Comprehensive analysis of clinical images contributions for melanoma classification using convolutional neural networks.

BACKGROUND: Timely diagnosis plays a critical role in determining melanoma prognosis, prompting the development of deep learning models to aid clinicians. Questions persist regarding the efficacy of clinical images alone or in conjunction with dermoscopy images for model training. This study aims to compare the classification performance for melanoma of three types of CNN models: those trained on clinical images, dermoscopy images, and a combination of paired clinical and dermoscopy images from the same lesion. MATERIALS AND METHODS: We divided 914 image pairs into training, validation, and test sets. Models were built using pre-trained Inception-ResNetV2 convolutional layers for feature extraction, followed by binary classification. Training comprised 20 models per CNN type using sets of random hyperparameters. Best models were chosen based on validation AUC-ROC. RESULTS: Significant AUC-ROC differences were found between clinical versus dermoscopy models (0.661 vs. 0.869, p < 0.001) and clinical versus clinical + dermoscopy models (0.661 vs. 0.822, p = 0.001). Significant sensitivity differences were found between clinical and dermoscopy models (0.513 vs. 0.799, p = 0.01), dermoscopy versus clinical + dermoscopy models (0.799 vs. 1.000, p = 0.02), and clinical versus clinical + dermoscopy models (0.513 vs. 1.000, p < 0.001). Significant specificity differences were found between dermoscopy versus clinical + dermoscopy models (0.800 vs. 0.288, p < 0.001) and clinical versus clinical + dermoscopy models (0.650 vs. 0.288, p < 0.001). CONCLUSION: CNN models trained on dermoscopy images outperformed those relying solely on clinical images under our study conditions. The potential advantages of incorporating paired clinical and dermoscopy images for CNN-based melanoma classification appear less clear based on our findings.

Pyoderma Gangrenosum of the Genitalia, Anus, and Perineum: Two Case Reports and a Review of Published Cases.

ABSTRACT: Pyoderma gangrenosum is an inflammatory skin disease that presents with rapidly progressive ulcers with violaceous, undermined borders. Despite most commonly affecting the lower extremities, pyoderma gangrenosum can rarely present in the genital, anal, and perineal regions. We describe two cases and report a review of published cases.

Perioperative management and clinical outcomes of peristomal pyoderma gangrenosum.

Peristomal pyoderma gangrenosum is an uncommon subtype of pyoderma gangrenosum mainly affecting stoma sites of patients with inflammatory bowel disease. While surgical treatments are often used to assist healing, little is known about the relationship between surgical interventions and the rate of recurrence of peristomal pyoderma gangrenosum. The aim of this study was to identify patient and clinical factors associated with peristomal pyoderma gangrenosum recurrence following surgical intervention. A multi-institutional retrospective case series and literature review was conducted to evaluate patient characteristics and perioperative treatment. Patients of any age with peristomal pyoderma gangrenosum undergoing surgical operations related to their pyoderma gangrenosum or due to another comorbidity were included. Descriptive statistics were used to characterize demographic information. Associations were evaluated using Wilcoxon's rank-sum test for continuous variables and Fisher's exact test for categorical data. Thirty-seven cases were included, 78.3% of which had a history of inflammatory bowel disease. Overall, 13 (35.1%) cases experienced recurrence at 30 days. There was no significant association identified between patient demographics, stoma location, surgical intervention, or perioperative treatment with rate of recurrence at 30 days post-operation. While no clinical risk factors or treatments were associated with recurrence, our work underscores the importance of a multidisciplinary approach to this disease to address gastrointestinal, dermatologic, and surgical components of treatment.

Association of vitiligo with multiple cutaneous and extra-cutaneous autoimmune diseases: a nationwide cross-sectional study.

Previous studies found conflicting results about associations of vitiligo with different autoimmune diseases. To evaluate associations of vitiligo with multiple autoimmune diseases. A cross-sectional study representative of 612,084,148 US patients from the Nationwide Emergency Department Sample (NEDS) 2015-2019 was performed. Vitiligo and autoimmune diseases were identified using International Classification of Diseases-10 codes. The most frequent autoimmune disorders in patients with vitiligo were type 1 diabetes, rheumatoid arthritis, systemic lupus erythematosus (SLE), autoimmune thyroiditis, Addison's disease, and systemic sclerosis (SSc). Vitiligo was associated with any autoimmune disorder (adjusted odds ratio [95% confidence interval] 1.45 [1.32-1.58]). Cutaneous disorders with largest effect-sizes were alopecia areata (186.22 [115.31-300.72]) and SSc (32.13 [25.28-40.82]). Non-cutaneous comorbidities with largest effect-sizes were primary sclerosing cholangitis (43.12 [18.98-97.99]), pernicious anemia (41.26 [31.66-53.78]), Addison's disease (33.85 [26.68-42.9]), and autoimmune thyroiditis (31.65 [26.34-38.02]). Vitiligo is associated with multiple cutaneous and non-cutaneous autoimmune diseases, especially in females and older age.

Association of Adult Atopic Dermatitis Severity With Bacterial, Viral, and Fungal Skin Infections.

Background: Little is known about the relationship of atopic dermatitis (AD) severity, phenotype, and persistence on different types of skin infections. Objective: To evaluate the relationship of AD characteristics and skin infections over time in adults. Methods: We performed a prospective dermatology practice-based study (n = 559). History of infection was assessed using questionnaires. AD severity was evaluated using Scoring Atopic Dermatitis (SCORAD), Eczema Area and Severity Index (EASI), Investigator's Global Assessment (IGA), and Patient-reported Global Assessment (PtGA). Results: At baseline, 160 (21.4%) patients reported history of ≥1 skin infection, including 14.3% with bacterial infections. In multivariable repeated measures logistic regression models, ≥1 cutaneous infection was associated with moderate (adjusted odds ratio [95% confidence interval]: 2.67 [1.67-4.28]) and severe (6.35 [3.36-12.01]) versus mild SCORAD; as well as severe SCORAD-itch; moderate and severe versus clear-mild EASI; moderate and severe versus clear-mild PtGA; mild, moderate, and severe versus clear-almost clear IGA. Cutaneous infections were not associated with ichthyosis, palmar hyperlinearity, nummular eczema, cheilitis, or hand eczema. Specific infections varied by AD severity and body site. Persistent moderate-severe disease was associated with higher odds of skin infection. Conclusion: Skin infections were associated with AD severity but not phenotype, and may be mitigated by improved AD severity.