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.

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.

A self-supervised vision transformer to predict survival from histopathology in renal cell carcinoma.

PURPOSE: To develop and validate an interpretable deep learning model to predict overall and disease-specific survival (OS/DSS) in clear cell renal cell carcinoma (ccRCC). METHODS: Digitised haematoxylin and eosin-stained slides from The Cancer Genome Atlas were used as a training set for a vision transformer (ViT) to extract image features with a self-supervised model called DINO (self-distillation with no labels). Extracted features were used in Cox regression models to prognosticate OS and DSS. Kaplan-Meier for univariable evaluation and Cox regression analyses for multivariable evaluation of the DINO-ViT risk groups were performed for prediction of OS and DSS. For validation, a cohort from a tertiary care centre was used. RESULTS: A significant risk stratification was achieved in univariable analysis for OS and DSS in the training (n = 443, log rank test, p < 0.01) and validation set (n = 266, p < 0.01). In multivariable analysis, including age, metastatic status, tumour size and grading, the DINO-ViT risk stratification was a significant predictor for OS (hazard ratio [HR] 3.03; 95%-confidence interval [95%-CI] 2.11-4.35; p < 0.01) and DSS (HR 4.90; 95%-CI 2.78-8.64; p < 0.01) in the training set but only for DSS in the validation set (HR 2.31; 95%-CI 1.15-4.65; p = 0.02). DINO-ViT visualisation showed that features were mainly extracted from nuclei, cytoplasm, and peritumoural stroma, demonstrating good interpretability. CONCLUSION: The DINO-ViT can identify high-risk patients using histological images of ccRCC. This model might improve individual risk-adapted renal cancer therapy in the future.

Dimethyl fumarate treatment in relapsed and refractory cutaneous T-cell lymphoma: a multicenter phase 2 study.

Targeted therapies for cutaneous T-cell lymphoma (CTCL) are limited and curative approaches are lacking. Furthermore, relapses and drug induced side effects are major challenges in the therapeutic management of patients with CTCL, creating an urgent need for new and effective therapies. Pathologic constitutive NF-κB activity leads to apoptosis resistance in CTCL cells and, thus, represents a promising therapeutic target in CTCL. In a preclinical study we showed the potential of dimethyl fumarate (DMF) to block NF-κB and, specifically, kill CTCL cells. To translate these findings to applications in a clinical setting, we performed a multicentric phase 2 study evaluating oral DMF therapy in 25 patients with CTCL stages Ib to IV over 24 weeks (EudraCT number 2014-000924-11/NCT number NCT02546440). End points were safety and efficacy. We evaluated skin involvement (using a modified severity weighted assessment tool [mSWAT]), pruritus, quality of life, and blood involvement, if applicable, as well as translational data. Upon skin analysis, 7 of 23 (30.4%) patients showed a response with >50% reduction in the mSWAT score. Patients with high tumor burden in the skin and blood responded best to DMF therapy. Although not generally significant, DMF also improved pruritus in several patients. Response in the blood was mixed, but we confirmed the NF-κB-inhibiting mechanism of DMF in the blood. The overall tolerability of the DMF therapy was very favorable, with mostly mild side effects. In conclusion, our study presents DMF as an effective and excellently tolerable therapeutic option in CTCL to be further evaluated in a phase 3 study or real-life patient care as well as in combination therapies. This trial was registered at www.clinicaltrials.gov as #NCT02546440.

Case report: Therapeutic potential of T-VEC in combination with MEK inhibitors in melanoma patients with NRAS mutation.

Mutations in the NRAS gene are common alterations in malignant melanoma. However, there are no specific treatment options approved for NRAS-mutated melanoma patients besides immune checkpoint inhibition. Since preclinical data suggests a synergistic effect of a MEK inhibitor (MEKi) and the oncolytic virus talimogene laherparepvec (T-VEC), we have treated three melanoma patients with this combination. All of the three patients had been suffering from recurring cutaneous and subcutaneous in-transit metastases. Upon treatment one patient (case 1) presented full regression of locoregional metastases and remained progression-free until date, for almost three years. The second patient (case 2) showed a partial regression of painful gluteal satellite metastases but died from brain metastases. The third patient (case 3) showed a durable response of locoregional metastases for seven months. The combination treatment was well tolerated with common adverse events known for each single agent. This report is the first case series presenting a clinical benefit of the combined T-VEC and MEKi treatment. We suggest the combination of T-VEC and MEKi as an off-label treatment option for patients with NRAS mutations, especially with recurrent in-transit or satellite metastases.

Multimodal integration of image, epigenetic and clinical data to predict BRAF mutation status in melanoma.

BACKGROUND: In machine learning, multimodal classifiers can provide more generalised performance than unimodal classifiers. In clinical practice, physicians usually also rely on a range of information from different examinations for diagnosis. In this study, we used BRAF mutation status prediction in melanoma as a model system to analyse the contribution of different data types in a combined classifier because BRAF status can be determined accurately by sequencing as the current gold standard, thus nearly eliminating label noise. METHODS: We trained a deep learning-based classifier by combining individually trained random forests of image, clinical and methylation data to predict BRAF-V600 mutation status in primary and metastatic melanomas of The Cancer Genome Atlas cohort. RESULTS: With our multimodal approach, we achieved an area under the receiver operating characteristic curve of 0.80, whereas the individual classifiers yielded areas under the receiver operating characteristic curve of 0.63 (histopathologic image data), 0.66 (clinical data) and 0.66 (methylation data) on an independent data set. CONCLUSIONS: Our combined approach can predict BRAF status to some extent by identifying BRAF-V600 specific patterns at the histologic, clinical and epigenetic levels. The multimodal classifiers have improved generalisability in predicting BRAF mutation status.

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.

Deep learning can predict survival directly from histology in clear cell renal cell carcinoma.

For clear cell renal cell carcinoma (ccRCC) risk-dependent diagnostic and therapeutic algorithms are routinely implemented in clinical practice. Artificial intelligence-based image analysis has the potential to improve outcome prediction and thereby risk stratification. Thus, we investigated whether a convolutional neural network (CNN) can extract relevant image features from a representative hematoxylin and eosin-stained slide to predict 5-year overall survival (5y-OS) in ccRCC. The CNN was trained to predict 5y-OS in a binary manner using slides from TCGA and validated using an independent in-house cohort. Multivariable logistic regression was used to combine of the CNNs prediction and clinicopathological parameters. A mean balanced accuracy of 72.0% (standard deviation [SD] = 7.9%), sensitivity of 72.4% (SD = 10.6%), specificity of 71.7% (SD = 11.9%) and area under receiver operating characteristics curve (AUROC) of 0.75 (SD = 0.07) was achieved on the TCGA training set (n = 254 patients / WSIs) using 10-fold cross-validation. On the external validation cohort (n = 99 patients / WSIs), mean accuracy, sensitivity, specificity and AUROC were 65.5% (95%-confidence interval [CI]: 62.9-68.1%), 86.2% (95%-CI: 81.8-90.5%), 44.9% (95%-CI: 40.2-49.6%), and 0.70 (95%-CI: 0.69-0.71). A multivariable model including age, tumor stage and metastasis yielded an AUROC of 0.75 on the TCGA cohort. The inclusion of the CNN-based classification (Odds ratio = 4.86, 95%-CI: 2.70-8.75, p < 0.01) raised the AUROC to 0.81. On the validation cohort, both models showed an AUROC of 0.88. In univariable Cox regression, the CNN showed a hazard ratio of 3.69 (95%-CI: 2.60-5.23, p < 0.01) on TCGA and 2.13 (95%-CI: 0.92-4.94, p = 0.08) on external validation. The results demonstrate that the CNN's image-based prediction of survival is promising and thus this widely applicable technique should be further investigated with the aim of improving existing risk stratification in ccRCC.

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.

Clinical and molecular characteristics associated with response to therapeutic PD-1/PD-L1 inhibition in advanced Merkel cell carcinoma.

BACKGROUND: Based on its viral-associated or UV-associated carcinogenesis, Merkel cell carcinoma (MCC) is a highly immunogenic skin cancer. Thus, clinically evident MCC occurs either in immuno-compromised patients or based on tumor-intrinsic immune escape mechanisms. This notion may explain that although advanced MCC can be effectively restrained by treatment with PD-1/PD-L1 immune checkpoint inhibitors (ICIs), a considerable percentage of patients does not benefit from ICI therapy. Biomarkers predicting ICI treatment response are currently not available. METHODS: The present multicenter retrospective study investigated clinical and molecular characteristics in 114 patients with unresectable MCC at baseline before treatment with ICI for their association with therapy response (best overall response, BOR). In a subset of 21 patients, pretreatment tumor tissue was analyzed for activation, differentiation and spatial distribution of tumor infiltrating lymphocytes (TIL). RESULTS: Of the 114 patients, n=74 (65%) achieved disease control (BOR=complete response/partial response/stable disease) on ICI. A Bayesian cumulative ordinal regression model revealed absence of immunosuppression and a limited number of tumor-involved organ systems was highly associated with a favorable therapy response. Unimpaired overall performance status, high age, normal serum lactate dehydrogenase and normal serum C reactive protein were moderately associated with disease control. While neither tumor Merkel cell polyomavirus nor tumor PD-L1 status showed a correlation with therapy response, treatment with anti-PD-1 antibodies was associated with a higher probability of disease control than treatment with anti-PD-L1 antibodies. Multiplexed immunohistochemistry demonstrated the predominance of CD8+ effector and central memory T cells (TCM) in close proximity to tumor cells in patients with a favorable therapy response. CONCLUSIONS: Our findings indicate the absence of immunosuppression, a limited number of tumor-affected organs, and a predominance of CD8+ TCM among TIL, as baseline parameters associated with a favorable response to PD-1/PD-L1 ICI therapy of advanced MCC. These factors should be considered when making treatment decisions in MCC patients.

Incidence of pulmonary embolism and impact on mortality in patients with malignant melanoma.

PURPOSE: Pulmonary embolism (PE) occurs frequently in patients with malignant melanoma (MM). The aim of this study is to determine the incidence of PE in patients with MM and to assess the clinical characteristics and mortality of MM patients with PE. MATERIAL AND METHODS: Medical records from 381 MM patients who underwent contrast-enhanced computed tomography were evaluated. Imaging parameters including location of PE and measurements of right heart dysfunction and clinical parameters including D-Dimer levels, local and distant tumor stage and time of death were analyzed. RESULTS: PE was found in 23/381 (6%) MM patients, whereby 17/23 (74%) were detected incidentally and only 6/23 (26%) were symptomatic. The presence of PE significantly correlated with elevated D-Dimers (p < 0.001), right ventricular dysfunction (p = 0.04), higher local tumor stage (≥T3) (p = 0.05), presence of visceral (p = 0.02) or cerebral metastases (p = 0.03) and increased mortality (p = 0.05). Further, patients with central PE showed an increased mortality compared to peripheral PE (p = 0.03), but no correlation was found between the localization of PE and the occurrence of clinical symptoms (p = 0.36). CONCLUSION: PE in patients with MM often occurs without clinical symptoms and is indicative for advanced disease and a poorer prognosis.

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.

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.

Gastrointestinal cancer classification and prognostication from histology using deep learning: Systematic review.

BACKGROUND: Gastrointestinal cancers account for approximately 20% of all cancer diagnoses and are responsible for 22.5% of cancer deaths worldwide. Artificial intelligence-based diagnostic support systems, in particular convolutional neural network (CNN)-based image analysis tools, have shown great potential in medical computer vision. In this systematic review, we summarise recent studies reporting CNN-based approaches for digital biomarkers for characterization and prognostication of gastrointestinal cancer pathology. METHODS: Pubmed and Medline were screened for peer-reviewed papers dealing with CNN-based gastrointestinal cancer analyses from histological slides, published between 2015 and 2020.Seven hundred and ninety titles and abstracts were screened, and 58 full-text articles were assessed for eligibility. RESULTS: Sixteen publications fulfilled our inclusion criteria dealing with tumor or precursor lesion characterization or prognostic and predictive biomarkers: 14 studies on colorectal or rectal cancer, three studies on gastric cancer and none on esophageal cancer. These studies were categorised according to their end-points: polyp characterization, tumor characterization and patient outcome. Regarding the translation into clinical practice, we identified several studies demonstrating generalization of the classifier with external tests and comparisons with pathologists, but none presenting clinical implementation. CONCLUSIONS: Results of recent studies on CNN-based image analysis in gastrointestinal cancer pathology are promising, but studies were conducted in observational and retrospective settings. Large-scale trials are needed to assess performance and predict clinical usefulness. Furthermore, large-scale trials are required for approval of CNN-based prediction models as medical devices.

Deep learning approach to predict sentinel lymph node status directly from routine histology of primary melanoma tumours.

AIM: Sentinel lymph node status is a central prognostic factor for melanomas. However, the surgical excision involves some risks for affected patients. In this study, we therefore aimed to develop a digital biomarker that can predict lymph node metastasis non-invasively from digitised H&E slides of primary melanoma tumours. METHODS: A total of 415 H&E slides from primary melanoma tumours with known sentinel node (SN) status from three German university hospitals and one private pathological practice were digitised (150 SN positive/265 SN negative). Two hundred ninety-one slides were used to train artificial neural networks (ANNs). The remaining 124 slides were used to test the ability of the ANNs to predict sentinel status. ANNs were trained and/or tested on data sets that were matched or not matched between SN-positive and SN-negative cases for patient age, ulceration, and tumour thickness, factors that are known to correlate with lymph node status. RESULTS: The best accuracy was achieved by an ANN that was trained and tested on unmatched cases (61.8% ± 0.2%) area under the receiver operating characteristic (AUROC). In contrast, ANNs that were trained and/or tested on matched cases achieved (55.0% ± 3.5%) AUROC or less. CONCLUSION: Our results indicate that the image classifier can predict lymph node status to some, albeit so far not clinically relevant, extent. It may do so by mostly detecting equivalents of factors on histological slides that are already known to correlate with lymph node status. Our results provide a basis for future research with larger data cohorts.

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.