Deep Learning-Based Evaluation of Ultrasound Images for Benign Skin Tumors.

In this study, a combined convolutional neural network for the diagnosis of three benign skin tumors was designed, and its effectiveness was verified through quantitative and statistical analysis. To this end, 698 sonographic images were taken and diagnosed at the Department of Dermatology at Severance Hospital in Seoul, Korea, between 10 November 2017 and 17 January 2020. Through an empirical process, a convolutional neural network combining two structures, which consist of a residual structure and an attention-gated structure, was designed. Five-fold cross-validation was applied, and the train set for each fold was augmented by the Fast AutoAugment technique. As a result of training, for three benign skin tumors, an average accuracy of 95.87%, an average sensitivity of 90.10%, and an average specificity of 96.23% were derived. Also, through statistical analysis using a class activation map and physicians' findings, it was found that the judgment criteria of physicians and the trained combined convolutional neural network were similar. This study suggests that the model designed and trained in this study can be a diagnostic aid to assist physicians and enable more efficient and accurate diagnoses.

A nomogram combining clinical factors and biomarkers for predicting the recurrence of high-risk cutaneous squamous cell carcinoma.

BACKGROUND: Although determining the recurrence of cutaneous squamous cell carcinoma (cSCC) is important, currently suggested systems and single biomarkers have limited power for predicting recurrence. OBJECTIVE: In this study, combinations of clinical factors and biomarkers were adapted into a nomogram to construct a powerful risk prediction model. METHODS: The study included 145 cSCC patients treated with Mohs micrographic surgery. Clinical factors were reviewed, and immunohistochemistry was performed using tumor tissue samples. A nomogram was constructed by combining meaningful clinical factors and protein markers. RESULTS: Among the various factors, four clinical factors (tumor size, organ transplantation history, poor differentiation, and invasion into subcutaneous fat) and two biomarkers (Axin2 and p53) were selected and combined into a nomogram. The concordance index (C-index) of the nomogram for predicting recurrence was 0.809, which was higher than that for the American Joint Committee on Cancer (AJCC) 7th, AJCC 8th, Brigham and Women's Hospital, and Breuninger staging systems in the patient data set. CONCLUSION: A nomogram model that included both clinical factors and biomarkers was much more powerful than previous systems for predicting cSCC recurrence.

High-speed combined reflectance confocal and moxifloxacin based two-photon microscopy.

Reflectance confocal microscopy (RCM) is a non-invasive high-resolution optical imaging technique used in clinical settings as a diagnostic method. However, RCM has limited diagnostic ability by providing non-specific morphological information only based on reflection contrast. Various multimodal imaging techniques have been developed to compensate the limitations of RCM, but multimodal techniques are often slow in imaging speed compared to RCM alone. In this report, we combined RCM with moxifloxacin based two-photon microscopy (TPM) for high-speed multimodal imaging. Moxifloxacin based TPM used clinically compatible moxifloxacin for cell labeling and could do non-invasive cellular imaging at 30 frames/s together with RCM. Performance of the combined microscopy was characterized in the imaging of mouse skin and cornea, in vivo. Detail tissue microstructures including cells, extra-cellular matrix (ECM), and vasculature were visualized. The combined microscopy was applied to human skin cancer specimens, and both cells and ECM in the skin cancer and normal skin regions were visualized at high imaging speeds. The combined microscopy can be useful in the clinical applications of RCM by providing multiple contrasts.

Correction: Acral melanoma detection using a convolutional neural network for dermoscopy images.

[This corrects the article DOI: 10.1371/journal.pone.0193321.].

Acral melanoma detection using a convolutional neural network for dermoscopy images.

BACKGROUND/PURPOSE: Acral melanoma is the most common type of melanoma in Asians, and usually results in a poor prognosis due to late diagnosis. We applied a convolutional neural network to dermoscopy images of acral melanoma and benign nevi on the hands and feet and evaluated its usefulness for the early diagnosis of these conditions. METHODS: A total of 724 dermoscopy images comprising acral melanoma (350 images from 81 patients) and benign nevi (374 images from 194 patients), and confirmed by histopathological examination, were analyzed in this study. To perform the 2-fold cross validation, we split them into two mutually exclusive subsets: half of the total image dataset was selected for training and the rest for testing, and we calculated the accuracy of diagnosis comparing it with the dermatologist's and non-expert's evaluation. RESULTS: The accuracy (percentage of true positive and true negative from all images) of the convolutional neural network was 83.51% and 80.23%, which was higher than the non-expert's evaluation (67.84%, 62.71%) and close to that of the expert (81.08%, 81.64%). Moreover, the convolutional neural network showed area-under-the-curve values like 0.8, 0.84 and Youden's index like 0.6795, 0.6073, which were similar score with the expert. CONCLUSION: Although further data analysis is necessary to improve their accuracy, convolutional neural networks would be helpful to detect acral melanoma from dermoscopy images of the hands and feet.

Dermoscopy guided dark-field multi-functional optical coherence tomography.

Dermoscopy is a skin surface microscopic technique allowing specular reflection free observation of the skin, and has been used to examine pigmented skin lesions. However, dermoscopy has limitations in providing depth information due to lack of 3D resolution. In order to overcome the limitations, we developed dermoscopy guided multi-functional optical coherence tomography (MF-OCT) providing both high-contrast superficial information and depth-resolved structural, birefringent, and vascular information of the skin simultaneously. Dermoscopy and MF-OCT were combined by using a dichroic mirror, and dark-field configuration was adapted for MF-OCT to reduce specular reflection. After characterization, dermoscopy guided MF-OCT was applied to several human skin lesions such as the scar, port-wine stain (PWS) as well as the normal skin for demonstration. Various features of the scar and PWS were elucidated by both dermoscopy and MF-OCT. Dermoscopy guided MF-OCT may be useful for evaluation and treatment monitoring of skin lesions in clinical applications.

Sequential Change of Wound Calculated by Image Analysis Using a Color Patch Method during a Secondary Intention Healing.

BACKGROUND: Photographs of skin wounds have the most important information during the secondary intention healing (SIH). However, there is no standard method for handling those images and analyzing them efficiently and conveniently. OBJECTIVE: To investigate the sequential changes of SIH depending on the body sites using a color patch method. METHODS: We performed retrospective reviews of 30 patients (11 facial and 19 non-facial areas) who underwent SIH for the restoration of skin defects and captured sequential photographs with a color patch which is specially designed for automatically calculating defect and scar sizes. RESULTS: Using a novel image analysis method with a color patch, skin defects were calculated more accurately (range of error rate: -3.39% ~ + 3.05%). All patients had smaller scar size than the original defect size after SIH treatment (rates of decrease: 18.8% ~ 86.1%), and facial area showed significantly higher decrease rate compared with the non-facial area such as scalp and extremities (67.05 ± 12.48 vs. 53.29 ± 18.11, P < 0.05). From the result of estimating the date corresponding to the half of the final decrement, all of the facial area showed improvements within two weeks (8.45 ± 3.91), and non-facial area needed 14.33 ± 9.78 days. CONCLUSION: From the results of sequential changes of skin defects, SIH can be recommended as an alternative treatment method for restoration with more careful dressing for initial two weeks.

The Investigation on the Distribution of Malassezia Yeasts on the Normal Korean Skin by 26S rDNA PCR-RFLP.

BACKGROUND: Malassezia yeasts are normal flora of the skin that are discovered in 75~98% of health subjects, but since its association with various skin disorders have been known, many studies have been conducted in the distribution of the yeasts. OBJECTIVE: To isolate, identify, and classify Malassezia yeasts from the normal human skin of Koreans by using the rapid and accurate molecular biology method (26S rDNA PCR-RFLP) which overcome the limits of morphological and biochemical methods, and to gather a basic database that will show its relation to various skin diseases. METHODS: Malassezia yeasts were cultured from clinically healthy human skin using scrub-wash technique at five sites (forehead, cheek, chest, upper arm, and thigh) and swabbing technique at scalp in 160 participants comprised of 80 males and 80 females aged from 0 to 80. Identification of obtained strains were placed into the one of the 11 species by 26S rDNA PCR-RFLP. RESULTS: An overall positive culture rate was 62.4% (599/960). As shown in the experiment groups by their age, the positive culture rate was the highest (74.2%) in the age 21~30 and 31~40 (89/120). In the experiment groups by different body areas, the scalp showed the highest positive culture rate of 90% (144/160). On analysis of 26S rDNA PCR-RFLP, M. globosa was the most predominant species in the age 0~10 (32.8%), 11~20 (28.9%), 21~30 (32.3%). M. restricta was identified as predominant species in the age 41~50 (27.9%), 61~70 (31.5%) and 71~80 (24.0%). In the age 31~40 years, M. sympodialis was found to be the most common species (24.6%). According to body site, M. restricta was more frequently recovered in the scalp (56.8%), forehead (39.8%) and cheek (24.0%) and while M. globosa was more frequently recovered in the chest (36.8%). Higher positive culture rates of Malassezia yeasts were shown in male subjects than female counterparts in all body areas except scalp (p<0.05). Especially in this study, M. dermatis, newly isolated Malassezia species from atopic dermatitis patient in Japan, was isolated and identified in 19 cases (1.9%) in healthy subjects. CONCLUSION: The key is to recognize the existence of a difference in the type of Malassezia species in different ages as well as body areas, which reflects differing skin lipid levels in various ages and different body areas. Moreover, 26S rDNA PCR-RFLP analysis which was opted in this study could provide a sensitive and rapid identification system for Malassezia species, which may be applied to epidemiological surveys and clinical practice.