Man against machine reloaded: performance of a market-approved convolutional neural network in classifying a broad spectrum of skin lesions in comparison with 96 dermatologists working under less artificial conditions.

BACKGROUND: Convolutional neural networks (CNNs) efficiently differentiate skin lesions by image analysis. Studies comparing a market-approved CNN in a broad range of diagnoses to dermatologists working under less artificial conditions are lacking. MATERIALS AND METHODS: One hundred cases of pigmented/non-pigmented skin cancers and benign lesions were used for a two-level reader study in 96 dermatologists (level I: dermoscopy only; level II: clinical close-up images, dermoscopy, and textual information). Additionally, dermoscopic images were classified by a CNN approved for the European market as a medical device (Moleanalyzer Pro, FotoFinder Systems, Bad Birnbach, Germany). Primary endpoints were the sensitivity and specificity of the CNN's dichotomous classification in comparison with the dermatologists' management decisions. Secondary endpoints included the dermatologists' diagnostic decisions, their performance according to their level of experience, and the CNN's area under the curve (AUC) of receiver operating characteristics (ROC). RESULTS: The CNN revealed a sensitivity, specificity, and ROC AUC with corresponding 95% confidence intervals (CI) of 95.0% (95% CI 83.5% to 98.6%), 76.7% (95% CI 64.6% to 85.6%), and 0.918 (95% CI 0.866-0.970), respectively. In level I, the dermatologists' management decisions showed a mean sensitivity and specificity of 89.0% (95% CI 87.4% to 90.6%) and 80.7% (95% CI 78.8% to 82.6%). With level II information, the sensitivity significantly improved to 94.1% (95% CI 93.1% to 95.1%; P < 0.001), while the specificity remained unchanged at 80.4% (95% CI 78.4% to 82.4%; P = 0.97). When fixing the CNN's specificity at the mean specificity of the dermatologists' management decision in level II (80.4%), the CNN's sensitivity was almost equal to that of human raters, at 95% (95% CI 83.5% to 98.6%) versus 94.1% (95% CI 93.1% to 95.1%); P = 0.1. In contrast, dermatologists were outperformed by the CNN in their level I management decisions and level I and II diagnostic decisions. More experienced dermatologists frequently surpassed the CNN's performance. CONCLUSIONS: Under less artificial conditions and in a broader spectrum of diagnoses, the CNN and most dermatologists performed on the same level. Dermatologists are trained to integrate information from a range of sources rendering comparative studies that are solely based on one single case image inadequate.

Standardization of dermoscopic terminology and basic dermoscopic parameters to evaluate in general dermatology (non-neoplastic dermatoses): an expert consensus on behalf of the International Dermoscopy Society.

BACKGROUND: Over the last few years, several articles on dermoscopy of non-neoplastic dermatoses have been published, yet there is poor consistency in the terminology among different studies. OBJECTIVES: We aimed to standardize the dermoscopic terminology and identify basic parameters to evaluate in non-neoplastic dermatoses through an expert consensus. METHODS: The modified Delphi method was followed, with two phases: (i) identification of a list of possible items based on a systematic literature review and (ii) selection of parameters by a panel of experts through a three-step iterative procedure (blinded e-mail interaction in rounds 1 and 3 and a face-to-face meeting in round 2). Initial panellists were recruited via e-mail from all over the world based on their expertise on dermoscopy of non-neoplastic dermatoses. RESULTS: Twenty-four international experts took part in all rounds of the consensus and 13 further international participants were also involved in round 2. Five standardized basic parameters were identified: (i) vessels (including morphology and distribution); (ii) scales (including colour and distribution); (iii) follicular findings; (iv) 'other structures' (including colour and morphology); and (v) 'specific clues'. For each of them, possible variables were selected, with a total of 31 different subitems reaching agreement at the end of the consensus (all of the 29 proposed initially plus two more added in the course of the consensus procedure). CONCLUSIONS: This expert consensus provides a set of standardized basic dermoscopic parameters to follow when evaluating inflammatory, infiltrative and infectious dermatoses. This tool, if adopted by clinicians and researchers in this field, is likely to enhance the reproducibility and comparability of existing and future research findings and uniformly expand the universal knowledge on dermoscopy in general dermatology. What's already known about this topic? Over the last few years, several papers have been published attempting to describe the dermoscopic features of non-neoplastic dermatoses, yet there is poor consistency in the terminology among different studies. What does this study add? The present expert consensus provides a set of standardized basic dermoscopic parameters to follow when evaluating inflammatory, infiltrative and infectious dermatoses. This consensus should enhance the reproducibility and comparability of existing and future research findings and uniformly expand the universal knowledge on dermoscopy in general dermatology.

Diagnostic performance of a deep learning convolutional neural network in the differentiation of combined naevi and melanomas.

BACKGROUND: Deep learning convolutional neural networks (CNN) may assist physicians in the diagnosis of melanoma. The capacity of a CNN to differentiate melanomas from combined naevi, the latter representing well-known melanoma simulators, has not been investigated. OBJECTIVE: To assess the diagnostic performance of a CNN when used to differentiate melanomas from combined naevi in comparison with dermatologists. METHODS: In this study, a CNN with regulatory approval for the European market (Moleanalyzer-Pro, FotoFinder Systems GmbH, Bad Birnbach, Germany) was used. We attained a dichotomous classification (benign, malignant) in dermoscopic images of 36 combined naevi and 36 melanomas with a mean Breslow thickness of 1.3 mm. Primary outcome measures were the CNN's sensitivity, specificity and the diagnostic odds ratio (DOR) in comparison with 11 dermatologists with different levels of experience. RESULTS: The CNN revealed a sensitivity, specificity and DOR of 97.1% (95% CI [82.7-99.6]), 78.8% (95% CI [62.8-89.1.3]) and 34 (95% CI [4.8-239]), respectively. Dermatologists showed a lower mean sensitivity, specificity and DOR of 90.6% (95% CI [84.1-94.7]; P = 0.092), 71.0% (95% CI [62.6-78.1]; P = 0.256) and 24 (95% CI [11.6-48.4]; P = 0.1114). Under the assumption that dermatologists use the CNN to verify their (initial) melanoma diagnosis, dermatologists achieve an increased specificity of 90.3% (95% CI [79.8-95.6]) at an almost unchanged sensitivity. The largest benefit was observed in 'beginners', who performed worst without CNN verification (DOR = 12) but best with CNN verification (DOR = 98). CONCLUSION: The tested CNN more accurately classified combined naevi and melanomas in comparison with trained dermatologists. Their diagnostic performance could be improved if the CNN was used to confirm/overrule an initial melanoma diagnosis. Application of a CNN may therefore be of benefit to clinicians.

Dermoscopic features of mammary Paget's disease: a retrospective case-control study by the International Dermoscopy Society.

BACKGROUND: Mammary Paget's disease (MPD) is a rare intraepidermal adenocarcinoma of the nipple-areola complex, associated with an underlying breast cancer in approximately 90% of cases. Delayed diagnosis of MPD is common. Its dermoscopic features have been ill defined in the literature. OBJECTIVES: To determine the clinical and dermoscopic features of MPD versus other dermatologic entities that involve nipple and areola. METHODS: Members of the IDS were invited to submit any case of histologically confirmed MPD, as well as other benign and malignant dermatoses that involve the nipple and areola complex. A standardized evaluation of the dermoscopic images was performed and the results were statistically analyzed. RESULTS: Sixty-five lesions were included in the study, 22 (33.8%) of them MPD and 43 (66.2%) controls. The most frequent dermoscopic criteria of MPD were white scales (86.4%) and pink structureless areas (81.8%), followed by dotted vessels (72.7%), erosion/ulceration (68.2%) and white shiny lines (63.6%). The multivariate analysis showed that white scales and pink structureless areas were significant predictors of MPD, posing a 68-fold and a 31-fold probability of MPD, respectively. Split of the population into pigmented and non-pigmented lesions showed that in pigmented MPD, pink structureless areas, white lines and grey granules and dots are positive predictors of the disease. Among non-pigmented lesions, pink structureless areas, white lines, erosion/ulceration and white scales served as predictors of MPD. CONCLUSIONS: The most frequent profile of an individual with MPD is an elderly female with unilateral, asymptomatic, erythematous plaque of the nipple, dermoscopically displaying pink structureless areas, fine white scales, dotted and a few short linear vessels. In case of pigmentation we may also observe brown structureless areas and pigmented granules. LIMITATIONS: Small sample size, retrospective design.

Composition determination of semiconductor alloys towards atomic accuracy by HAADF-STEM.

This paper presents a comprehensive investigation of an extended method to determine composition of materials by scanning transmission electron microscopy (STEM) high angle annular darkfield (HAADF) images and using complementary multislice simulations. The main point is to understand the theoretical capabilities of the algorithm and address the intrinsic limitations of using STEM HAADF intensities for composition determination. A special focus is the potential of the method regarding single-atom accuracy. All-important experimental parameters are included into the multislice simulations to ensure the best possible fit between simulation and experiment. To demonstrate the capabilities of the extended method, results for three different technical important semiconductor samples are presented. Overall the method shows a high lateral resolution combined with a high accuracy towards single-atom accuracy.

[Dermoscopy in special locations : Nails, acral skin, face, and mucosa]. / Dermatoskopie in Sonderlokalisationen : Nägel, akrale Haut, Gesicht und Mukosa.

The use of dermoscopy by dermatologists across Europe has become a standard examination for benign and malignant skin lesions and increasingly also for inflammatory skin diseases. However, based on the experience of the authors from numerous dermoscopy courses, knowledge about important dermoscopic features in special locations such as mucosa or nails is often limited. This may be explained by (1) a different anatomy of the skin and its adnexa in special locations in comparison to the remaining integument, (2) difficult technical access to special locations with a dermatoscope, and (3) a rather low incidence of malignant skin neoplasms in areas of special locations (with the exception of facial skin/scalp). This article aims at explaining dermoscopic characteristics and features of important benign and malignant lesions of nails, acral skin, face, and mucosa.

Second-harmonic generation as a probe for structural and electronic properties of buried GaP/Si(0 0 1) interfaces.

Optical second-harmonic generation is demonstrated to be a sensitive probe of the buried interface between the lattice-matched semiconductors gallium phosphide and silicon with (0 0 1) orientation. Ex situ rotational anisotropy measurements on GaP/Si heterostructures show a strong isotropic component of the second-harmonic response not present for pure Si(0 0 1) or GaP(0 0 1). The strength of the overlaying anisotropic response directly correlates with the quality of the interface as determined by atomically resolved scanning transmission electron microscopy. Systematic comparison of samples fabricated under different growth conditions in metal-organic vapor phase epitaxy reveals that the anisotropy for different polarization combinations can be used as a selective fingerprint for the occurrence of anti-phase domains and twins. This all-optical technique can be applied as an in situ and non-invasive monitor even during growth.

Enhanced Absorption by Linewidth Narrowing in Optically Excited Type-II Semiconductor Heterostructures.

We experimentally report a surprising linewidth narrowing of the direct exitonic 1 s heavy-hole transition in a type-II quantum well system. This narrowing, which builds up on a pico- to nanosecond timescale, causes a transient enhanced absorption at the spectral peak position of the excitonic resonance. We discuss how this effect depends on experimental parameters such as excitation density, temperature, and barrier width. We cannot attribute this effect to known physical mechanisms.

Author Correction: High-temperature operation of electrical injection type-II (GaIn)As/Ga(AsSb)/(GaIn)As "W"-quantum well lasers emitting at 1.3 µm.

A correction to this article has been published and is linked from the HTML and PDF versions of this paper. The error has not been fixed in the paper.

An experimental approach for real time mass spectrometric CVD gas phase investigations.

This is a report on the first setup of a recently developed, extremely sensitive and very fast 3D quadrupole ion trap mass spectrometer inline in a metalorganic vapour phase epitaxy (MOVPE) system. This setup was developed ultimately for the decomposition- and the interaction analysis of various established as well as novel metalorganic sources for MOVPE deposition of III/V semiconductors. To make in-situ gas phase and growth interaction analysis on a new level of sensitivity possible without disturbing the MOVPE growth process itself, an optimized experimental connection of the mass spectrometer to the MOVPE system is required. This work reports on the realization of such an experimental setup and provides first proof of concept for decomposition analysis. In addition, a comparison to previous studies and gas-phase analysis at MOVPE systems will be given in this work.

High-temperature operation of electrical injection type-II (GaIn)As/Ga(AsSb)/(GaIn)As "W"-quantum well lasers emitting at 1.3 µm.

Electrical injection lasers emitting in the 1.3 µm wavelength regime based on (GaIn)As/Ga(AsSb)/(GaIn)As type-II double "W"-quantum well heterostructures grown on GaAs substrate are demonstrated. The structure is designed by applying a fully microscopic theory and fabricated using metal organic vapor phase epitaxy. Temperature-dependent electroluminescence measurements as well as broad-area edge-emitting laser studies are carried out in order to characterize the resulting devices. Laser emission based on the fundamental type-II transition is demonstrated for a 975 µm long laser bar in the temperature range between 10 °C and 100 °C. The device exhibits a differential efficiency of 41 % and a threshold current density of 1.0 kA/cm2 at room temperature. Temperature-dependent laser studies reveal characteristic temperatures of T0 = (132 ± 3) K over the whole temperature range and T1 = (159 ± 13) K between 10 °C and 70 °C and T1 = (40 ± 1) K between 80 °C and 100 °C.

Atomic structure of 'W'-type quantum well heterostructures investigated by aberration-corrected STEM.

The atomic structure of (GaIn)As/Ga(AsSb)/(GaIn)As-'W'-type quantum well heterostructures ('W'-QWHs) is investigated by scanning transmission electron microscopy (STEM). These structures were grown by metal organic vapour phase epitaxy and are built for type-II laser systems in the infrared wavelength regime. For two samples grown at 525°C and 550°C, intensity profiles are extracted from the STEM images for each sublattice separately. These intensity profiles are compared to the one obtained from an image simulation of an ideal 'W'-QWH that is modelled in close agreement with the experiment. From the intensity profiles, the width of the different quantum wells (QWs) can be determined. Additionally, characteristics connected to the growth of the structures, such as segregation coefficients and material homogeneity, are calculated. Finally, composition profiles are derived from the STEM intensity profiles to a first approximation. For these composition profiles, the expected photoluminescence (PL) is computed based using the semiconductor luminescence equations. The PL spectra are then compared to experimental measurements for both samples.

[Dermoscopy for malignant and benign skin tumors : Indication and standardized terminology]. / Dermatoskopie bei malignen und benignen Hauttumoren : Indikation und standardisierte Terminologie.

Dermoscopy has a high diagnostic accuracy in pigmented and nonpigmented malignant and benign skin tumors. These microscopic in vivo examinations with polarized and nonpolarized light are effective in the early detection of malignant skin tumors and reduce the number of unnecessary excisions of benign skin tumors. The selection of the skin lesions is crucial for the diagnostic accuracy of the dermoscopic examination. Not only large pigmented skin lesions, but also small hypo-, de-, or nonpigmented skin lesions, should be examined dermatoscopically as well as skin lesions that have changed in shape and/or color. In clinical routine, research and teaching, the dermoscopic diagnosis should be performed by describing the visible structures, their distribution and colors by means of descriptive and/or metaphoric terminology. Optionally, a diagnostic algorithm can also be used. Especially in benign lesions, the dermatoscopic diagnosis should be uniform for the complete area. Comparison with other nearby skin tumors of the same patient (comparative approach) is helpful in the evaluation of numerous melanocytic skin tumors. If it is unclear whether the lesion is malignant, a biopsy or complete excision should be performed with subsequent histopathological examination.

Recombination dynamics of type-II excitons in (Ga,In)As/GaAs/Ga(As,Sb) heterostructures.

(Ga,In)As/GaAs/Ga(As,Sb) multi-quantum well heterostructures have been investigated using continuous wave and time-resolved photoluminescence spectroscopy at various temperatures. A complex interplay was observed between the excitonic type-II transitions with electrons in the (Ga,In)As well and holes in the Ga(As,Sb) well and the type-I excitons in the (Ga,In)As and Ga(As,Sb) wells. The type-II luminescence exhibits a strongly non-exponential temporal behavior below a critical temperature of T c = 70 K. The transients were analyzed in the framework of a rate-equation model. It was found that the exciton relaxation and hopping in the localized states of the disordered ternary Ga(As,Sb) are the decisive processes to describe the dynamics of the type-II excitons correctly.

Clinical and dermoscopic clues to differentiate pigmented nail bands: an International Dermoscopy Society study.

BACKGROUND: Longitudinal melanonychia might be difficult to differentiate and the use of dermoscopy can be useful for the preoperative evaluation and management decision. OBJECTIVES: The aim of our study was to investigate clinical and dermoscopic criteria of acquired longitudinal melanonychia in adults to identify the best predictors of melanoma using a multivariate analysis and to explore eventual new dermoscopic criteria for nail melanoma diagnosis. METHODS: In this retrospective observational study, 82 histopathologically diagnosed, acquired nail pigmented bands were collected and examined. All variables were included in the analysis and examined as possible predictors of nail melanoma. Both univariate and multivariable analyses have been performed. RESULTS: Among 82 cases, 25 were diagnosed as nail melanoma and 57 as benign lesions (including 32 melanocytic nevi and 25 benign melanocytic hyperplasia). Melanoma cases were significantly associated with a width of the pigmented band higher than 2/3 of the nail plate, grey and black colours, irregularly pigmented lines, Hutchinson and micro-Hutchinson signs, and nail dystrophy. Granular pigmentation, a newly defined dermoscopic criterion, was found in 40% of melanomas and only in 3.51% of benign lesions. CONCLUSIONS: Dermoscopic examination of longitudinal melanonychia provides useful information that could help clinicians to improve melanoma recognition.