Automated decision support in melanocytic lesion management.
ABSTRACT: An automated melanocytic lesion image-analysis algorithm is described that aims to reproduce the decision-making of a dermatologist. The utility of the algorithm lies in its ability to identify lesions requiring excision from lesions not requiring excision. Using only wavelet coefficients as features, and testing three different machine learning algorithms, a cohort of 250 images of pigmented lesions is classified based on expert dermatologists' recommendations of either excision (165 images) or no excision (85 images). It is shown that the best algorithm utilises the Shannon4 wavelet coupled to the support vector machine, where the latter is used as the classifier. In this case the algorithm, utilising only 22 othogonal features, achieves a 10-fold cross validation sensitivity and specificity of 0.96 and 0.87, resulting in a diagnostic-odds ratio of 261. The advantages of this method over diagnostic algorithms-which make a melanoma/no melanoma decision-are twofold: first, by reproducing the decision-making of a dermatologist, the average number of lesions excised per melanoma among practioners in general can be reduced without compromising the detection of melanoma; and second, the intractable problem of clinically differentiating between many atypical dysplastic naevi and melanoma is avoided. Since many atypical naevi that require excision on clinical grounds will not be melanoma, the algorithm-in contrast to diagnostic algorithms-can aim for perfect specificities without clinical concerns, thus lowering the excision rate of non-melanoma. Finally, the algorithm has been implemented as a smart phone application to investigate its utility in clinical practice and to streamline the assimilation of hitherto unseen tested images into the training set.
Project description:BACKGROUND:Computer vision may aid in melanoma detection. OBJECTIVE:We sought to compare melanoma diagnostic accuracy of computer algorithms to dermatologists using dermoscopic images. METHODS:We conducted a cross-sectional study using 100 randomly selected dermoscopic images (50 melanomas, 44 nevi, and 6 lentigines) from an international computer vision melanoma challenge dataset (n = 379), along with individual algorithm results from 25 teams. We used 5 methods (nonlearned and machine learning) to combine individual automated predictions into "fusion" algorithms. In a companion study, 8 dermatologists classified the lesions in the 100 images as either benign or malignant. RESULTS:The average sensitivity and specificity of dermatologists in classification was 82% and 59%. At 82% sensitivity, dermatologist specificity was similar to the top challenge algorithm (59% vs. 62%, P = .68) but lower than the best-performing fusion algorithm (59% vs. 76%, P = .02). Receiver operating characteristic area of the top fusion algorithm was greater than the mean receiver operating characteristic area of dermatologists (0.86 vs. 0.71, P = .001). LIMITATIONS:The dataset lacked the full spectrum of skin lesions encountered in clinical practice, particularly banal lesions. Readers and algorithms were not provided clinical data (eg, age or lesion history/symptoms). Results obtained using our study design cannot be extrapolated to clinical practice. CONCLUSION:Deep learning computer vision systems classified melanoma dermoscopy images with accuracy that exceeded some but not all dermatologists.
Project description:INTRODUCTION:Having many melanocytic naevi or 'moles' on the skin is the strongest predictor of melanoma; thus, much can be learnt from investigating naevi in the general population. We aim to improve the understanding of the epidemiology and biology of naevi by conducting a 3-year prospective study of melanocytic naevi in adults. METHODS AND ANALYSIS:This is a population-based cohort study of melanocytic naevi in 200 adults aged 20-69 years recruited via the Australian electoral roll. At baseline, participants will complete a questionnaire on their sun behaviour and health and undergo a clinical examination. Three-dimensional (3D) total-body photography will be used to record the images of skin lesions. Pigmented naevi will be analysed in terms of number, diameter, colour and border irregularity using automated analysis software (excluding scalp, beneath underwear and soles of feet). All naevi ?5?mm will be recorded using the integrated dermoscopy photographic system. A saliva sample will be obtained at baseline for genomic DNA analysis of pigmentation, naevus and melanoma-associated genes using the Illumina HumanCoreExome platform. The sun behaviour and health follow-up questionnaire, clinical examination and 3D total-body photography will be repeated every 6 months for 3 years. The first 50 participants will also undergo manual counts of naevi ?2?mm and ?5?mm at baseline, 6-month and 12-month follow-ups. Microbiopsy and excision of naevi of research interest is planned to commence at the 18-month time point among those who agree to donate samples for detailed histopathological and molecular assessment. ETHICS AND DISSEMINATION:This study was approved by the Metro South Health Human Research Ethics Committee in April 2016 (approval number: HREC/16/QPAH/125). The findings will be disseminated through peer-reviewed and non-peer-reviewed publications and presentations at conferences.
Project description:BACKGROUND: Early detection and treatment of melanoma is important for optimal clinical outcome, leading to biopsy of pigmented lesions deemed suspicious for the disease. The vast majority of such lesions are benign. Thus, a more objective and accurate means for detection of melanoma is needed to identify lesions for excision. OBJECTIVES: To provide proof-of-principle that epidermal genetic information retrieval (EGIR™; DermTech International, La Jolla, CA, U.S.A.), a method that noninvasively samples cells from stratum corneum by means of adhesive tape stripping, can be used to discern melanomas from naevi. METHODS: Skin overlying pigmented lesions clinically suspicious for melanoma was harvested using EGIR. RNA isolated from the tapes was amplified and gene expression profiled. All lesions were removed for histopathological evaluation. RESULTS: Supervised analysis of the microarray data identified 312 genes differentially expressed between melanomas, naevi and normal skin specimens (P<0·001, false discovery rate q<0·05). Surprisingly, many of these genes are known to have a role in melanocyte development and physiology, melanoma, cancer, and cell growth control. Subsequent class prediction modelling of a training dataset, consisting of 37 melanomas and 37 naevi, discovered a 17-gene classifier that discriminates these skin lesions. Upon testing with an independent dataset, this classifier discerned in situ and invasive melanomas from naevi with 100% sensitivity and 88% specificity, with an area under the curve for the receiver operating characteristic of 0·955. CONCLUSIONS: These results demonstrate that EGIR-harvested specimens can be used to detect melanoma accurately by means of a 17-gene genomic biomarker.
Project description:Overlapping histological features between benign and malignant lesions and a lack of firm diagnostic criteria for malignancy result in high rates of inter-observer variation in the diagnosis of melanocytic lesions. We aimed to investigate the differential expression of five miRNAs (21, 200c, 204, 205, and 211) in benign naevi (n?=?42), dysplastic naevi (n?=?41), melanoma in situ (n?=?42), and melanoma (n =?42) and evaluate their potential as diagnostic biomarkers of melanocytic lesions. Real-time PCR showed differential miRNA expression profiles between benign naevi; dysplastic naevi and melanoma in situ; and invasive melanoma. We applied a random forest machine learning algorithm to classify cases based on their miRNA expression profiles, which resulted in a ROC curve analysis of 0.99 for malignant melanoma and greater than 0.9 for all other groups. This indicates an overall very high accuracy of our panel of miRNAs as a diagnostic biomarker of benign, dysplastic, and malignant melanocytic lesions. However, the impact of variable lesion percentage and spatial expression patterns of miRNAs on these real-time PCR results was also considered. In situ hybridisation confirmed the expression of miRNA 21 and 211 in melanocytes, while demonstrating expression of miRNA 205 only in keratinocytes, thus calling into question its value as a biomarker of melanocytic lesions. In conclusion, we have validated some miRNAs, including miRNA 21 and 211, as potential diagnostic biomarkers of benign, dysplastic, and malignant melanocytic lesions. However, we also highlight the crucial importance of considering tissue morphology and spatial expression patterns when using molecular techniques for the discovery and validation of new biomarkers.
Project description:BACKGROUND:Several smartphone applications (app) with an automated risk assessment claim to be able to detect skin cancer at an early stage. Various studies that have evaluated these apps showed mainly poor performance. However, all studies were done in patients and lesions were mainly selected by a specialist. OBJECTIVES:To investigate the performance of the automated risk assessment of an app by comparing its assessment to that of a dermatologist in lesions selected by the participants. METHODS:Participants of a National Skin Cancer Day were enrolled in a multicentre study. Skin lesions indicated by the participants were analysed by the automated risk assessment of the app prior to blinded rating by the dermatologist. The ratings of the automated risk assessment were compared to the assessment and diagnosis of the dermatologist. Due to the setting of the Skin Cancer Day, lesions were not verified by histopathology. RESULTS:We included 125 participants (199 lesions). The app was not able to analyse 90 cases (45%) of which nine BCC, four atypical naevi and one lentigo maligna. Thirty lesions (67%) with a high and 21 with a medium risk (70%) rating by the app were diagnosed as benign naevi or seborrhoeic keratoses. The interobserver agreement between the ratings of the automated risk assessment and the dermatologist was poor (weighted kappa = 0.02; 95% CI -0.08-0.12; P = 0.74). CONCLUSIONS:The rating of the automated risk assessment was poor. Further investigations about the diagnostic accuracy in real-life situations are needed to provide consumers with reliable information about this healthcare application.
Project description:This article refers to the Computer Aided Diagnosis of the melanoma skin cancer. We derive wavelet-based features of melanoma from the dermoscopic images of pigmental skin lesions and apply binary C-SVM classifiers to discriminate malignant melanoma from dysplastic nevus. The aim of this research is to select the most efficient model of the SVM classifier for various image resolutions and to search for the best resolution-invariant wavelet bases. We show AUC as a function of the wavelet number and SVM kernels optimized by the Bayesian search for two independent data sets. Our results are compatible with the previous experiments to discriminate melanoma in dermoscopy images with ensembling and feed-forward neural networks.
Project description:Many medical professions require practitioners to perform visual categorizations in domains such as radiology, dermatology, and neurology. However, acquiring visual expertise is tedious and time-consuming and the perceptual strategies mediating visual categorization skills are poorly understood. In this paper, the Ease algorithm was developed to predict an item's categorization difficulty (Ease value) based on the item's perceptual similarity to all within-category items versus between-category items in the dataset. In this study, Ease values were used to construct an easy-to-hard and hard-to-easy training schedule for teaching melanoma diagnosis. Whereas previous visual training studies suggest that an easy-to-hard schedule benefits learning outcomes, no studies to date have demonstrated the easy-to-hard advantage with complex, real-world images. In our study, 237 melanoma and benign images were collected for training and testing purposes. The diagnostic accuracy of images was verified by an expert dermatologist. Based on their Ease values, the items were grouped into easy, medium, and hard categories, each containing an equal number of melanoma and benign lesions. During training, participants categorized images of skin lesions as either benign or melanoma and were given corrective feedback after each trial. In the easy-to-hard training condition, participants learned to categorize all the easy items first, followed by the medium items, and finally the hard items. Participants in the hard-to-easy training condition learned items in the reverse order. Post-training results showed that training in both conditions transferred to the classification of new melanoma and benign images. Participants in the easy-to-hard condition showed modest advantages both in the acquisition and retention of the melanoma diagnosis skills, but neither scheduling condition exhibited a gross advantage. The Ease values of the items predicted categorization accuracy after, but not before training, suggesting that the Ease algorithm is a promising tool for optimizing medical training in visual categorization.
Project description:BACKGROUND:The benign and malignant patterns of acral melanocytic naevi (AMN) and acral melanomas (AM) have been defined in a series of retrospective studies. A three-step algorithm was developed to determine when to biopsy acral melanocytic lesions. This algorithm has only been applied to a Japanese population. OBJECTIVES:Our study aimed to review the current management strategy of acral melanocytic lesions and to investigate the utility of the three-step algorithm in a predominately Caucasian cohort. METHODS:A retrospective search of the pathology and image databases at Mayo Clinic was performed between the years 2006 and 2016. Only cases located on a volar surface with dermoscopic images were included. Two dermatologists reviewed all dermoscopic images and assigned a global dermoscopic pattern. Clinical and follow-up data were gathered by chart review. All lesions with known diameter and pathological diagnosis were used for the three-step algorithm. RESULTS:Regular fibrillar and ridge patterns were more likely to be biopsied (P = 0.01). The majority of AMN (58.1%) and AM (60%) biopsied were due to physician-deemed concerning dermoscopic patterns. 39.2% of these cases were parallel furrow, lattice-like or regular fibrillar. When patients were asked to follow-up within a 3- to 6-month period, only 16.7% of the patients returned within that interval. The three-step algorithm would have correctly identified four of five AM for biopsy, missing a 6 mm, multicomponent, invasive melanoma. CONCLUSION:We found one major educational gap in the recognition of low-risk lesions with high rates of biopsy of the fibrillary pattern. Recognizing low-risk dermoscopic patterns could reduce the rate of biopsy of AMN by 23.3%. We identified two major practice gaps, poor patient compliance with follow-up and the potential insensitivity of the three-step algorithm to small multicomponent acral melanocytic lesions.
Project description:The prognostic significance of the major redox regulator nuclear factor erythroid-2-related factor (NRF2) is recognized in many cancers, but the role of NRF1 is not generally well understood in cancer. Our aim was to investigate these redox transcription factors in conjunction with redox-related microRNAs in naevi and melanoma. We characterized the immunohistochemical expression of NRF1 and NRF2 in 99 naevi, 88 primary skin melanomas, and 67 lymph node metastases. In addition, NRF1 and NRF2 mRNA and miR-23B, miR-93, miR-144, miR-212, miR-340, miR-383, and miR-510 levels were analysed with real-time qPCR from 54 paraffin-embedded naevi and melanoma samples. The immunohistochemical expression of nuclear NRF1 decreased from benign to dysplastic naevi (p < 0.001) and to primary melanoma (p < 0.001) and from primary melanoma to metastatic lesions (p = 0.012). Also, NRF1 mRNA levels decreased from benign naevi to dysplastic naevi (p = 0.034). Similarly, immunopositivity of NRF2 decreased from benign to dysplastic naevi (p = 0.02) and to primary lesions (p = 0.018). NRF2 mRNA decreased from benign to dysplastic naevi and primary melanomas (p = 0.012). Analysis from the Gene Expression Omnibus datasets supported the mRNA findings. High nuclear immunohistochemical NRF1 expression in pigment cells associated with a worse survival (p = 0.048) in patients with N0 disease at the time of diagnosis, and high nuclear NRF2 expression in pigment cells associated with a worse survival (p = 0.033) in patients with M0 disease at the time of diagnosis. In multivariate analysis, neither of these variables exceeded the prognostic power of Breslow. The levels of miR-144 and miR-212 associated positively with ulceration (p = 0.012 and p = 0.027, respectively) while miR-510 levels associated positively with lymph node metastases at the time of diagnosis (p = 0.004). Furthermore, the miRNAs correlated negatively with the immunohistochemical expression of NRF1 and NRF2 but positively with their respective mRNA. Together, this data sheds new light about NFE2L family factors in pigment tumors and suggests that these factors are worth for further explorations.
Project description:Importance:A high proportion of suspicious pigmented skin lesions referred for investigation are benign. Techniques to improve the accuracy of melanoma diagnoses throughout the patient pathway are needed to reduce the pressure on secondary care and pathology services. Objective:To determine the accuracy of an artificial intelligence algorithm in identifying melanoma in dermoscopic images of lesions taken with smartphone and digital single-lens reflex (DSLR) cameras. Design, Setting, and Participants:This prospective, multicenter, single-arm, masked diagnostic trial took place in dermatology and plastic surgery clinics in 7 UK hospitals. Dermoscopic images of suspicious and control skin lesions from 514 patients with at least 1 suspicious pigmented skin lesion scheduled for biopsy were captured on 3 different cameras. Data were collected from January 2017 to July 2018. Clinicians and the Deep Ensemble for Recognition of Malignancy, a deterministic artificial intelligence algorithm trained to identify melanoma in dermoscopic images of pigmented skin lesions using deep learning techniques, assessed the likelihood of melanoma. Initial data analysis was conducted in September 2018; further analysis was conducted from February 2019 to August 2019. Interventions:Clinician and algorithmic assessment of melanoma. Main Outcomes and Measures:Area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity of the algorithmic and specialist assessment, determined using histopathology diagnosis as the criterion standard. Results:The study population of 514 patients included 279 women (55.7%) and 484 white patients (96.8%), with a mean (SD) age of 52.1 (18.6) years. A total of 1550 images of skin lesions were included in the analysis (551 [35.6%] biopsied lesions; 999 [64.4%] control lesions); 286 images (18.6%) were used to train the algorithm, and a further 849 (54.8%) images were missing or unsuitable for analysis. Of the biopsied lesions that were assessed by the algorithm and specialists, 125 (22.7%) were diagnosed as melanoma. Of these, 77 (16.7%) were used for the primary analysis. The algorithm achieved an AUROC of 90.1% (95% CI, 86.3%-94.0%) for biopsied lesions and 95.8% (95% CI, 94.1%-97.6%) for all lesions using iPhone 6s images; an AUROC of 85.8% (95% CI, 81.0%-90.7%) for biopsied lesions and 93.8% (95% CI, 91.4%-96.2%) for all lesions using Galaxy S6 images; and an AUROC of 86.9% (95% CI, 80.8%-93.0%) for biopsied lesions and 91.8% (95% CI, 87.5%-96.1%) for all lesions using DSLR camera images. At 100% sensitivity, the algorithm achieved a specificity of 64.8% with iPhone 6s images. Specialists achieved an AUROC of 77.8% (95% CI, 72.5%-81.9%) and a specificity of 69.9%. Conclusions and Relevance:In this study, the algorithm demonstrated an ability to identify melanoma from dermoscopic images of selected lesions with an accuracy similar to that of specialists.