Clinician's Ability to Identify Non-Melanoma Skin Cancer on 3D-Total Body Photography Sectors that Were Initially Identified during In-Person Skin Examination with Dermoscopy.

INTRODUCTION: Non-melanoma skin cancer (NMSC) is a cause of significant morbidity and mortality in high-risk individuals. Total body photography (TBP) is currently used to monitor melanocytic lesions in patients with high risk for melanoma. The authors examined if three-dimensional (3D)-TBP could be useful for diagnosis of NMSC. METHODS: Patients (n = 129; 52 female, 77 male) with lesions suspicious for NMSC who had not yet had a biopsy underwent clinical examination followed by examination of each lesion with 3D-TBP Vectra®WB360 (Canfield Scientific, Parsippany, NJ, USA) and dermoscopy. RESULTS: The 129 patients had a total of 182 lesions. Histological examination was performed for 158 lesions; the diagnoses included basal cell carcinoma (BCC; n = 107), squamous cell carcinoma (SCC; n = 27), in-situ SCC (n = 15). Lesions were located in the head/neck region (n = 138), trunk (n = 21), and limbs (n = 23). Of the 182 lesions examined, 12 were not visible on 3D-TBP; reasons for not being visible included location under hair and on septal of nose. Two lesions appeared only as erythema in 3D-TBP but were clearly identifiable on conventional photographs. Sensitivity of 3D-TBP was lower than that of dermoscopy for BCC (73% vs. 79%, p = 0.327), higher for SCC (81% vs. 74%, p = 0.727), and lower for in-situ SCC (0% vs. 33%, p = 125). Specificity of 3D-TBP was lower than that of dermoscopy for BCC (77% vs. 82%, 0.581), lower for SCC (75% vs. 84%, p = 0.063), and higher for in-situ SCC (97% vs. 94%, p = 0.344). Diagnostic accuracy of 3D-TBP was lower than that of dermoscopy for BCC (75% vs. 80%), lower for SCC (76% vs. 82%), and lower for in-situ SCC (88% vs. 89%). Lesion location was not associated with diagnostic confidence in dermoscopy (p = 0.152) or 3D-TBP (p = 0.353). If only lesions with high confidence were included in the calculation, diagnostic accuracy increased for BCC (n = 27; sensitivity 85%, specificity 85%, diagnostic accuracy 85%), SCC (n = 10; sensitivity 90%, specificity 80%, diagnostic accuracy 83%), and for in-situ SCC (n = 2; sensitivity 0%, specificity 100%, diagnostic accuracy 95%). CONCLUSION: Diagnostic accuracy appears to be slightly lower for 3D-TBP in comparison to dermoscopy. However, there is no statistically significant difference in the sensitivity and specificity of 3D-TBP and dermoscopy for NMSC. Diagnostic accuracy increases, if only lesions with high confidence are included in the calculation. Further studies are necessary to determine if 3D-TBP can improve management of NMSC.

Examination of Subungual Hematomas and Subungual Melanocytic Lesions by Using Optical Coherence Tomography and Dermoscopy.

INTRODUCTION: Examination of subungual pigmented lesions is sometimes a diagnostic challenge for clinicians. OBJECTIVES: The study was aimed to investigate characteristic patterns in optical coherence tomography (OCT) of subungual hematomas and determine distinctive features that can differentiate them from subungual melanocytic lesions. METHODS: VivoSight® (Michelson Diagnostics, Maidstone, UK) was used to examine 71 subungual hematomas and 11 subungual melanocytic lesions in 69 patients (18 female and 51 male patients). RESULTS: On OCT, bleeding was related to sharply defined black sickle-shaped (p < 0.001) or globular regions (not significant [ns]) with a hyperreflective margin (0.002), a grey center (0.013), hyperreflective lines in the area (ns) or periphery (p = 0.031), peripheral fading (p = 0.029), and red dots in the area (p = 0.001). In the 1 case of melanoma in situ examined, we found curved vessels with irregular sizes and distribution on the dermis of the nailbed, while subungual hematomas and subungual benign nevi presented as clustered red dots and/or regularly distributed curved vessels. CONCLUSION: Our findings indicate that the use of OCT in addition to dermoscopy provides high-resolution optical imaging information for the diagnosis of subungual hematoma and facilitates the differential diagnosis of subungual hematomas and subungual melanocytic lesions.

[Successful treatment of cutaneous leishmaniasis with simulated daylight photodynamic therapy]. / Erfolgreiche Therapie der kutanen Leishmaniose mit simulierter Tageslicht-photodynamischer Therapie.

A 5-year-old Syrian boy , presented with a complex cutaneous leishmaniasis (CL) of the right ankle caused by Leishmania (L.) tropica. The patient received photodynamic therapy (PDT; 6 cycles with application of 5­aminolevulinic acid and foil occlusion for 3 h). Due to pain during exposure to red light, exposure was continued with simulated daylight (sDL-PDT). The lesion healed with an atrophic scar. Due to fewer side effects and less pain, sDL-PDT seems to be a good therapeutic strategy for CL caused by L. tropica.

[Optical coherence tomography for the diagnosis and differentiation of cutaneous cysts: a case series]. / Diagnose und Differenzierung von kutanen Zysten mit optischer Kohärenztomographie: eine Fallserie.

Cutaneous cystic lesions (n = 35) were examined with optical coherence tomography. Cysts were visible as a hyporeflective roundish area with a clear margin; in some cases, the epidermis was thinned. Epidermal cysts, trichilemmal cysts, and hidrocystomas had a linear margin representing the epithelium of the cyst, whereas mucoid pseudocysts showed no linear margin. Trichilemmal and epidermal cysts presented with hyperreflective content that corresponds to keratin. By visualizing the margin and the content of the cyst, it was possible to differentiate between different types of cysts.

Comparison of Extended Skin Cancer Screening Using a Three-Step Advanced Imaging Programme vs. Standard-of-Care Examination in a High-Risk Melanoma Patient Cohort.

Modern diagnostic procedures, such as three-dimensional total body photography (3D-TBP), digital dermoscopy (DD), and reflectance confocal microscopy (RCM), can improve melanoma diagnosis, particularly in high-risk patients. This study assessed the benefits of combining these advanced imaging techniques in a three-step programme in managing high-risk patients. This study included 410 high-risk melanoma patients who underwent a specialised imaging consultation in addition to their regular skin examinations in outpatient care. At each visit, the patients underwent a 3D-TBP, a DD for suspicious findings, and an RCM for unclear DD findings. The histological findings of excisions initiated based on imaging consultation and outpatient care were compared. Imaging consultation detected sixteen confirmed melanomas (eight invasive and eight in situ) in 39 excised pigmented lesions. Outpatient care examination detected seven confirmed melanomas (one invasive and six in situ) in 163 excised melanocytic lesions. The number needed to excise (NNE) in the imaging consultation was significantly lower than that in the outpatient care (2.4 vs. 23.3). The NNE was 2.6 for DD and 2.3 for RCM. DD, 3D-TBP, or RCM detected melanomas that were not detected by the other imaging methods. The three-step imaging programme improves melanoma detection and reduces the number of unnecessary excisions in high-risk patients.

Evaluating deep learning-based melanoma classification using immunohistochemistry and routine histology: A three center study.

Pathologists routinely use immunohistochemical (IHC)-stained tissue slides against MelanA in addition to hematoxylin and eosin (H&E)-stained slides to improve their accuracy in diagnosing melanomas. The use of diagnostic Deep Learning (DL)-based support systems for automated examination of tissue morphology and cellular composition has been well studied in standard H&E-stained tissue slides. In contrast, there are few studies that analyze IHC slides using DL. Therefore, we investigated the separate and joint performance of ResNets trained on MelanA and corresponding H&E-stained slides. The MelanA classifier achieved an area under receiver operating characteristics curve (AUROC) of 0.82 and 0.74 on out of distribution (OOD)-datasets, similar to the H&E-based benchmark classification of 0.81 and 0.75, respectively. A combined classifier using MelanA and H&E achieved AUROCs of 0.85 and 0.81 on the OOD datasets. DL MelanA-based assistance systems show the same performance as the benchmark H&E classification and may be improved by multi stain classification to assist pathologists in their clinical routine.

Dermatologist-like explainable AI enhances trust and confidence in diagnosing melanoma.

Artificial intelligence (AI) systems have been shown to help dermatologists diagnose melanoma more accurately, however they lack transparency, hindering user acceptance. Explainable AI (XAI) methods can help to increase transparency, yet often lack precise, domain-specific explanations. Moreover, the impact of XAI methods on dermatologists' decisions has not yet been evaluated. Building upon previous research, we introduce an XAI system that provides precise and domain-specific explanations alongside its differential diagnoses of melanomas and nevi. Through a three-phase study, we assess its impact on dermatologists' diagnostic accuracy, diagnostic confidence, and trust in the XAI-support. Our results show strong alignment between XAI and dermatologist explanations. We also show that dermatologists' confidence in their diagnoses, and their trust in the support system significantly increase with XAI compared to conventional AI. This study highlights dermatologists' willingness to adopt such XAI systems, promoting future use in the clinic.

Federated Learning for Decentralized Artificial Intelligence in Melanoma Diagnostics.

Importance: The development of artificial intelligence (AI)-based melanoma classifiers typically calls for large, centralized datasets, requiring hospitals to give away their patient data, which raises serious privacy concerns. To address this concern, decentralized federated learning has been proposed, where classifier development is distributed across hospitals. Objective: To investigate whether a more privacy-preserving federated learning approach can achieve comparable diagnostic performance to a classical centralized (ie, single-model) and ensemble learning approach for AI-based melanoma diagnostics. Design, Setting, and Participants: This multicentric, single-arm diagnostic study developed a federated model for melanoma-nevus classification using histopathological whole-slide images prospectively acquired at 6 German university hospitals between April 2021 and February 2023 and benchmarked it using both a holdout and an external test dataset. Data analysis was performed from February to April 2023. Exposures: All whole-slide images were retrospectively analyzed by an AI-based classifier without influencing routine clinical care. Main Outcomes and Measures: The area under the receiver operating characteristic curve (AUROC) served as the primary end point for evaluating the diagnostic performance. Secondary end points included balanced accuracy, sensitivity, and specificity. Results: The study included 1025 whole-slide images of clinically melanoma-suspicious skin lesions from 923 patients, consisting of 388 histopathologically confirmed invasive melanomas and 637 nevi. The median (range) age at diagnosis was 58 (18-95) years for the training set, 57 (18-93) years for the holdout test dataset, and 61 (18-95) years for the external test dataset; the median (range) Breslow thickness was 0.70 (0.10-34.00) mm, 0.70 (0.20-14.40) mm, and 0.80 (0.30-20.00) mm, respectively. The federated approach (0.8579; 95% CI, 0.7693-0.9299) performed significantly worse than the classical centralized approach (0.9024; 95% CI, 0.8379-0.9565) in terms of AUROC on a holdout test dataset (pairwise Wilcoxon signed-rank, P < .001) but performed significantly better (0.9126; 95% CI, 0.8810-0.9412) than the classical centralized approach (0.9045; 95% CI, 0.8701-0.9331) on an external test dataset (pairwise Wilcoxon signed-rank, P < .001). Notably, the federated approach performed significantly worse than the ensemble approach on both the holdout (0.8867; 95% CI, 0.8103-0.9481) and external test dataset (0.9227; 95% CI, 0.8941-0.9479). Conclusions and Relevance: The findings of this diagnostic study suggest that federated learning is a viable approach for the binary classification of invasive melanomas and nevi on a clinically representative distributed dataset. Federated learning can improve privacy protection in AI-based melanoma diagnostics while simultaneously promoting collaboration across institutions and countries. Moreover, it may have the potential to be extended to other image classification tasks in digital cancer histopathology and beyond.

Obstructive sleep apnoea is associated with the development of diastolic dysfunction after myocardial infarction with preserved ejection fraction.

BACKGROUND: Left ventricular diastolic dysfunction is a predictor of adverse outcome after acute myocardial infarction (AMI). We aimed to test if sleep-disordered breathing (SDB) contributes to the development of diastolic dysfunction in patients with preserved left ventricular ejection fraction after AMI. METHOD: Patients with AMI, percutaneous coronary intervention and an ejection fraction ≥50% were included in this sub-analysis of a prospective observational study. Patients with AMI (n = 41) underwent cardiovascular magnetic resonance imaging (volume-time curve analysis) to define diastolic function by means of the normalised peak filling rate [nPFR; (end diastolic volume/second)]. In patients with AMI, the nPFR was assessed within <5 days and three months after AMI. Patients with AMI were stratified in patients with (apnoea-hypopnoea index, AHI ≥15/h) and without (AHI <15/h) SDB as assessed by polysomnography. RESULTS: At the time of AMI, the nPFR was similar between patients with and without SDB (2.90 ± 0.54 vs. 3.03 ± 1.20, p = 0.662). Within three months after AMI, diastolic function was significantly lower in patients with SDB than in patients without SDB (ΔnPFR: -0.83 ± 0.14 vs. 0.03 ± 0.14; p < 0.001; ANCOVA, adjusted for baseline nPFR). In contrast to central AHI, obstructive AHI was associated with a lower nPFR three months after AMI, after accounting for established risk factors for diastolic dysfunction [multiple linear regression analysis, B (95%CI): -0.036 (-0.063 to -0.009), p = 0.011]. CONCLUSION: Our data indicate that obstructive sleep apnoea impairs diastolic function early after myocardial infarction.

Metabolic Syndrome in Psoriasis Is Associated With Upregulation of CXCL16 on Monocytes and a Dysbalance in Innate Lymphoid Cells.

Psoriasis is frequently associated with the metabolic syndrome and occurs more often in obese individuals. In order to understand innate immune mechanisms mediating this inflammatory pattern we investigated expression of the chemokine and lipid scavenger receptor CXCL16 in patients with psoriasis and associated comorbidities. CXCL16 expression was enhanced on all monocyte subsets in psoriatic patients compared with healthy controls and positively correlated with psoriasis activity and severity index, body mass index and the risk for cardiovascular disease indicated by PROCAM score. The intensity of CXCL16 expression on monocytes further correlated with their capability to phagocytose oxidized LDL indicating the possibility to transform into foam cells in atherosclerotic plaques. Patients with psoriasis and atherosclerosis or obesity displayed elevated numbers of innate lymphoid cells in blood with specific increase of the IFN-γ or IL-17 producing ILC1 and ILC3 subpopulations. The expression of the CXCL16 receptor, CXCR6, was increased in ILCs and co-expressed with CCR6 but not CCR7 indicating their migratory potential to psoriatic skin or adipose tissue that is characterized by strong CXCL16 and CCL20 expression. This hypothesis was supported by the finding that the percentage of CXCR6 expressing ILCs was alleviated in blood of psoriatic patients. Together these data link a strong expression of CXCL16 to metabolic syndrome in psoriasis and indicate a possible link to ILC activation and tissue distribution in obese psoriatic patients. These data contribute to the understanding of the complex interaction of innate immunity and metabolic state in psoriasis.

Explainable artificial intelligence in skin cancer recognition: A systematic review.

BACKGROUND: Due to their ability to solve complex problems, deep neural networks (DNNs) are becoming increasingly popular in medical applications. However, decision-making by such algorithms is essentially a black-box process that renders it difficult for physicians to judge whether the decisions are reliable. The use of explainable artificial intelligence (XAI) is often suggested as a solution to this problem. We investigate how XAI is used for skin cancer detection: how is it used during the development of new DNNs? What kinds of visualisations are commonly used? Are there systematic evaluations of XAI with dermatologists or dermatopathologists? METHODS: Google Scholar, PubMed, IEEE Explore, Science Direct and Scopus were searched for peer-reviewed studies published between January 2017 and October 2021 applying XAI to dermatological images: the search terms histopathological image, whole-slide image, clinical image, dermoscopic image, skin, dermatology, explainable, interpretable and XAI were used in various combinations. Only studies concerned with skin cancer were included. RESULTS: 37 publications fulfilled our inclusion criteria. Most studies (19/37) simply applied existing XAI methods to their classifier to interpret its decision-making. Some studies (4/37) proposed new XAI methods or improved upon existing techniques. 14/37 studies addressed specific questions such as bias detection and impact of XAI on man-machine-interactions. However, only three of them evaluated the performance and confidence of humans using CAD systems with XAI. CONCLUSION: XAI is commonly applied during the development of DNNs for skin cancer detection. However, a systematic and rigorous evaluation of its usefulness in this scenario is lacking.

Model soups improve performance of dermoscopic skin cancer classifiers.

BACKGROUND: Image-based cancer classifiers suffer from a variety of problems which negatively affect their performance. For example, variation in image brightness or different cameras can already suffice to diminish performance. Ensemble solutions, where multiple model predictions are combined into one, can improve these problems. However, ensembles are computationally intensive and less transparent to practitioners than single model solutions. Constructing model soups, by averaging the weights of multiple models into a single model, could circumvent these limitations while still improving performance. OBJECTIVE: To investigate the performance of model soups for a dermoscopic melanoma-nevus skin cancer classification task with respect to (1) generalisation to images from other clinics, (2) robustness against small image changes and (3) calibration such that the confidences correspond closely to the actual predictive uncertainties. METHODS: We construct model soups by fine-tuning pre-trained models on seven different image resolutions and subsequently averaging their weights. Performance is evaluated on a multi-source dataset including holdout and external components. RESULTS: We find that model soups improve generalisation and calibration on the external component while maintaining performance on the holdout component. For robustness, we observe performance improvements for pertubated test images, while the performance on corrupted test images remains on par. CONCLUSIONS: Overall, souping for skin cancer classifiers has a positive effect on generalisation, robustness and calibration. It is easy for practitioners to implement and by combining multiple models into a single model, complexity is reduced. This could be an important factor in achieving clinical applicability, as less complexity generally means more transparency.

Combining CNN-based histologic whole slide image analysis and patient data to improve skin cancer classification.

BACKGROUND: Clinicians and pathologists traditionally use patient data in addition to clinical examination to support their diagnoses. OBJECTIVES: We investigated whether a combination of histologic whole slides image (WSI) analysis based on convolutional neural networks (CNNs) and commonly available patient data (age, sex and anatomical site of the lesion) in a binary melanoma/nevus classification task could increase the performance compared with CNNs alone. METHODS: We used 431 WSIs from two different laboratories and analysed the performance of classifiers that used the image or patient data individually or three common fusion techniques. Furthermore, we tested a naive combination of patient data and an image classifier: for cases interpreted as 'uncertain' (CNN output score <0.7), the decision of the CNN was replaced by the decision of the patient data classifier. RESULTS: The CNN on its own achieved the best performance (mean ± standard deviation of five individual runs) with AUROC of 92.30% ± 0.23% and balanced accuracy of 83.17% ± 0.38%. While the classification performance was not significantly improved in general by any of the tested fusions, naive strategy of replacing the image classifier with the patient data classifier on slides with low output scores improved balanced accuracy to 86.72% ± 0.36%. CONCLUSION: In most cases, the CNN on its own was so accurate that patient data integration did not provide any benefit. However, incorporating patient data for lesions that were classified by the CNN with low 'confidence' improved balanced accuracy.

A benchmark for neural network robustness in skin cancer classification.

BACKGROUND: One prominent application for deep learning-based classifiers is skin cancer classification on dermoscopic images. However, classifier evaluation is often limited to holdout data which can mask common shortcomings such as susceptibility to confounding factors. To increase clinical applicability, it is necessary to thoroughly evaluate such classifiers on out-of-distribution (OOD) data. OBJECTIVE: The objective of the study was to establish a dermoscopic skin cancer benchmark in which classifier robustness to OOD data can be measured. METHODS: Using a proprietary dermoscopic image database and a set of image transformations, we create an OOD robustness benchmark and evaluate the robustness of four different convolutional neural network (CNN) architectures on it. RESULTS: The benchmark contains three data sets-Skin Archive Munich (SAM), SAM-corrupted (SAM-C) and SAM-perturbed (SAM-P)-and is publicly available for download. To maintain the benchmark's OOD status, ground truth labels are not provided and test results should be sent to us for assessment. The SAM data set contains 319 unmodified and biopsy-verified dermoscopic melanoma (n = 194) and nevus (n = 125) images. SAM-C and SAM-P contain images from SAM which were artificially modified to test a classifier against low-quality inputs and to measure its prediction stability over small image changes, respectively. All four CNNs showed susceptibility to corruptions and perturbations. CONCLUSIONS: This benchmark provides three data sets which allow for OOD testing of binary skin cancer classifiers. Our classifier performance confirms the shortcomings of CNNs and provides a frame of reference. Altogether, this benchmark should facilitate a more thorough evaluation process and thereby enable the development of more robust skin cancer classifiers.

Skin cancer classification via convolutional neural networks: systematic review of studies involving human experts.

BACKGROUND: Multiple studies have compared the performance of artificial intelligence (AI)-based models for automated skin cancer classification to human experts, thus setting the cornerstone for a successful translation of AI-based tools into clinicopathological practice. OBJECTIVE: The objective of the study was to systematically analyse the current state of research on reader studies involving melanoma and to assess their potential clinical relevance by evaluating three main aspects: test set characteristics (holdout/out-of-distribution data set, composition), test setting (experimental/clinical, inclusion of metadata) and representativeness of participating clinicians. METHODS: PubMed, Medline and ScienceDirect were screened for peer-reviewed studies published between 2017 and 2021 and dealing with AI-based skin cancer classification involving melanoma. The search terms skin cancer classification, deep learning, convolutional neural network (CNN), melanoma (detection), digital biomarkers, histopathology and whole slide imaging were combined. Based on the search results, only studies that considered direct comparison of AI results with clinicians and had a diagnostic classification as their main objective were included. RESULTS: A total of 19 reader studies fulfilled the inclusion criteria. Of these, 11 CNN-based approaches addressed the classification of dermoscopic images; 6 concentrated on the classification of clinical images, whereas 2 dermatopathological studies utilised digitised histopathological whole slide images. CONCLUSIONS: All 19 included studies demonstrated superior or at least equivalent performance of CNN-based classifiers compared with clinicians. However, almost all studies were conducted in highly artificial settings based exclusively on single images of the suspicious lesions. Moreover, test sets mainly consisted of holdout images and did not represent the full range of patient populations and melanoma subtypes encountered in clinical practice.