Search

We investigated whether human preferences hold the potential to improve diagnostic artificial intelligence (AI)-based decision support using skin cancer diagnosis as a use case. We utilized nonuniform rewards and penalties based on expert-generated tables, balancing the benefits and harms of various diagnostic errors, which were applied using reinforcement learning. Compared with supervised learning, the reinforcement learning model improved the sensitivity for melanoma from 61.4% to 79.5% (95% confidence interval (CI): 73.5–85.6%) and for basal cell carcinoma from 79.4% to 87.1% (95% CI: 80.3–93.9%). AI overconfidence was also reduced while simultaneously maintaining accuracy. Reinforcement learning increased the rate of correct diagnoses made by dermatologists by 12.0% (95% CI: 8.8–15.1%) and improved the rate of optimal management decisions from 57.4% to 65.3% (95% CI: 61.7–68.9%). We further demonstrated that the reward-adjusted reinforcement learning model and a threshold-based model outperformed naïve supervised learning in various clinical scenarios. Our findings suggest the potential for incorporating human preferences into image-based diagnostic algorithms.

ABSTRACTIntroductionTildrakizumab is a monoclonal antibody that targets the p19 subunit of IL-23, a crucial cytokine for Th17 cells. Tildrakizumab has been assessed in several Phase I, II, and III clinical trials and is approved for treatment of adults with moderate to severe plaque psoriasis who are indicated for systemic therapy.Areas coveredThe available evidence on the efficacy, safety, and use of tildrakizumab in special populations was evaluated by 14 experts who critically reviewed the current literature.Expert opinionTildrakizumab has good efficacy that lasts for at least 5 years in patients with moderate to severe psoriasis, and appears to be safe and well tolerated in the long-term with no apparent dose-related differences in adverse events, a low incidence of discontinuation due to adverse events, and no evidence of increased risk of malignancies. The safety and the efficacy of tildrakizumab has also been confirmed in special populations such as those with inflammatory bowel disease, cardiovascular disease, metabolic syndrome, and advanced age. Early intervention with IL-23-inhibitors, such as tildrakizumab, may help to control symptoms and change the long-term course of the disease in patients affected by plaque psoriasis, while improving the quality of life and potentially minimizing the risk of developing comorbidities.

AimsThe diagnosis of metastatic cutaneous melanoma (CM) on lymph node fine needle aspiration samples may be challenging and usually requires confirmation by immunocytochemistry. However, the cytological material could be too scant to order a broad panel of markers. In this case, the pathologist is forced to choose the most advantageous antibodies. The most commonly used melanocytic markers include S100, Melan-A, HMB45 and SOX10 but their diagnostic yield on cytological samples has been poorly studied. The current work aimed to evaluate the diagnostic performance of melanocytic markers when applied to cell blocks obtained from fine needle aspiration cytology (FNAC) of lymph node metastases from CM.MethodsS100, Melan-A, HMB45 and SOX10 were tested on cell block sections of 38 lymphnode metastases from CM diagnosed by cytology. A combined score was built to evaluate each immunostaining, considering the intensity of the staining and the percentage of stained neoplastic cells.ResultsS100 and SOX10 revealed a higher sensitivity (100%) than Melan-A and HMB45 for the diagnosis of metastatic CM. Furthermore, SOX10 emerged as the melanocytic marker with the best staining performance.ConclusionSOX10 has a 100% detection rate and the most easily interpretable staining pattern compared with other melanocytic markers. Therefore, it is strongly recommended that SOX10 is included in the minimal immunocytochemical panel for the diagnostic evaluation of lymph node FNAC in patients with suspected CM metastasis.

Dermoscopy aids in melanoma detection; however, agreement on dermoscopic features, including those of high clinical relevance, remains poor. In this study, we attempted to evaluate agreement among experts on exemplar images not only for the presence of melanocytic-specific features but also for spatial localization. This was a cross-sectional, multicenter, observational study. Dermoscopy images exhibiting at least 1 of 31 melanocytic-specific features were submitted by 25 world experts as exemplars. Using a web-based platform that allows for image markup of specific contrast-defined regions (superpixels), 20 expert readers annotated 248 dermoscopic images in collections of 62 images. Each collection was reviewed by five independent readers. A total of 4,507 feature observations were performed. Good-to-excellent agreement was found for 14 of 31 features (45.2%), with eight achieving excellent agreement (Gwet’s AC >0.75) and seven of them being melanoma-specific features. These features were peppering/granularity (0.91), shiny white streaks (0.89), typical pigment network (0.83), blotch irregular (0.82), negative network (0.81), irregular globules (0.78), dotted vessels (0.77), and blue–whitish veil (0.76). By utilizing an exemplar dataset, a good-to-excellent agreement was found for 14 features that have previously been shown useful in discriminating nevi from melanoma. All images are public (www.isic-archive.com) and can be used for education, scientific communication, and machine learning experiments.

ABSTRACTBackground: The susceptibility of patients with chronic plaque psoriasis and the risks or benefits related to the use of biological therapies for COVID-19 are unknown.Few data about prevalence, clinical course and outcomes of COVID-19 among psoriatic patients were reported. The aims of this study were 1) to assess the prevalence and severity of COVID-19 in psoriatic patients treated with biologic agents during the first phase of the emergency (22 February to 22 April 2020) in Italy, and 2) to report the clinical outcomes of patients who have been exposed to individuals with confirmed SARS-CoV-2 infection.Methods: Patients with moderate-to-severe chronic plaque psoriasis, aged ≥18 years and undergoing treatment with biologic agents as of 22 February 2020, were eligible to be included in PSO-BIO-COVID study. Demographic and clinical characteristics of patients using any biologic for psoriasis treatment between 22 February and 22 April 2020 were registered.Results: A total of 12,807 psoriatic patients were included in the PSO-BIO-COVID study.In this cohort 26 patients (0.2%) had a swab confirmation of SARS-CoV-2 infection. Eleven patients required hospitalization and two died.Conclusion: The incidence of COVID-19 observed in our cohort of psoriatic patients (0.2%) is similar to that seen in the general population (0.31%) in Italy. However, the course of the disease was mild in most patients. Biological therapies may likely lessen ‘cytokine storm’ of COVID-19, which sometimes lead to multiple organ failure, ARDS, and death.

The rapid increase in telemedicine coupled with recent advances in diagnostic artificial intelligence (AI) create the imperative to consider the opportunities and risks of inserting AI-based support into new paradigms of care. Here we build on recent achievements in the accuracy of image-based AI for skin cancer diagnosis to address the effects of varied representations of AI-based support across different levels of clinical expertise and multiple clinical workflows. We find that good quality AI-based support of clinical decision-making improves diagnostic accuracy over that of either AI or physicians alone, and that the least experienced clinicians gain the most from AI-based support. We further find that AI-based multiclass probabilities outperformed content-based image retrieval (CBIR) representations of AI in the mobile technology environment, and AI-based support had utility in simulations of second opinions and of telemedicine triage. In addition to demonstrating the potential benefits associated with good quality AI in the hands of non-expert clinicians, we find that faulty AI can mislead the entire spectrum of clinicians, including experts. Lastly, we show that insights derived from AI class-activation maps can inform improvements in human diagnosis. Together, our approach and findings offer a framework for future studies across the spectrum of image-based diagnostics to improve human–computer collaboration in clinical practice.

IMPORTANCE: The clinical and dermoscopic features of the vast majority of uncommon variants of cutaneous melanoma have been rarely reported, leading to difficulty in making accurate diagnoses. OBJECTIVE: To define the main clinical and dermoscopic features of the most frequently reported but uncommon histologic variants of cutaneous melanoma. EVIDENCE REVIEW: A 2-step systematic review of the literature was performed (from inception to November 2018) using PubMed, Embase, and Cochrane Central Register of Controlled Trials databases. The first step identified those uncommon variants of melanoma for which at least 1 case reporting dermoscopy was described in the literature. The keywords searched were melanoma, uncommon, rare, dermoscopy, and dermatoscopy. In the second step, each previously identified uncommon variant was searched for in the same databases by combining the following terms with melanoma, dermoscopy, and dermatoscopy: amelanotic, hypopigmented, animal, melanocytoma, balloon, desmoplastic, follicular, nested, nevoid, dermal, spitz*, spindle, and verrucous. The institution’s database was also searched from January 2012 to September 2019 for histopathologically confirmed cases of the same melanoma variants. Each reviewer also assessed the quality of reporting in the included articles based on previously described guidelines. FINDINGS: In total, 62 articles met the inclusion criteria, reporting 433 melanoma cases. An additional 56 cases of uncommon melanoma variants were retrieved from the institution’s database for a total of 489 cases: 283 cases of amelanotic superficial spreading melanoma, 18 cases of animal-type and pigmented epithelioid melanocytoma, 7 cases of balloon cell melanoma, 71 cases of desmoplastic melanoma, 3 cases of follicular melanoma, 10 cases of nested melanoma, 33 cases of nevoid melanoma, 2 cases of primary dermal melanoma, 57 cases of spitzoid melanoma, and 5 cases of verrucous melanoma. These variants of melanoma occurred more frequently in women than men (147 cases vs 132 cases). Clinically, these tumors were mainly palpable (162 of 217 [74.7%]) or amelanotic (283 of 489 [57.9%]) lesions that could resemble other benign or malignant skin conditions; dermoscopy typically revealed a homogeneous pinkish background, white structures, and polymorphic vessels. The mean age of all included was 58 years (range, 1-89 years). CONCLUSIONS AND RELEVANCE: Uncommon melanoma variants may resemble both inflammatory disorders and other cutaneous neoplasms, representing a diagnostic pitfall even for the most experienced dermatologist. The purpose of this systematic review was to provide an extensive and detailed overview of specific clinical and dermoscopic features of each uncommon melanoma variant, highlighting the main criteria for differentiating these variants from other benign or malignant skin lesions.

Background: Cutaneous melanoma is one of the most severe skin diseases. Nodular melanoma is the second melanoma subtype in order of frequency. The prognosis of skin melanoma depends on the vertical growth of the tumor (Breslow index). For this measurement, excisional biopsy is strongly recommended. This is, however, an invasive procedure and may cause damage to the lymphatic drainage system. The HFUS system, , can be extremely useful for determining tumor thickness in the preoperative phase, given its high resolution capacity. The aim of this preliminary study is to define the role of HFUS for the nodular skin melanoma Breslow thickness in adults before surgery by making a comparison with histological features. Methods: In this study, 14 melanocytic lesions (8 male and 6 female) were evaluated with dermatoscopic clinical features strongly indicative of nodular melanoma. Out of these, excisional biopsy of 7 lesions was requested. The ultrasounds were performed preoperatively. The images were acquired through the first ultrasound scanner with ultra-high frequency probes (range from 50MHz to 70 MHz) available on the market under the EEC mark (Vevo "MD, FUJIFILM Visual Sonics, Amsterdam, the Netherlands) equipped with a linear probe of 50-70 MHz. Results: From the ultrasonographic analysis of 14 nodular melanoma thickness was determined for the presence of two hyperechogenic laminae, separated by a hypo / anechoic space. The twelve lesions were in situ while the other two lesions showed ultrasonography for example; the satellite lesions (less than two centimeters from the primary lesion) and in transit (localizable to more than two centimeters from the primary lesion). Four of these lesions were ulcerated. A comparsion was made the 7 lesions on between the thickness calculated with this method, and that obtained on the bioptic piece. The presence of a positive concordance has been evident in all of the cases. Conclusion: If further studies are needed to support its widespread clinical use, its is believed that, in expert hands and with an interdisciplinary team, HFUS is already capable to reliably calculate a Breslow index in a large majority of patients with cutaneous melanoma.

ABSTRACTBackground: Long term data on the real-life use of secukinumab are scant. The aim of this study was to investigate the real-life effectiveness, safety and treatment persistence of secukinumab in patients with moderate-to-severe psoriasis.Research design and methods: This 84-week, multicenter (n = 7) retrospective study analyzed data from patients who initiated and received at least 6 months of secukinumab treatment between June 2016 and June 2018 in the Campania region of Italy. Patient demographic and treatment characteristics, duration of treatment and reasons for discontinuation as well as Psoriasis Area and Severity Index (PASI), Body Surface Area (BSA), and Dermatology Life Quality Index (DLQI) scores were assessed.Results: 324 patients (63% male, mean age 50.2 years) were enrolled and received a mean 11.7 months of secukinumab treatment. Overall, 9.5% discontinued secukinumab, including 5.2% who discontinued due to secondary inefficacy and 1.8% due to adverse events. PASI, BSA and DLQI scores were significantly improved from baseline at every follow-up visit (p < 0.001) and mean PASI decreased from 15.3 ± 6.3 at baseline to 0.5 ± 1.0 at week 84. Secukinumab had comparable effectiveness in biologic naïve and non-naïve patients.Conclusions: This study confirmed the effectiveness and safety of secukinumab in real-world patients with psoriasis.

Metastatic melanoma treatment has dramatically changed in the last few years, having a breakthrough with the introduction of targeted agents and immunotherapy. PD-1/PD-L1 pathway is one of the physiologic mechanisms of peripheral immune tolerance, but it also represents a mechanism of tumor immune escape. PD-1/PD-L1 inhibitors represent new immune-checkpoint drugs currently used in metastatic melanoma treatment. Resistance to PD-1/PD-L1 axis blockade, which is the main cause of therapeutic failure during therapeutic use of these drugs, could be linked to several mechanism of immune escape. In fact, other inhibitory receptor such as CTLA-4, LAG-3, TIM-3 and TIGIT might be co-expressed on T cells, deleting the effect of anti-PD-1/PD-L1; overexpression of the enzyme IDO could cause immunosuppression through the depletion of tryptophan in the tumor microenvironment; defective c ostimulation (through reduced activity of 4-1BB and OX40 receptors) could result in T-cell energy. Combination of anti-PD-1/PD-L1 with drugs targeting inhibitory or costimulatory receptors, intracellular pathways, enzymes or neoangiogenesis could be a possible strategy to overcome resistance to single PD-1/PD-L1 blockade. Clinical trials evaluating combination therapies have already showed interesting results, although most of them are still on going.

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 (&lt;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 &lt; .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.

Cutaneous vascular lesions (VLs) represent a very common reason for dermatologic consultation for patients. In most cases, VLs are benign and self-limiting. However, because they often mimic malignant skin tumors, their correct and prompt identification is very important in daily practice. Dermoscopy may play a key role in achieving that purpose. This article reviews current knowledge of dermoscopic features of the most frequent VLs.

IMPORTANCE: The accuracy of melanoma-specific dermoscopic criteria has been tested mainly in studies including invasive tumors. Scarce evidence exists on the usefulness of these criteria for the diagnosis of melanoma in situ (MIS). OBJECTIVE: To investigate the diagnostic accuracy of dermoscopic criteria for the diagnosis of MIS. DESIGN, SETTING, AND PARTICIPANTS: A diagnostic accuracy study with retrospective patient enrollment was conducted in 3 centers specializing in skin cancer diagnosis and management. A total of 1285 individuals with histopathologically diagnosed MIS or other flat, pigmented skin tumors that were histopathologically diagnosed or monitored for at least 1 year were included. Dermoscopic images of MIS and other flat, pigmented skin tumors were evaluated by 3 independent investigators for the presence of predefined criteria. Evaluators were blinded to the clinic dermoscopic and histopathologic diagnosis. MAIN OUTCOMES AND MEASURES: Frequencies of dermoscopic criteria per diagnosis were calculated. Crude odds ratios, adjusted odds ratios, and corresponding 95% CIs were calculated by univariate and multivariate logistic regression, respectively. RESULTS: A total of 1285 patients were included in the study (642 [50%] male); mean age was 45.9 years (range, 9-91 years). Of a total of 1285 lesions obtained from these patients, 325 (25.3%) were MIS; 574 (44.7%) were nevi (312 [24.3%] excised and 262 [20.4%] not excised); 67 (5.2%) were seborrheic keratoses, solar lentigines, or lichen planus–like keratoses; 91 (7.1%) were pigmented superficial basal cell carcinomas; 26 (2.0%) were pigmented intraepithelial carcinomas; 100 (7.8%) were Reed nevi; and 102 (7.9%) were invasive melanomas with a Breslow thickness less than 0.75 mm. The most frequent dermoscopic criteria for MIS were regression (302 [92.9%]), atypical network (278 [85.5%]), and irregular dots and/or globules (163 [50.2%]). The multivariate analysis revealed 5 main positive dermoscopic indicators of MIS: atypical network (3.7-fold; 95% CI, 2.5-5.4), regression (4.7-fold; 95% CI, 2.8-8.1), irregular hyperpigmented areas (5.4-fold; 95% CI, 3.7-8.0), prominent skin markings (3.4-fold; 95% CI, 1.9-6.1), and angulated lines (2.2-fold; 95% CI, 1.2-4.1). When compared only with excised nevi, 2 of these criteria remained potent MIS indicators, namely, irregular hyperpigmented areas (4.3-fold; 95% CI, 2.7-6.8) and prominent skin markings (2.7-fold; 95% CI, 1.3-5.7). CONCLUSIONS AND RELEVANCE: Clinicians should take into consideration the aforementioned dermoscopic indicators for the diagnosis of MIS.

The literature has not been able to conclude whether pregnancy influences the prognosis of melanoma. The aim of this study was to explore the prognosis of melanoma diagnosed during pregnancy or post partum[pregnancy-associated melanoma (PAM)] compared with melanoma in female patients who were not pregnant. We systematically searched for studies of female patients with melanoma that reported outcomes related to survival. Fifteen eligible studies were found. Overall, PAM was associated with a 17% higher mortality compared with melanoma diagnosed in female patients who were not pregnant (hazard ratio=1.17, 95% confidence interval: 1.03–1.33, P=0.02). The heterogeneity associated with this test was moderate (P=0.07; I2=38%). PAM was also associated with a 50% higher recurrence rate compared with melanoma not associated with pregnancy (hazard ratio=1.50, 95% confidence interval: 1.19–1.90, P<0.001). The heterogeneity associated with this test was low (P=0.69; I2=0%). A limitation of this meta-analysis is the definition of PAM, which is not unanimous among the studies included. Our results indicate that PAM is associated with a worse prognosis than melanoma not related to pregnancy, both in terms of overall survival and disease-free survival. On the basis of our data, we anticipate that the survival difference we report here will be further amplified with the addition of future well-carried out studies. We suggest that detection of PAM requires particular awareness by healthcare professionals.

Actinic keratoses (AKs) develop as a consequence of chronic ultraviolet (UV) exposure and exist on a continuum with squamous cell carcinoma (SCC). As one of the most common conditions treated by dermatologists, AK places a significant burden on patients and the healthcare system. A range of treatments are used, including topical treatments that target the visible and subclinical lesions. The goal of such therapies is to achieve complete clearance of AKs and eliminate the risk of progression to SCC. Robust meta-analyses of trial data can provide valuable information for the optimal management of AK and cost-effectiveness evaluations of topical treatments, such as ingenol mebutate gel and diclofenac. These outcomes can facilitate prescribing physicians’ decisions and shape therapeutic guidelines. Peer-reviewed meta-analysis publications and treatment guidelines favoured ingenol mebutate efficacy over diclofenac and the relative cost-effectiveness of ingenol mebutate. We discuss and critique recent evidence, from a cost-effectiveness analysis of 3% diclofenac sodium and ingenol mebutate in the treatment of AK in Italy, which has challenged this view.

ABSTRACTIntroduction: The sudden eruption of melanocytic nevi has been associated with a number of conditions, such as bullous skin diseases, immunodeficiency and immunosuppression. The exact mechanisms leading to the development of eruptive melanocytic nevi are unknown.Areascovered: The aim of this article is to review the literature concerning eruptive melanocytic nevi following the administration of immunosuppressive drugs and other medications.Expertopinion: The literature regarding the development of eruptive nevi in association with pharmacological therapies includes a relatively low number of reports. Prevalence of this phenomenon is likely to be underestimated, thus reporting should be encouraged in order to better define the actual significance and related clinical implications. The development of multiple melanocytic nevi during immunosuppressive treatments highlights the importance of immune system integrity in the regulation of nevi growth. The observation of eruptive nevi as an unexpected effect of targeted therapies for specific types of cancer, including melanoma, provided intriguing hints to understand the mechanisms underlying this paradoxical event. The synergistic role of additional triggers in the occurrence of drug-induced eruptive nevi has not been explored and may be an interesting area of research.

We report the case of an 8-year-old child with subungual exostosis, whose diagnosis was suspected on the basis of dermoscopic findings and subsequently confirmed by X-ray and histopathology.

The field of skin cancer detection offers a compelling use case for the application of artificial intelligence (AI) within the realm of image-based diagnostic medicine. Through the analysis of large datasets, AI algorithms have the capacity to classify clinical or dermoscopic images with remarkable accuracy. Although these AI-based applications can operate both autonomously and under human supervision, the best results are achieved through a collaborative approach that leverages the expertise of both AI and human experts. However, it is important to note that most studies focus on assessing the diagnostic accuracy of AI in artificial settings rather than in real-world scenarios. Consequently, the practical utility of AI-assisted diagnosis in a clinical environment is still largely unknown. Furthermore, there exists a knowledge gap concerning the optimal use cases and deployment settings for these AI systems as well as the practical challenges that may arise from widespread implementation. This review explores the advantages and limitations of AI in a variety of real-world contexts, with a specific focus on its value to consumers, general practitioners, and dermatologists.

IMPORTANCE: Reflectance confocal microscopy (RCM) improves diagnostic accuracy in skin cancer detection when combined with dermoscopy; however, little evidence has been gathered regarding its real impact on routine clinical workflow, and, to our knowledge, no studies have defined the terms for its optimal application. OBJECTIVE: To identify lesions on which RCM performs better in terms of diagnostic accuracy and consequently to outline the best indications for use of RCM. DESIGN, SETTING, AND PARTICIPANTS: Prospectively acquired and evaluated RCM images from consecutive patients with at least 1 clinically and/or dermoscopically equivocal skin lesion referred to RCM imaging, from January 2012 to October 2014, carried out in a tertiary referral academic center. MAIN OUTCOMES AND MEASURES: A total of 1279 equivocal skin lesions were sent for RCM imaging. Spearman correlation, univariate, and multivariate regression models were performed to find features significantly correlated with RCM outcome. RESULTS: In a total of 1279 lesions in 1147 patients, RCM sensitivity and specificity were 95.3% and 83.9%, respectively. The number of lesions needed to excise to rule out a melanoma was 2.4. After univariate and multivariate regression analysis, head and neck resulted as the most appropriate body location for confocal examination; RCM showed a high diagnostic accuracy for lesions located on sun-damaged skin (adjusted odds ratio [aOR], 2.13; 95% CI, 1.37-3.30; P=.001) and typified by dermoscopic regression (aOR, 2.13; 95% CI, 1.31-3.47; P=.002) or basal-cell carcinoma specific criteria (aOR, 9.35; 95% CI, 1.28-68.58; P=.03). CONCLUSIONS AND RELEVANCE: Lesions located on the head and neck, damaged by chronic sun-exposure, and dermoscopically typified by regression represent best indications for the use of RCM.

IMPORTANCE: The comparative diagnostic performance of dermoscopic algorithms and their individual criteria are not well studied. OBJECTIVES: To analyze the discriminatory power and reliability of dermoscopic criteria used in melanoma detection and compare the diagnostic accuracy of existing algorithms. DESIGN, SETTING, AND PARTICIPANTS: This was a retrospective, observational study of 477 lesions (119 melanomas [24.9%] and 358 nevi [75.1%]), which were divided into 12 image sets that consisted of 39 or 40 images per set. A link on the International Dermoscopy Society website from January 1, 2011, through December 31, 2011, directed participants to the study website. Data analysis was performed from June 1, 2013, through May 31, 2015. Participants included physicians, residents, and medical students, and there were no specialty-type or experience-level restrictions. Participants were randomly assigned to evaluate 1 of the 12 image sets. MAIN OUTCOMES AND MEASURES: Associations with melanoma and intraclass correlation coefficients (ICCs) were evaluated for the presence of dermoscopic criteria. Diagnostic accuracy measures were estimated for the following algorithms: the ABCD rule, the Menzies method, the 7-point checklist, the 3-point checklist, chaos and clues, and CASH (color, architecture, symmetry, and homogeneity). RESULTS: A total of 240 participants registered, and 103 (42.9%) evaluated all images. The 110 participants (45.8%) who evaluated fewer than 20 lesions were excluded, resulting in data from 130 participants (54.2%), 121 (93.1%) of whom were regular dermoscopy users. Criteria associated with melanoma included marked architectural disorder (odds ratio [OR], 6.6; 95% CI, 5.6-7.8), pattern asymmetry (OR, 4.9; 95% CI, 4.1-5.8), nonorganized pattern (OR, 3.3; 95% CI, 2.9-3.7), border score of 6 (OR, 3.3; 95% CI, 2.5-4.3), and contour asymmetry (OR, 3.2; 95% CI, 2.7-3.7) (P &lt; .001 for all). Most dermoscopic criteria had poor to fair interobserver agreement. Criteria that reached moderate levels of agreement included comma vessels (ICC, 0.44; 95% CI, 0.40-0.49), absence of vessels (ICC, 0.46; 95% CI, 0.42-0.51), dark brown color (ICC, 0.40; 95% CI, 0.35-0.44), and architectural disorder (ICC, 0.43; 95% CI, 0.39-0.48). The Menzies method had the highest sensitivity for melanoma diagnosis (95.1%) but the lowest specificity (24.8%) compared with any other method (P &lt; .001). The ABCD rule had the highest specificity (59.4%). All methods had similar areas under the receiver operating characteristic curves. CONCLUSIONS AND RELEVANCE: Important dermoscopic criteria for melanoma recognition were revalidated by participants with varied experience. Six algorithms tested had similar but modest levels of diagnostic accuracy, and the interobserver agreement of most individual criteria was poor.

IMPORTANCE: Nevi are among the strongest risk factors for melanoma. However, little is known about the association of many total nevi (TN) or atypical nevi (AN) with tumor thickness. OBJECTIVES: To examine the association between age and the number of TN and AN and to explore whether there was a relationship between TN or AN and tumor thickness, controlling for multiple variables. DESIGN, SETTING, AND PARTICIPANTS: Survey of patients with melanoma at 2 academic sites and an affiliated Veteran Affairs medical center. Participants included 566 patients surveyed within 3 months of diagnosis. Patients were surveyed in the melanoma clinics from May 17, 2006, through March 31, 2009, within 3 months of diagnostic biopsy. The dates of the analysis were April 1, 2015, to August 1, 2015. MAIN OUTCOMES AND MEASURES: Counts of TN and AN were performed at the first visit after diagnosis and were categorized as 0 to 20, 20 to 50, or more than 50 for TN and as 0, 1 to 5, or more than 5 for AN. Tumor thickness was categorized as 2.00 mm or less or as 2.01 mm or greater. All analyses were stratified by patient age (&lt;60 or ≥60 years). Logistic regression was used to test associations, controlling for age, sex, anatomic location of melanoma, institution, histologic subtype, marital status, performance of skin self-examination, number of health care visits in the past year, mode of melanoma discovery, and receipt of skin examination by a physician. RESULTS: The study population included 566 patients. Their mean (SD) age was 56.7 (15.9) years, and 39.0% (n = 221) were female. Of 566 patients, the number of TN was classified as 0 to 20 (66.4% [n = 376]), 20 to 50 (20.5% [n = 116]), or more than 50 (13.1% [n = 74]). Atypical nevus counts were 0 (73.3% [n = 415]), 1 to 5 (14.5% [n = 82]), or more than 5 (12.2% [n = 69]). For those younger than 60 years, the presence of more than 50 TN was associated with a sharply reduced risk of thick melanoma (odds ratio, 0.32; 95% CI, 0.12-0.81), and the presence of more than 5 AN compared with no AN was associated with thicker melanoma (odds ratio, 2.43; 95% CI, 1.02-5.75). CONCLUSIONS AND RELEVANCE: Most patients with melanoma had few nevi and no AN. In younger patients (&lt;60 years), thick melanomas were commonly found in those with fewer TN but more AN, suggesting that physicians and patients should not rely on the total nevus count as a sole reason to perform skin examinations or to determine a patient’s at-risk status. Younger patients should be educated on the increased risk of thicker melanomas that is associated with having more AN.

IMPORTANCE: Incidence rates of skin cancer are increasing globally, and the correct classification of skin lesions (SLs) into benign and malignant tissue remains a continuous challenge. A collective intelligence approach to skin cancer detection may improve accuracy. OBJECTIVE: To evaluate the performance of 2 well-known collective intelligence rules (majority rule and quorum rule) that combine the independent conclusions of multiple decision makers into a single decision. DESIGN, SETTING, AND PARTICIPANTS: Evaluations were obtained from 2 large and independent data sets. The first data set consisted of 40 experienced dermoscopists, each of whom independently evaluated 108 images of SLs during the Consensus Net Meeting of 2000. The second data set consisted of 82 medical professionals with varying degrees of dermatology experience, each of whom evaluated a minimum of 110 SLs. All SLs were evaluated via the Internet. Image selection of SLs was based on high image quality and the presence of histopathologic information. Data were collected from July through October 2000 for study 1 and from February 2003 through January 2004 for study 2 and evaluated from January 5 through August 7, 2015. MAIN OUTCOMES AND MEASURES: For both collective intelligence rules, we determined the true-positive rate (ie, the hit rate or specificity) and the false-positive rate (ie, the false-alarm rate or 1 − sensitivity) and compared these rates with the performance of single decision makers. Furthermore, we evaluated the effect of group size on true- and false-positive rates. RESULTS: One hundred twenty-two medical professionals performed 16 029 evaluations. Use of either collective intelligence rule consistently outperformed single decision makers. The groups achieved an increased true-positive rate and a decreased false-positive rate. For example, individual decision makers in study 1, using the pattern analysis as diagnostic algorithm, achieved a true-positive rate of 0.83 and a false-positive rate of 0.17. Groups of 3 individuals achieved a true-positive rate of 0.91 and a false-positive rate of 0.14. These improvements increased with increasing group size. CONCLUSIONS AND RELEVANCE: Collective intelligence might be a viable approach to increase diagnostic accuracy in skin cancer and reduce skin cancer–related mortality.