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: Age-associated changes in epidermal hydration, pigmentation, thickness and cell renewal influence skin appearance and can lead to laxity, dryness and poor skin tone. The aim of this pilot study was to assess the synergistic effects of a new bipolar radiofrequency plus non-crosslinked hyaluronic acid (HA) mesotherapy protocol compared with radiofrequency alone on skin appearance and markers of epidermal function. Methods: This prospective, single-center, split-face pilot study recruited women aged 25–65 years with dryness and laxity of the facial skin defined by a trans-epidermal water loss (TEWL) value of ≥26 g/m²/h. Subjects were treated with a bipolar radiofrequency device on both sides of the face. This was immediately followed by needle hyaluronic acid (HA) treatment on one side of the face with 2.5 mL of a non-crosslinked HA. Photographic documentation, analysis of epidermal barrier function parameters, and high frequency (HF) ultrasound analysis were performed prior to treatment and at 28 days. Results: Twenty female subjects with a mean age of 46 (range 29 to 54) years and dry and lax facial skin were included. TEWL was reduced and skin hydration improved to a greater extent with the combined radiofrequency plus mesotherapy protocol compared with radiofrequency alone (−5.8% vs. +3.9% and +23.1% vs. +1.0%, respectively). The combined protocol was also associated with greater improvements in melanin (−7.5% vs. −1.5%) and erythema values (−7.2% vs. +3.0%), respectively. Ultrasound measures of epidermal thickness and epidermal density were greater after the combined protocol compared with radiofrequency alone (12.0% vs. 5.6% and 57.7% vs. 7.1%, respectively). Both treatments were well-tolerated. Conclusions: The combined bipolar radiofrequency and HA mesotherapy protocol provided greater improvements in skin hydration, firmness and tone compared with radiofrequency alone. The combination treatment was also associated with greater epidermal thickness and density and increased keratinocyte differentiation suggesting a synergistic effect of both treatments on epidermal homeostasis and barrier function. Both treatments were well-tolerated and led to improvements in facial appearance.