuts obtained from the fuzzy sets.

RESULTS

The method was tested on a database of 875 images -by far the largest used in the state of the art to detect pigment network- extracted from a public Atlas of Dermoscopy, obtaining AUC results of 0.912 and 88%% accuracy, with 90.71%% sensitivity and 83.44%% specificity.

CONCLUSION

The main contribution of this method is the very design of the algorithm, highly innovative, which could also be used to deal with other pattern recognition problems of a similar nature. Other contributions are: 1. The good performance in discriminating between the pattern and the disturbing artefacts -which means that no prior preprocessing is required in this method- and between the pattern and other dermoscopic patterns; 2. It puts forward a new methodological approach for work of this kind, introducing the system specification as a required step prior to algorithm design and development, being this specification the basis for a required parameterisation -in the form of configurable parameters (with their value ranges) and set threshold values- of the algorithm and the subsequent conducting of the experiments.

Texture based skin lesion abruptness quantification to detect malignancy

BACKGROUND

Abruptness of pigment patterns at the periphery of a skin lesion is one of the most important dermoscopic features for detection of malignancy. In current clinical setting, abrupt cutoff of a skin lesion determined by an examination of a dermatologist. This process is subjective, nonquantitative, and error-prone. We present an improved computational model to quantitatively measure abruptness of a skin lesion over our previous method. To achieve this, we quantitatively analyze the texture features of a region within the lesion boundary. This region is bounded by an interior border line of the lesion boundary which is determined using level set propagation (LSP) method. This method provides a fast border contraction without a need for extensive boolean operations. Then, we build feature vectors of homogeneity, standard deviation of pixel values, and mean of the pixel values of the region between the contracted border and the original border. These vectors are then classified using neural networks (NN) and SVM classifiers.

RESULTS

As lower homogeneity indicates sharp cutoffs, suggesting melanoma, we carried out our experiments on two dermoscopy image datasets, which consist of 800 benign and 200 malignant melanoma cases. LSP method helped produce better results than Kaya et al., 2016 study. By using texture homogeneity at the periphery of a lesion border determined by LSP, as a classification results, we obtained 87% f1-score and 78% specificity; that we obtained better results than in the previous study. We also compared the performances of two different NN classifiers and support vector machine classifier. The best results obtained using combination of RGB color spaces with the fully-connected multi-hidden layer NN.

CONCLUSIONS

Computational results also show that skin lesion abrupt cutoff is a reliable indicator of malignancy. Results show that computational model of texture homogeneity along the periphery of skin lesion borders based on LSP is an effective way of quantitatively measuring abrupt cutoff of a lesion.

Incorporating Colour Information for Computer-Aided Diagnosis of Melanoma from Dermoscopy Images: A Retrospective Survey and Critical Analysis

Cutaneous melanoma is the most life-threatening form of skin cancer. Although advanced melanoma is often considered as incurable, if detected and excised early, the prognosis is promising. Today, clinicians use computer vision in an increasing number of applications to aid early detection of melanoma through dermatological image analysis (dermoscopy images, in particular). Colour assessment is essential for the clinical diagnosis of skin cancers. Due to this diagnostic importance, many studies have either focused on or employed colour features as a constituent part of their skin lesion analysis systems. These studies range from using low-level colour features, such as simple statistical measures of colours occurring in the lesion, to availing themselves of high-level semantic features such as the presence of blue-white veil, globules, or colour variegation in the lesion. This paper provides a retrospective survey and critical analysis of contributions in this research direction.

Computer-assisted melanoma diagnosis: a new integrated system

In dermatology, attempts at synergy between man and machine have mainly been made to improve melanoma diagnosis. The aim of the present study was to test an 'integrated digital dermoscopy analysis' (i-DDA) system with a series of melanocytic lesions that were benign and malignant in nature, and to evaluate its discriminating power with respect to histological diagnosis. In a retrospective study we used an i-DDA system to evaluate a series of 856 excised, clinically atypical pigmented skin lesions (584 benign and 272 malignant). The system evaluated 48 parameters to be studied as possible discriminant variables, grouped into four categories (geometries, colours, textures and islands of colour) integrated with three personal metadata items (sex, age and site of lesion) and presence/absence of three dermoscopic patterns (regression structures, blue-white veil and polymorphic vascular structures). Stepwise multivariate logistic regression of i-DDA data selected nine variables with the highest possible discriminant power. At the end of the stepwise procedure the percentage of cases correctly classified by i-DDA was 89.2% (100% sensitivity and 40.8% specificity). The limitations of the study included those associated with a retrospective design and the 'a priori' exclusion of nonmelanocytic skin lesions. By incorporating numerical digital features with personal data and some dermoscopic patterns into the learning process, the proposed i-DDA improved the performance of assisted melanoma diagnosis, with the advantage that our results can be objectively repeated in any other clinical setting.

Automatic Classification of Specific Melanocytic Lesions Using Artificial Intelligence

Background. Given its propensity to metastasize, and lack of effective therapies for most patients with advanced disease, early detection of melanoma is a clinical imperative. Different computer-aided diagnosis (CAD) systems have been proposed to increase the specificity and sensitivity of melanoma detection. Although such computer programs are developed for different diagnostic algorithms, to the best of our knowledge, a system to classify different melanocytic lesions has not been proposed yet. Method. In this research we present a new approach to the classification of melanocytic lesions. This work is focused not only on categorization of skin lesions as benign or malignant but also on specifying the exact type of a skin lesion including melanoma, Clark nevus, Spitz/Reed nevus, and blue nevus. The proposed automatic algorithm contains the following steps: image enhancement, lesion segmentation, feature extraction, and selection as well as classification. Results. The algorithm has been tested on 300 dermoscopic images and achieved accuracy of 92% indicating that the proposed approach classified most of the melanocytic lesions correctly. Conclusions. A proposed system can not only help to precisely diagnose the type of the skin mole but also decrease the amount of biopsies and reduce the morbidity related to skin lesion excision.

Automatic differentiation of melanoma from dysplastic nevi

Malignant melanoma causes the majority of deaths related to skin cancer. Nevertheless, it is the most treatable one, depending on its early diagnosis. The early prognosis is a challenging task for both clinicians and dermatologist, due to the characteristic similarities of melanoma with other skin lesions such as dysplastic nevi. In the past decades, several computerized lesion analysis algorithms have been proposed by the research community for detection of melanoma. These algorithms mostly focus on differentiating melanoma from benign lesions and few have considered the case of melanoma against dysplastic nevi. In this paper, we consider the most challenging task and propose an automatic framework for differentiation of melanoma from dysplastic nevi. The proposed framework also considers combination and comparison of several texture features beside the well used colour and shape features based on "ABCD" clinical rule in the literature. Focusing on dermoscopy images, we evaluate the performance of the framework using two feature extraction approaches, global and local (bag of words) and three classifiers such as support vector machine, gradient boosting and random forest. Our evaluation revealed the potential of texture features and random forest as an almost independent classifier. Using texture features and random forest for differentiation of melanoma and dysplastic nevi, the framework achieved the highest sensitivity of 98% and specificity of 70%.

Melanoma and other skin lesion detection using smart handheld devices

Smartphones of the latest generation featuring advanced multicore processors, dedicated microchips for graphics, high-resolution cameras, and innovative operating systems provide a portable platform for running sophisticated medical screening software and delivering point-of-care patient diagnostic services at a very low cost. In this chapter, we present a smartphone digital dermoscopy application that can analyze high-resolution images of skin lesions and provide the user with feedback about the likelihood of malignancy. The same basic procedure has been adapted to evaluate other skin lesions, such as the flesh-eating bacterial disease known as Buruli ulcer. When implemented on the iPhone, the accuracy and speed achieved by this application are comparable to that of a desktop computer, demonstrating that smartphone applications can combine portability and low cost with high performance. Thus, smartphone-based systems can be used as assistive devices by primary care physicians during routine office visits, and they can have a significant impact in underserved areas and in developing countries, where health-care infrastructure is limited.

Improving dermoscopy image classification using color constancy

Robustness is one of the most important characteristics of computer-aided diagnosis systems designed for dermoscopy images. However, it is difficult to ensure this characteristic if the systems operate with multisource images acquired under different setups. Changes in the illumination and acquisition devices alter the color of images and often reduce the performance of the systems. Thus, it is important to normalize the colors of dermoscopy images before training and testing any system. In this paper, we investigate four color constancy algorithms: Gray World, max-RGB, Shades of Gray, and General Gray World. Our results show that color constancy improves the classification of multisource images, increasing the sensitivity of a bag-of-features system from 71.0% to 79.7% and the specificity from 55.2% to 76% using only 1-D RGB histograms as features.

Computer aided diagnostic support system for skin cancer: a review of techniques and algorithms

Image-based computer aided diagnosis systems have significant potential for screening and early detection of malignant melanoma. We review the state of the art in these systems and examine current practices, problems, and prospects of image acquisition, pre-processing, segmentation, feature extraction and selection, and classification of dermoscopic images. This paper reports statistics and results from the most important implementations reported to date. We compared the performance of several classifiers specifically developed for skin lesion diagnosis and discussed the corresponding findings. Whenever available, indication of various conditions that affect the technique's performance is reported. We suggest a framework for comparative assessment of skin cancer diagnostic models and review the results based on these models. The deficiencies in some of the existing studies are highlighted and suggestions for future research are provided.

Unified approach for lesion border detection based on mixture modeling and local entropy thresholding

BACKGROUND/PURPOSE

Computer-aided design (CAD) methods are highly valuable for the analysis of skin lesions using digital dermoscopy due to low rate of diagnostic accuracy of expert dermatologist. In computerized diagnostic methods, automatic border detection is the first and crucial step.

METHOD

In this study, a novel unified approach is proposed for automatic border detection (ABD). A preprocessing step is performed by normalized smoothing filter (NSF) to reduce background noise. Mixture models technique is then utilized to initially segment the lesion area roughly. Afterward, local entropy thresholding is performed to extract the lesion candidate pixels and the lesion border is smoothed using morphological reconstruction.

RESULTS

The overall ABD system is tested on a set of 100 dermoscopy images with ground truth. A comparative study was conducted with the other three state-of-the-art methods using statistical metrics. This ABD technique has the minimal average error probability rate of 5%, true detection of 92.10% and false positive rate of 6.41%.

CONCLUSION

Results demonstrate that the proposed method segments the lesion area accurately. Sample dataset and execute software are available online and can be downloaded from: http://cs.ntu.edu.pk/research.

A simple scoring system for the diagnosis of palmo-plantar pigmented skin lesions by digital dermoscopy analysis

BACKGROUND

Many research groups have recently developed equipments and statistical methods enabling pattern classification of pigmented skin lesions. To differentiate between benign and malignant ones, the mathematical extraction of digital patterns together with the use of appropriate statistical approaches is a challenging task.

OBJECTIVE

To design a simple scoring model that provides accurate classification of benign and malignant palmo-plantar pigmented skin lesions, by evaluation of parameters obtained by digital dermoscopy analysis (DDA).

PATIENTS AND METHODS

In the present study we used a digital dermoscopy analyser to evaluate a series of 445 palmo-plantar melanocytic skin lesion images (25 melanomas 420 nevi). Area under the receiver operator curve, sensitivity and specificity were calculated to evaluate the diagnostic performance of our scoring model for the differentiation of benign and malignant palmo-plantar melanocytic lesions.

RESULTS

Model performance reached a very high value (0.983). The DDA parameters selected by the model that proved statistically significant were: area, peripheral dark regions, total imbalance of colours, entropy, dark area and red and blue multicomponent. When all seven model variables were used in a multivariate mode, setting sensitivity at 100% to avoid false negatives, we estimated a minimum specificity of about 80%.

CONCLUSIONS

Simplicity of use and effectiveness of implementation are important requirements for the success of quantitative methods in routine clinical practice. Scoring systems meet these requirements. Their outcomes are accessible in real time without the use of any data processing system, thus allowing decisions to be made quickly and effectively.

A system for the detection of pigment network in dermoscopy images using directional filters

A pigment network is one of the most important dermoscopic structures. This paper describes an automatic system that performs its detection in dermoscopy images. The proposed system involves a set of sequential steps. First, a preprocessing algorithm is applied to the dermoscopy image. Then, a bank of directional filters and a connected component analysis are used in order to detect the "lines" of the pigment network. Finally, features are extracted from the detected network and used to train an AdaBoost algorithm to classify each lesion regarding the presence of the pigment network. The algorithm was tested on a dataset of 200 medically annotated images from the database of Hospital Pedro Hispano (Matosinhos), achieving a sensitivity = 91.1% and a specificity = 82.1%.

On the role of texture and color in the classification of dermoscopy images

This paper addresses the detection of melanoma lesions in dermoscopy images, using texture and color features. Although melanoma detection has been studied in several works, using different types of texture, color and shape features, it is not always clear what is the role of each set of features and which features are most discriminative. This papers aims at clarifying the role of texture and color features. Furthermore, the proposed systems is based on features which can be easily implemented and tested by other researchers. It is concluded that both types of features achieve good detection scores when used alone. The best results (SE=94.1%, SP=77.4%) are achieved by combining them both.

A possible melanoma discrimination index based on hyperspectral data: a pilot study

BACKGROUND

Early detection and proper excision of the primary lesions of malignant melanoma (MM) are crucial for reducing melanoma-related deaths. To support the early detection of melanoma, automated melanoma screening systems have been extensively studied and developed. In this article, we present a hyperspectral melanoma screening system and propose a possible melanoma discrimination index derived from the characteristics of the pigment molecules in the skin, both of which have been derived from hyperspectral data (HSD).

METHODS

The index expresses the disordered nature of each lesion including variegation in color based on variation in spectral information obtained from each lesion. Performance of the index in discriminating melanomas from other pigmented skin lesions has been studied in five cases of melanoma (41 HSD sets), one case of Spitz nevus (13 HSD sets), 10 cases of seborrheic keratosis (78 HSD sets), three cases of basal cell carcinoma (16 HSD sets), and nine cases of melanocytic nevus (21 HSD sets), obtained from patients and volunteers, all of whom were Japanese.

RESULTS

Performance of the index, which reflects the disordered nature of a lesion, discriminates melanomas with a sensitivity of 90%, a specificity of 84%, and an area under the receiver operating characteristic curve of 0.93, on resubstitution.

CONCLUSION

An objective melanoma discrimination index at a molecular pigmentary level, derived from HSD, has been proposed, and its performance evaluated. This index was highly successful in discriminating MM from non-melanoma, although the statistical population was small.

Patient acceptance and diagnostic utility of automated digital image analysis of pigmented skin lesions

BACKGROUND

Computerized analysis of pigmented skin lesions may help to increase diagnostic accuracy for melanoma, help to avoid unnecessary procedures and reduce health care costs.

OBJECTIVES

We evaluated both the patient acceptance and diagnostic utility of such an analysis tool in a real clinical setting.

METHODS

Two hundred nine consecutive patients (median age: 34 years, range: 2-73 years), who were concerned about a pigmented skin lesion, answered a questionnaire about their attitude towards computerized analysis and their confidence in the resulting findings. Using a dermoscopy analyser, their skin lesions (n = 219) were then grouped into the categories, benign, suspicious and malignant, and results were compared with those obtained by in-person examination of dermato-oncologic experts.

RESULTS

More than half of the patients (n = 114) would accept the use of computer analysis for melanoma screening; although 16 (14.0%) patients would accept this method solely, 98 (86.0%) patients would prefer an additional in-person examination by a dermatologist. Of the 219 pigmented skin lesions, the dermoscopic experts rated 171 (78.1%) as benign, 36 (16.4%) as suspicious and 12 (5.5%) as malignant, whereas computer analysis revealed 102 (46.6%) benign, 78 (35.6%) suspicious and 39 (17.8%) malignant lesions. At the expense of specificity (48.8%), the sensitivity of computerized analysis was excellent (100%) and equal to that of in-person examination.

CONCLUSIONS

Most patients would accept computer analysis for melanoma screening, some of them even without reservations. However, due to a high rate of false positive computer assessments, it cannot be recommended as a screening tool at this time.

Skin tumor area extraction using an improved dynamic programming approach

BACKGROUND/PURPOSE

Border (B) description of melanoma and other pigmented skin lesions is one of the most important tasks for the clinical diagnosis of dermoscopy images using the ABCD rule. For an accurate description of the border, there must be an effective skin tumor area extraction (STAE) method. However, this task is complicated due to uneven illumination, artifacts present in the lesions and smooth areas or fuzzy borders of the desired regions.

METHODS

In this paper, a novel STAE algorithm based on improved dynamic programming (IDP) is presented. The STAE technique consists of the following four steps: color space transform, pre-processing, rough tumor area detection and refinement of the segmented area. The procedure is performed in the CIE L(*) a(*) b(*) color space, which is approximately uniform and is therefore related to dermatologist's perception. After pre-processing the skin lesions to reduce artifacts, the DP algorithm is improved by introducing a local cost function, which is based on color and texture weights.

RESULTS

The STAE method is tested on a total of 100 dermoscopic images. In order to compare the performance of STAE with other state-of-the-art algorithms, various statistical measures based on dermatologist-drawn borders are utilized as a ground truth. The proposed method outperforms the others with a sensitivity of 96.64%, a specificity of 98.14% and an error probability of 5.23%.

CONCLUSION

The results demonstrate that this STAE method by IDP is an effective solution when compared with other state-of-the-art segmentation techniques. The proposed method can accurately extract tumor borders in dermoscopy images.

Classification of melanocytic skin lesions from non-melanocytic lesions

In this paper, we present a classification method of dermoscopy images between melanocytic skin lesions (MSLs) and non-melanocytic skin lesions (NoMSLs). The motivation of this research is to develop a pre-processor of an automated melanoma screening system. Since NoMSLs have a wide variety of shapes and their border is often ambiguous, we developed a new tumor area extraction algorithm to account for these difficulties. We confirmed that this algorithm is capable of handling different dermoscopy images not only those of NoMSLs but also MSLs as well. We determined the tumor area from the image using this new algorithm, calculated a total 428 features from each image, and built a linear classifier. We found only two image features, "the skewness of bright region in the tumor along its major axis" and "the difference between the average intensity in the peripheral part of the tumor and that in the normal skin area using the blue channel" were very efficient at classifying NoMSLs and MSLs. The detection accuracy of MSLs by our classifier using only the above mentioned image feature has a sensitivity of 98.0% and a specificity of 86.6% in a set of 107 non-melanocytic and 548 melanocytic dermoscopy images using a cross-validation test.

Skin cancer recognition by using a neuro-fuzzy system

Skin cancer is the most prevalent cancer in the light-skinned population and it is generally caused by exposure to ultraviolet light. Early detection of skin cancer has the potential to reduce mortality and morbidity. There are many diagnostic technologies and tests to diagnose skin cancer. However many of these tests are extremely complex and subjective and depend heavily on the experience of the clinician. To obviate these problems, image processing techniques, a neural network system (NN) and a fuzzy inference system were used in this study as promising modalities for detection of different types of skin cancer. The accuracy rate of the diagnosis of skin cancer by using the hierarchal neural network was 90.67% while using neuro-fuzzy system yielded a slightly higher rate of accuracy of 91.26% in diagnosis skin cancer type. The sensitivity of NN in diagnosing skin cancer was 95%, while the specificity was 88%. Skin cancer diagnosis by neuro-fuzzy system achieved sensitivity of 98% and a specificity of 89%.

Detecting the pigment network in dermoscopy images: a directional approach

Several algorithms have been recently proposed for the analysis of dermoscopy images and the detection of melanomas. However, the pigment network is not considered in most of these works, although this cue plays a major role in clinical diagnosis routines. This paper proposes an algorithm for the detection of the pigment network. The algorithm is based on a bank of directional filters (difference of Gaussians) and explores color, directionality and topological properties of the network.

Objective melanoma progression

BACKGROUND/PURPOSE

Many aspects of the natural history of malignant melanoma (MM) are still unclear, specifically its appearance at onset and particularly how it changes in time. The purpose of our study was to retrospectively determine objective changes in melanoma over a 3-24-month observation period.

MATERIALS AND METHODS

Our study was carried out in two Italian dermatology centers. Digital dermoscopy analyzers (DB-Mips System) were used to retrospectively evaluate dermoscopic images of 59 MM (with no initial clinical aspects suggesting melanoma) under observation for 3-24 months. The analyzer evaluates 49 parameters grouped into four categories: geometries, colors, textures and islands of color. Multivariate analysis of variance for repeated measures was used to evaluate the statistical significance of the changes in the digital dermoscopy variables of melanomas.

RESULTS

Within-lesion analysis indicated that melanomas increased in dimension (Area, Minimum, and Maximum Diameter), manifested greater disorganization of the internal components (Red, Green and Blue Multicomponent, Contrast, and Entropy) and increased in clusters of milky pink color (Light Red Area).

CONCLUSION

Analysis of the parameters of our model and statistical analysis enabled us to interpret/identify the most significant factors of melanoma modification, providing quantitative insights into the natural history of this cutaneous malignancy.

A support vector machine for decision support in melanoma recognition

The early diagnosis of melanoma is critical to achieving reduced mortality and increased survival. Although clinical examination is currently the method of choice for melanocytic lesion assessment, difficulties may arise in the diagnosis of atypical lesions. From both the naked eye and dermoscopic perspective, dysplastic naevi often exhibit a prominent heterogeneity of structure that renders their clinical distinction from melanoma difficult. To address these problems in diagnosis, there exists a heightened interest among researchers regarding the utility of machine learning techniques in computerised image analysis. Here we report on the utility, for dermatologists, of support vector machine (SVM) technology in melanoma diagnosis, using an archive of 199 digital dermoscopic images of excised atypical melanocytic lesions. Our best validation models achieved an average sensitivity and specificity for melanoma diagnosis of 0.86 and 0.72, respectively. Applying the best model to our test set yielded a sensitivity of 0.89, a diagnostic odds ratio of 14.09 and an area under the receiver operated characteristic curve (AUC) of 0.76. Advantages of the procedure implemented are the simplicity of feature extraction and the computationally cheap and efficient nature of SVMs. The derived model generalises well with outcomes that compare favourably with competing algorithms and expert assessment. In line with the concept of the utility of decision support systems in clinical practice, we propose that to reduce the risk of missed melanomas, both the dermatologists' assessment and the SVM diagnosis be incorporated into the clinical decision-making process.

Lacunarity analysis: a promising method for the automated assessment of melanocytic naevi and melanoma

The early diagnosis of melanoma is critical to achieving reduced mortality and increased survival. Although clinical examination is currently the method of choice for melanocytic lesion assessment, there is a growing interest among clinicians regarding the potential diagnostic utility of computerised image analysis. Recognising that there exist significant shortcomings in currently available algorithms, we are motivated to investigate the utility of lacunarity, a simple statistical measure previously used in geology and other fields for the analysis of fractal and multi-scaled images, in the automated assessment of melanocytic naevi and melanoma. Digitised dermoscopic images of 111 benign melanocytic naevi, 99 dysplastic naevi and 102 melanomas were obtained over the period 2003 to 2008, and subject to lacunarity analysis. We found the lacunarity algorithm could accurately distinguish melanoma from benign melanocytic naevi or non-melanoma without introducing many of the limitations associated with other previously reported diagnostic algorithms. Lacunarity analysis suggests an ordering of irregularity in melanocytic lesions, and we suggest the clinical application of this ordering may have utility in the naked-eye dermoscopic diagnosis of early melanoma.