Alopecia areata developing paralell to improvement of psoriasis during ustekinumab therapy

This study presents the first application of convolutional neural networks to high-frequency ultrasound skin image classification. This type of imaging opens up new opportunities in dermatology, showing inflammatory diseases such as atopic dermatitis, psoriasis, or skin lesions. We collected a database of 631 images with healthy skin and different skin pathologies to train and assess all stages of the methodology. The proposed framework starts with the segmentation of the epidermal layer using a DeepLab v3+ model with a pre-trained Xception backbone. We employ transfer learning to train the segmentation model for two purposes: to extract the region of interest for classification and to prepare the skin layer map for classification confidence estimation. For classification, we train five models in different input data modes and data augmentation setups. We also introduce a classification confidence level to evaluate the deep model’s reliability. The measure combines our skin layer map with the heatmap produced by the Grad-CAM technique designed to indicate image regions used by the deep model to make a classification decision. Moreover, we propose a multicriteria model evaluation measure to select the optimal model in terms of classification accuracy, confidence, and test dataset size. The experiments described in the paper show that the DenseNet-201 model fed with the extracted region of interest produces the most reliable and accurate results.

Background: Genital warts are the manifestation of the human papillomavirus (HPV) infection, which may last for weeks or months before the clinical presentation. The primary aim of the study was the correlation of the DNA HPV genotypes eradication with the treatment response in male patients with persistent genital warts. Methods: Twenty-one male patients (age range: 22–58) after failure of cryotherapy and podophyllotoxin treatment were enrolled in the study. Genetic tests (Real Time - PCR method) analyzed the presence of DNA-HPV before and 6 months after four sessions (4 weeks apart) of photodynamic therapy with 5-aminolaevulinic acid (ALA-PDT). The treatment efficacy was evaluated before each PDT session and at the end of the study. Results: The single HPV DNA type was present in 15/21 of the patients (13/15 HPV6). The high-risk HPV types were found in 8/21 subjects, of which 6/8 had several types. Six months after four sessions of PDT, complete response was found in 16/21 (76.19%; p = 0.0007) of patients, and DNA HPV clearance was found in 66.67% (p = 0.03). The eradication rate differed among patients with primary low-risk and high-risk HPV types—76.92% (10/13; p = 0.0003) and 50% (4/8; p = 0.05) respectively. Conclusion: ALA-PDT is an effective treatment even after the failure of previous modalities. The persistence of clinical lesions and high oncological risk HPV types should be an indication for treatment prolongation.

Objective: The aim of the study was to verify two hypotheses. The first concerned the possibility of diagnostic dermoscopic differentiation between cutaneous melanomas of the histopathological category in situ (pTis) and thin melanomas (pT1a) in terms of their diameter. The second assessed the diagnostic feasibility of two dermoscopic algorithms aiming to detect ≤ 5.0 mm-sized melanomas histopathologically confirmed as pTis and pT1a. Methods: Dermoscopic images of consecutive cases of histopathologically confirmed melanomas were evaluated by three independent investigators for the presence of the predefined criteria. The melanomas were subdivided according to their diameter into small melanomas, so-called micromelanomas (microM)—sized ≤ 5.0 mm and >5.0 mm, according to published definitions of small melanocytic lesions. The Triage Amalgamated Dermoscopic Algorithm (TADA) and the revisited 7-point checklist of dermoscopy (7-point) algorithm were chosen for the diagnostic feasibility. Odds ratios and corresponding 95% confidence limits (CL) were calculated using the logistic regression adjusted for age for the melanoma-specific dermoscopic structures, the dermoscopic patterns and the diagnostic feasibility of the 7-point checklist and TADA algorithms. The p-values of the results were corrected using the Bonferroni method. Results: In total, 106 patients with 109 melanomas, 50 sized ≤ 5.0 mm and 59 exceeding the diameter of 5.0 mm, were retrospectively analyzed. The prevalent general pattern of microM was the spitzoid one (48% vs. 11.86%, p = 0.0013). Furthermore, 40% of microM vs. 6.78% melanomas sized > 5.0 mm (p = 0.0023) did not present melanoma-specific patterns. The asymmetric multicomponent pattern was present in 64.41% melanomas sized > 5.0 mm and in 26.00% microM (p = 0.0034). The asymmetry of structures or colors was detected in 56% microM vs. 89.83% (p = 0.0020) and 56% microM and 94.92% (p = 0.000034) melanoma sized > 5.0 mm, respectively. The differences in frequency of the detected dermoscopic structures specific to melanomas revealed that microM are almost deprived of negative networks (p = 0.04), shiny white structures (p = 0.0027) and regression features (p = 0.00003). Neither prominent skin markings nor angulated lines were found in the entire study group. Out of the vascular structures, microM presented only dotted (32%) or polymorphous (28%) vessels, although more rarely than melanomas sized > 5.0 mm (66.1% p = 0.017 and 49% p > 0.05, respectively). The diagnostic feasibility revealed a score ≥ 3 of the 7-point algorithm (indicative for malignancy) in 60% microM and 98.31% melanomas sized > 5.0 mm (p = 0.000006). The TADA algorithm revealed melanoma-specific patterns in 64% microM and 96.61% > 5.0 mm-sized melanomas (p = 0.00006) and melanoma-specific structures in 72% and 91.53% (p > 0.05), respectively. Conclusion: In the dermoscopy, 40% of micromelanomas histopathologically staged as pTis and pT1a did not reveal melanoma-specific patterns. Among the general melanocytic patterns, the spitzoid one was the most frequently found in melanomas sized ≤ 5.0 mm. The 7-point checklist and TADA dermoscopic algorithms were helpful in the identification of the majority of melanomas sized ≤ 5.0 mm.

The co‑occurrence of melanoma and non-melanoma skin cancer (NMSC) can lead to increased morbidity. However, there has been limited research into the dermoscopic characteristics of melanomas and clinical factors during co‑occurrence. A total of 264 patients with melanoma, including 63 with NMSC comorbidity, were enrolled in the present study to retrospectively analyse the coexistence of melanoma morphology, as determined by dermoscopic examination, pathological report, tumour location and clinically manifested risk factors. The frequency of solar lentiginosis (SL) was compared between 264 patients with melanoma and 233 patients with NMSC without melanoma. In 83.4% of cases, skin cancer occurred before or concomitantly with the melanoma. The leading indicators of comorbidity were age (median 70 years; P<0.0001) and SL on the trunk and arms (P<0.0001). Melanomas in patients with NMSC comorbidity were significantly more frequently located on the head and neck [P<0.001; Bonferroni adjusted P‑value (P‑adj.)<0.01], then on the trunk, but less frequently occurred on the lower limbs (P<0.05). The dermoscopic multicomponent asymmetric pattern was the predominant pattern in both groups. The most characteristic pattern in the NMSC group was facial melanoma (P<0.005; P‑adj.<0.05); the spitzoid pattern (P<0.001; P‑adj.<0.01) was rare. Dermoscopic regression was more common (P<0.001) in the NMSC group. Regression and the number of nevi were independent of age. Differences in the incidence of SL were evaluated based on the presence of melanoma (P<0.01) and in patients without melanoma based on the presence of squamous cell carcinoma (SCC; P<0.01), multiple basal cell carcinoma (P<0.0001) and multiple SCC (P<0.005). Patients with melanoma were 10 years younger on average compared with patients with NMSC (P<0.0001). The differentiation factors identified in the present study may improve the precision of dermoscopic examinations and potentially lead to modifications in the diagnostic workflow for patients with multiple NMSCs with comorbid melanoma.