Explainability agreement between dermatologists and five visual explanations techniques in deep neural networks for melanoma AI classification.

Introduction: The use of deep convolutional neural networks for analyzing skin lesion images has shown promising results. The identification of skin cancer by faster and less expensive means can lead to an early diagnosis, saving lives and avoiding treatment costs. However, to implement this technology in a clinical context, it is important for specialists to understand why a certain model makes a prediction; it must be explainable. Explainability techniques can be used to highlight the patterns of interest for a prediction. Methods: Our goal was to test five different techniques: Grad-CAM, Grad-CAM++, Score-CAM, Eigen-CAM, and LIME, to analyze the agreement rate between features highlighted by the visual explanation maps to 3 important clinical criteria for melanoma classification: asymmetry, border irregularity, and color heterogeneity (ABC rule) in 100 melanoma images. Two dermatologists scored the visual maps and the clinical images using a semi-quantitative scale, and the results were compared. They also ranked their preferable techniques. Results: We found that the techniques had different agreement rates and acceptance. In the overall analysis, Grad-CAM showed the best total+partial agreement rate (93.6%), followed by LIME (89.8%), Grad-CAM++ (88.0%), Eigen-CAM (86.4%), and Score-CAM (84.6%). Dermatologists ranked their favorite options: Grad-CAM and Grad-CAM++, followed by Score-CAM, LIME, and Eigen-CAM. Discussion: Saliency maps are one of the few methods that can be used for visual explanations. The evaluation of explainability with humans is ideal to assess the understanding and applicability of these methods. Our results demonstrated that there is a significant agreement between clinical features used by dermatologists to diagnose melanomas and visual explanation techniques, especially Grad-Cam.

Artificial intelligence for skin cancer detection and classification for clinical environment: a systematic review.

Background: Skin cancer is one of the most common forms worldwide, with a significant increase in incidence over the last few decades. Early and accurate detection of this type of cancer can result in better prognoses and less invasive treatments for patients. With advances in Artificial Intelligence (AI), tools have emerged that can facilitate diagnosis and classify dermatological images, complementing traditional clinical assessments and being applicable where there is a shortage of specialists. Its adoption requires analysis of efficacy, safety, and ethical considerations, as well as considering the genetic and ethnic diversity of patients. Objective: The systematic review aims to examine research on the detection, classification, and assessment of skin cancer images in clinical settings. Methods: We conducted a systematic literature search on PubMed, Scopus, Embase, and Web of Science, encompassing studies published until April 4th, 2023. Study selection, data extraction, and critical appraisal were carried out by two independent reviewers. Results were subsequently presented through a narrative synthesis. Results: Through the search, 760 studies were identified in four databases, from which only 18 studies were selected, focusing on developing, implementing, and validating systems to detect, diagnose, and classify skin cancer in clinical settings. This review covers descriptive analysis, data scenarios, data processing and techniques, study results and perspectives, and physician diversity, accessibility, and participation. Conclusion: The application of artificial intelligence in dermatology has the potential to revolutionize early detection of skin cancer. However, it is imperative to validate and collaborate with healthcare professionals to ensure its clinical effectiveness and safety.

Brazilian Medical Survey on Telemedicine since the onset of COVID-19

ABSTRACT Objective The primary aim of this study was to understand the difference in the use of Telemedicine by Brazilian physicians before and after the onset of COVID-19 pandemic and their intention to continue using it post the pandemic period. The secondary objective was to analyze the differences of opinion between physicians in the private and public sectors. Methods We conducted an online medical survey through the SurveyMonkey platform in a large hospital in São Paulo, Brazil, from May to July 2022. Results Three-hundred-and -two physicians responded to the survey. We found that there was a significant increase in the number of physicians who started using Telemedicine in both the public and private sectors (p<0.0001) since the onset of COVID-19 pandemic and that >50% of them intend to continue using Telemedicine in their daily practice. Most responders consider that Telemedicine is useful in screening, diagnosis and management of patients; that it facilitates the physician's daily practice; that it can maintain or even add financial gains through reducing office expenses; and that is another medium for dispensing medical care. They also believe that Telemedicine should be regulated by the Brazilian Federal Council of Medicine. There were no significant differences between the responses from private and public sector physicians. Conclusion Telemedicine has played a major role in healthcare since the onset of COVID-19 pandemic and most of the physicians approve its use and intend to continue using Telemedicine in their daily practice.

Diagnosis and treatment of systemic mastocytosis in Brazil: Recommendations of a multidisciplinary expert panel

ABSTRACT Introduction: Systemic Mastocytosis comprises a group of neoplastic diseases characterized by clonal expansion and infiltration of mast cells into several organs. The diagnosis and treatment of this disease may be challenging for non-specialists. Objective: Make suggestions or recommendations in Systemic Mastocytosis based in a panel of Brazilian specialists. Method and results: An online expert panel with 18 multidisciplinary specialists was convened to propose recommendations on the diagnosis and treatment of Systemic Mastocytosis in Brazil. Recommendations were based on discussions of topics and multiple-choice questions and were graded using the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence Chart. Conclusion: Twenty-two recommendations or suggestions were proposed based on a literature review and graded according to the findings.

Diagnosis and treatment of systemic mastocytosis in Brazil: Recommendations of a multidisciplinary expert panel.

INTRODUCTION: Systemic Mastocytosis comprises a group of neoplastic diseases characterized by clonal expansion and infiltration of mast cells into several organs. The diagnosis and treatment of this disease may be challenging for non-specialists. OBJECTIVE: Make suggestions or recommendations in Systemic Mastocytosis based in a panel of Brazilian specialists. METHOD AND RESULTS: An online expert panel with 18 multidisciplinary specialists was convened to propose recommendations on the diagnosis and treatment of Systemic Mastocytosis in Brazil. Recommendations were based on discussions of topics and multiple-choice questions and were graded using the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence Chart. CONCLUSION: Twenty-two recommendations or suggestions were proposed based on a literature review and graded according to the findings.

Implementation of artificial intelligence algorithms for melanoma screening in a primary care setting.

Skin cancer is currently the most common type of cancer among Caucasians. The increase in life expectancy, along with new diagnostic tools and treatments for skin cancer, has resulted in unprecedented changes in patient care and has generated a great burden on healthcare systems. Early detection of skin tumors is expected to reduce this burden. Artificial intelligence (AI) algorithms that support skin cancer diagnoses have been shown to perform at least as well as dermatologists' diagnoses. Recognizing the need for clinically and economically efficient means of diagnosing skin cancers at early stages in the primary care attention, we developed an efficient computer-aided diagnosis (CAD) system to be used by primary care physicians (PCP). Additionally, we developed a smartphone application with a protocol for data acquisition (i.e., photographs, demographic data and short clinical histories) and AI algorithms for clinical and dermoscopic image classification. For each lesion analyzed, a report is generated, showing the image of the suspected lesion and its respective Heat Map; the predicted probability of the suspected lesion being melanoma or malignant; the probable diagnosis based on that probability; and a suggestion on how the lesion should be managed. The accuracy of the dermoscopy model for melanoma was 89.3%, and for the clinical model, 84.7% with 0.91 and 0.89 sensitivity and 0.89 and 0.83 specificity, respectively. Both models achieved an area under the curve (AUC) above 0.9. Our CAD system can screen skin cancers to guide lesion management by PCPs, especially in the contexts where the access to the dermatologist can be difficult or time consuming. Its use can enable risk stratification of lesions and/or patients and dramatically improve timely access to specialist care for those requiring urgent attention.

Accuracy of Deep Neural Network in Triaging Common Skin Diseases of Primary Care Attention.

Access to dermatological care can be challenging in certain regions of the world. The triage process is usually conducted by primary care physicians; however, they may not be able to diagnose and assign the correct referral and level of priority for different dermatosis. The present research aimed to test different deep neural networks to obtain the highest level of accuracy for the following: (1) diagnosing groups of dermatoses; (2) correct referrals; and (3) the level of priority given to the referral compared to dermatologists. Using 140,446 images from a teledermatology project, previously labeled with the clinical diagnosis, and their respective referrals, namely biopsy, in-person dermatologist visits or monitoring the case via teledermatology along with the general physician, 27 different scenarios of neural networks were derived, and the algorithm accuracies in classifying different dermatosis, according to the group of the diagnosis they belong to, were calculated. The most accurate algorithm was then tested for accuracy in diagnosis, referral, and level of priority given to 6,945 cases. The GoogLeNet architecture, trained with 24,000 images and 1,000 epochs, using weight random initialization and learning rates of 10-3 was found to be the most accurate network, showing an accuracy of 89.72% for diagnosis, 96.03% for referrals and 92.54% for priority level in 6,975 image testing. Our study population, however, was confined to individuals with chronic skin conditions and, therefore, it has limited value as a triage tool because it has not been tested for acute conditions. Deep neural networks are accurate in triaging, correct referral and prioritizing common chronic skin diseases related to primary care attention. They can also help health-care systems optimize patients' access to dermatologists.

Dermatologists' perceptions on the utility and limitations of teledermatology after examining 55,000 lesions.

INTRODUCTION: Few studies have assessed the perception of teledermatologists about the utility and limitations of teledermatology, especially to diagnose a broad range of skin diseases. This study aimed to evaluate dermatologists' confidence in teledermatology, its utility and limitations for dermatological conditions in primary care. METHODS: An analytical study that used a survey for dermatologists who diagnosed 30,916 patients with 55,012 lesions through teledermatology during a 1-year project in São Paulo, Brazil. RESULTS: Dermatologists found teledermatology useful for triage and diagnosis, especially for xerotic eczema, pigmentary disorders and superficial infections. Their confidence in teledermatology was statistically higher by the end of the project (p = 0.0012). Limitations included some technical issues and the impossibility to suggest how soon the patient should be assisted face-to-face by a dermatologist. The most treatable group of diseases by teledermatology was superficial infections (92%). The use of dermoscopy images would significantly increase the confidence to treat atypical naevi and malignant tumours (p < 0.0001 and p = 0.0003 respectively). Follow-ups by teledermatology or feedback from primary-care physicians would be desirable, according to the dermatologists. DISCUSSION: We found it interesting that dermatologists became increasingly confident in teledermatology after the project and how they classified teledermatology as useful for triage, diagnosis and even treatment of most types of skin conditions followed at primary care. Dermoscopy should definitely be added to the photographs, especially for malignant tumours and atypical naevi. Most of the technical limitations found could be solved with a few improvements in the software/platform.

Efficacy and safety of dupilumab in two adolescents with severe atopic dermatitis / Eficácia e segurança do uso de dupilumabe em dois adolescentes com dermatite atópica grave

ABSTRACT We report the cases of two adolescent siblings with severe atopic dermatitis, who, despite weighing approximately 40kg, presented a good response to dupilumab with the off-label dose for individuals aged 12 years and weighing 60kg. Both had already used cyclosporine, azathioprine, methotrexate and oral corticosteroids for long periods, plus topical treatments with no adequate disease control. Skin lesions were constant and widespread, with frequent skin infections and very poor quality of life, with numerous physical and psychosocial consequences, such as dropping out of school activities due to severe itching, appearance and bullying. They also showed delayed growth and development. In 2018, dupilumab, an immunobiological agent, was approved for treatment of moderate to severe atopic dermatitis in adults and, in 2019, extended to the 12-17-year age group. Although it had already been approved by the Brazilian Health Surveillance Agency, the 200mg presentation (indicated for the weight of patients) was not available, with no expected arrival date. Therefore, weighing the risks and benefits of the situation of both, we chose to treat them with an adult dose (loading dose of 600mg subcutaneously, and 300mg subcutaneously every 2 weeks) despite the low weight. So far, they have received eight injections, showing significant improvement of disease and quality of life. There were no major adverse effects, only worsening of allergic conjunctivitis in one of them. The patients and their family are very satisfied, and we believe that the therapy has been successful.

RESUMO Relatamos os casos de dois irmãos adolescentes com dermatite atópica grave e que, apesar de pesarem cerca de 40kg, apresentaram boa resposta ao dupilumabe com a dose off-label para indivíduos com 12 anos e peso de 60kg. Ambos já tinham usado ciclosporina, azatioprina, metotrexato e corticoide oral por longos períodos, acrescidos de tratamentos tópicos sem controle adequado da doença. As lesões cutâneas eram constantes e disseminadas, e os irmãos apresentavam infeções de pele frequentes e qualidade de vida muito ruim, com inúmeras consequências físicas e psicossociais, como o abandono da atividade escolar pelo prurido intenso, pela aparência e pelo bullying sofrido. Apresentavam também retardo de crescimento e de desenvolvimento. Em 2018, o dupilumabe, um agente imunobiológico, foi aprovado para o tratamento de dermatite atópica moderada a severa para adultos e, em 2019, ampliado para faixa etária de 12 a 17 anos. Embora já tivesse a aprovação da Agência Nacional de Vigilância Sanitária no Brasil, a apresentação de 200mg (indicada para o peso dos pacientes) não estava disponível, sem previsão de chegada. Assim, pesando os riscos e benefícios da situação de ambos, optamos por tratá-los com dose de adulto (ataque de 600mg por via subcutânea e 300mg por via subcutânea a cada 2 semanas) apesar do baixo peso. Até o momento, eles realizaram oito aplicações, apresentando importante melhora da doença e da qualidade de vida. Não houve efeitos adversos importantes - apenas a piora da conjuntivite alérgica em um deles. Os pacientes e sua família estão muito satisfeitos, e nós avaliamos que a terapia está sendo bem-sucedida.

Part II: Accuracy of Teledermatology in Skin Neoplasms.

Teledermatology has been proving to be of great help for delivering healthcare, especially now, during the SARS-CoV-2 pandemic. It is crucial to assess how accurate this method can be for evaluating different dermatoses. Such knowledge can contribute to the dermatologists' decision of whether to adhere to teledermatology or not. Our objective was to determine the accuracy of teledermatology in the 10 most frequent skin neoplasms in our population, comparing telediagnosis to histopathological report and in-person dermatologists' diagnosis. A retrospective cohort study was conducted in São Paulo, Brazil, where a store-and-forward teledermatology project was implemented under primary-care attention to triage surgical, more complex, or severe dermatoses. A total of 30,976 patients presenting 55,012 lesions took part in the project. Thirteen teledermatologists who participated in the project had three options to refer the patients: send them directly to biopsy, to the in-person dermatologist, or back to the general physician with the most probable diagnosis and management. In the groups referred to the in-person dermatologist and biopsy, we looked for the 10 most frequent International Statistical Classification of Diseases and Related Health Problems-10 (ICD-10) of skin neoplasms, which resulted in 289 histopathologic reports and 803 in-person dermatologists' diagnosis. We were able to compare the ICD-10 codes filled by teledermatologists, in-person dermatologists, and from histopathological reports. The proportion of complete, partial, and no agreement rates between the in-person dermatologist's, histopathologic report, and the teledermatologist's diagnosis was assessed. We also calculated Cohen's kappa, for complete and complete plus partial agreement. The mean complete agreement rate comparing telediagnosis to histopathological report was 54% (157/289; kappa = 0.087), being the highest for malign lesions; to in-person dermatologists was 61% (487/803; kappa = 0.213), highest for benign lesions. When accuracy of telediagnosis for either malign or benign lesions was evaluated, the agreement rate with histopathology was 70% (kappa = 0.529) and with in-person dermatologist, 81% (kappa = 0.582). This study supports that teledermatology for skin neoplasms has moderate accuracy. This result reassures that it can be a proper option for patient care, especially when the goal is to differentiate benign from malign lesions.

Part I: Accuracy of Teledermatology in Inflammatory Dermatoses.

Teledermatology is assuming a progressively greater role as a healthcare delivery method, especially now, during this pandemic time. It is important to know how accurate this tool is for different skin diseases. Most of the studies have focused on skin neoplasms or general dermatology. Studies based on a large number of inflammatory dermatoses have not yet been performed. Such knowledge can help dermatologists to decide whether endorsing this method or not. Our objective was to determine the accuracy of teledermatology in inflammatory dermatoses in a robust number of cases. A retrospective cohort study was conducted in São Paulo, Brazil, from July 2017-18, where a store-and-forward Teledermatology project was implemented under primary-care attention to triage surgical, more complex, or severe dermatoses. A total of 30,976 patients presenting 55,012 lesions took part in the project. Thirteen participating teledermatologists had three options to refer the patients: directly to biopsy, to the in-person dermatologist or back to the general physician with most probable diagnosis and management. In the group referred to the in-person dermatologist, we looked for the 20 most frequent International Classification of Diseases and Related Health Problems- 10th revision (ICD-10) of inflammatory dermatoses, which resulted in 739 patients and 739 lesions. As patients had been triaged by teledermatology previously, we were able to compare ICD-10 codes filled both by teledermatogists and by in-person dermatologists. The proportion of complete, partial, and no agreement rates between the in-person dermatologist's and the teledermatologist's diagnoses was used for accuracy. We also calculated Cohen's kappa, a statistical measure of inter-rater agreement, for complete agreement. The mean complete agreement rate for all twenty dermatoses was 78% (31-100%) and kappa = 0.743; partial agreement 8%; and no agreement 14%, presenting variability according to the disease. Our study showed that teledermatology for inflammatory dermatoses has a high accuracy. This result reassures that it can be a proper option for patient care.

Nodular primary cutaneous melanoma is associated with PD-L1 expression.

BACKGROUND: In previous studies, patients with Stage III melanomas expressing PD-L1 in more than 5% of their neoplastic cells had improved recurrence-free survival with anti-PD1 adjuvant therapy. OBJECTIVES: We examined PD-L1 expression as a possible biomarker of primary cutaneous melanomas in the vertical growth phase. MATERIALS AND METHODS: This was a retrospective study including 66 patients with invasive primary cutaneous melanomas. We assessed patient clinical and histopathological data and performed immunohistochemical assays with melanoma specimens from the patients to evaluate PD-L1, PD-1, CD3, CD8 and FoxP3 expression. RESULTS: We observed PD-L1 expression in 21% (14/66) of our samples, and this expression correlated with increased melanoma thickness (p = 0.002) and nodular-type melanoma (p = 0.001). After adjusting for tumor thickness using a logistic regression test, the association of PD-L1 with nodular-type melanoma persisted. Nodular-type melanoma was 6.48 times more likely to be positive for PD-L1 than other histological types (p = 0.014; 95% CI: 1.46-28.82). As expected, PD-L1 expression correlated with the number of PD-1-expressing cells in the tumor-infiltrating lymphocyte population (p = 0.04). No correlation with PD-L1 was observed for age, sex, tumor site, skin phototype, ulceration status, sentinel lymph node status, metastasis development or survival. Regarding the immune profile of the tumor-infiltrating lymphocytes of PD-L1-positive and -negative groups, no significant differences were observed in the numbers of CD3 + , CD8 + FoxP3-, CD8-FoxP3+ and CD8 + FoxP3+ cells by immunohistochemistry. CONCLUSION: Nodular-type melanoma is associated with PD-L1 expression and may be a suitable candidate for adjuvant therapy of primary melanomas treated with immunotherapy.

Escabiose mascarada por mastocitose sistêmica / Scabies masked by systemic mastocytosis

Paciente do sexo feminino, com 59 anos de idade, portadora de mastocitose sistêmica há 20 anos. A mastocitose é doença rara, caracterizada pela proliferação excessiva e o subsequente acúmulo de mastócitos em órgãos e tecidos, principalmente na medula óssea, pele e no trato gastrointestinal. Há 1 mês, relatava história de novas lesões cutâneas caracterizadas por pápulas e placas eritemato-edematosas com escoriação e intenso prurido. Feito o raspado da pele com confirmação diagnóstica de escabiose.

A 59-year-old female patient had a diagnosis of systemic mastocytosis for 20 years. Mastocytosis is a rare disease characterized by excessive proliferation and accumulation of mast cells in organs and tissues, especially in the bone marrow, skin and gastrointestinal tract. She reported new skin lesions characterized by erythematous papules and plaques with excoriation and intense itching for one month. Skin scraping confirmed the diagnosis of scabies.

Worse survival of invasive melanoma patients in men and "de novo" lesions.

BACKGROUND: The incidence and mortality of melanoma is increasing in many countries, including Brazil. Survival studies are still scarce in our country, but much needed to know and address this problem better. OBJECTIVE: To analyze the disease-specific survival of patients with invasive melanoma and to correlate it with clinical and histopathological variables. METHODS: Retrospective cohort analysis of 565 cases of invasive melanoma in a tertiary hospital with the objective of testing variables that could be associated with a worse prognosis, such as gender, phototype, thickness, histological type and presence of pre-existing clinical lesion at the site of the tumor. RESULTS: The worst survival rates were significantly associated with thicker tumors (p<0.001), male sex (p=0.014), high phototype (p=0.047), nodular melanoma (p=0.024) and "de novo" lesions (p=0.005). When all variables were adjusted for melanoma thickness, male patients (p=0.011) and "de novo" melanomas (p=0.025) remained associated with worse survival. STUDY LIMITATIONS: Retrospective study of a single tertiary hospital. CONCLUSIONS: Although the causes are still unknown, melanoma-specific survival was statistically worse for males and for "de novo" melanomas even after adjustment of tumor thickness.

Accuracy and efficiency of telemedicine in atopic dermatitis.

BACKGROUND: Telemedicine provides accurate diagnoses for skin disorders and has gained emphasis. It may be used for the triage and management of common skin diseases in primary care, improving patients' access and reducing time to treatment. OBJECTIVE: To evaluate the proportion of atopic dermatitis patients who could be managed with the support of telemedicine and its accuracy. Second, we aimed to assess the frequency of atopic dermatitis, demographics, clinical features, and therapies dispensed in relation to the disease. METHODS: Retrospective study in a population of 30,976 individuals, assisted by telemedicine. We assessed patients with the diagnosis of atopic dermatitis and evaluated the proportion of cases referred to biopsy, in-person dermatologists, or to be managed by primary care; the treatments suggested; and telemedicine accuracy to diagnose atopic dermatitis. RESULTS: Atopic dermatitis was diagnosed in 1648 patients (5.3%), the sixth most common dermatosis, with 2058 lesions (3.7%) analyzed. Primary care physicians were able to manage 72% of the atopic dermatitis patients, whereas 28% of them were referred to in-person dermatologists. Accuracy for atopic dermatitis diagnosis was 84.4%. CONCLUSION: Telemedicine was an accurate method and helped primary care physicians to treat 72% of the atopic dermatitis lesions, thereby optimizing the availability of in-person appointments with dermatologists for more severe cases.