Background: Recent studies have successfully demonstrated the use of deep-learning algorithms for dermatologist-level classification of suspicious lesions by the use of excessive proprietary image databases and limited numbers of dermatologists. For the first time, the performance of a deep-learning algorithm trained by open-source images exclusively is compared to a large number of dermatologists covering all levels within the clinical hierarchy. Methods: We used methods from enhanced deep learning to train a convolutional neural network (CNN) with 12,378 open-source dermoscopic images. We used 100 images to compare the performance of the CNN to that of the 157 dermatologists from 12 university hospitals in Germany.

Optical coherence tomography (OCT) is a noninvasive diagnostic method that offers a view into the superficial layers of the skin in vivo in real-time. An infrared broadband light source allows the investigation of skin architecture and changes up to a depth of 1 to 2 mm with a resolution between 15 and 3 μm, depending on the system used. Thus OCT enables evaluation of skin lesions, especially nonmelanoma skin cancers and inflammatory diseases, quantification of skin changes, visualization of parasitic infestations, and examination of other indications such as the investigation of nails. OCT provides a quick and useful diagnostic imaging technique for a number of clinical questions and is a valuable addition or complement to other noninvasive imaging tools such as dermoscopy, high-frequency ultrasound, and confocal laser scan microscopy.

No abstract

has received speaker's honoraria from Almirall Hermal, Biofrontera, Galderma, Meda Pharma and JanssenCilag. C.B. has received speaker's and advisory board member's honoraria from, and has been involved in clinical trials sponsored by Almirall SummaryBackground The diagnostic criteria for basal cell carcinoma (BCC) using optical coherence tomography (OCT) have been described previously, but the clinical value of these findings remains unknown. Objectives To investigate the diagnostic value of OCT for BCC in a typical clinical setting. The primary efficacy end point was a diagnosis of BCC for each lesion. Secondary end points were the diagnosis of other possible conditions. Methods This was an observational, prospective, multicentre study in which consecutive patients with nonpigmented pink lesions suspicious for BCC underwent clinical assessment, dermoscopy and OCT, with the diagnosis recorded at each stage. Once all diagnoses had been recorded, the histological results were disclosed. In total 164 patients with 256 lesions were recruited. Histology was missing for 21 lesions, leaving 235 lesions in 155 patients for analysis. Results Sixty per cent of lesions (141 of 235) were identified as BCC by histology. A slight increase of sensitivity was noted following OCT, which did not reach statistical significance. The specificity increased significantly from 28Á6% by clinical assessment to 54Á3% using dermoscopy and to 75Á3% with the addition of OCT (P < 0Á001). The positive predictive value for the diagnosis of BCC using OCT was 85Á2% [95% confidence interval (CI) 78Á6-90Á4], and the negative predictive value was 92Á1% (95% CI 83Á6-97Á0). The accuracy of diagnosis for all lesions increased from 65Á8% with clinical evaluation to 76Á2% following additional dermoscopy and to 87Á4% with the addition of OCT. Conclusions OCT significantly improved the diagnostic specificity for BCC compared with clinical assessment and dermoscopy alone.

Onychomycosis is common and can mimic several different nail disorders. Accurate diagnosis is essential to choose the optimum antifungal therapy. The aim of this study was to evaluate the use of confocal laser scanning microscopy (CLSM) and optical coherence tomography (OCT) as new non-invasive diagnostic tools in onychomycosis and to compare them with the established techniques. In a prospective trial, 50 patients with suspected onychomycosis and 10 controls were examined by CLSM and OCT. Parallel KOH preparation, culture, PAS-staining and PCR were performed. PCR showed the highest sensitivity, followed by CLSM, PAS and KOH preparation. OCT offered the second best sensitivity but displayed the lowest specificity. CLSM and KOH preparation showed a high specificity and CLSM offered the best positive predictive value, similar to KOH preparation and OCT. Fungal culture showed the lowest sensitivity and the worst negative predictive value, yet culture and PCR are the only techniques able to identify genus and species. In summary, CLSM was comparable to PAS staining and superior to KOH preparation. Due to the low specificity we assess OCT not as appropriate. In the differentiation of species PCR outplays the fungal culture in terms of time and sensitivity.