Evaluation of Melanoma Thickness with Clinical Close-up and Dermoscopic Images Using a Convolutional Neural Network.

Convolutional neural networks (CNNs) have shown promise in discriminating between invasive and in situ melanomas. The aim of this study was to analyse how a CNN model, integrating both clinical close-up and dermoscopic images, performed compared with 6 independent dermatologists. The secondary aim was to address which clinical and dermoscopic features dermatologists found to be suggestive of invasive and in situ melanomas, respectively. A retrospective investigation was conducted including 1,578 cases of paired images of invasive (n = 728, 46.1%) and in situ melanomas (n = 850, 53.9%). All images were obtained from the Department of Dermatology and Venereology at Sahlgrenska University Hospital and were randomized to a training set (n = 1,078), a validation set (n = 200) and a test set (n = 300). The area under the receiver operating characteristics curve (AUC) among the dermatologists ranged from 0.75 (95% confidence interval 0.70-0.81) to 0.80 (95% confidence interval 0.75-0.85). The combined dermatologists' AUC was 0.80 (95% confidence interval 0.77-0.86), which was significantly higher than the CNN model (0.73, 95% confidence interval 0.67-0.78, p = 0.001). Three of the dermatologists significantly outperformed the CNN. Shiny white lines, atypical blue-white structures and polymorphous vessels displayed a moderate interobserver agreement, and these features also correlated with invasive melanoma. Prospective trials are needed to address the clinical usefulness of CNN models in this setting.

Interobserver Agreement on Dermoscopic Features and their Associations with In Situ and Invasive Cutaneous Melanomas.

Several melanoma-specific dermoscopic features have been described, some of which have been reported as indicative of in situ or invasive melanomas. To assess the usefulness of these features to differentiate between these 2 categories, a retrospective, single-centre investigation was conducted. Dermoscopic images of melanomas were reviewed by 7 independent dermatologists. Fleiss' kappa (κ) was used to analyse interobserver agreement of predefined features. Logistic regression and odds ratios were used to assess whether specific features correlated with melanoma in situ or invasive melanoma. Overall, 182 melanomas (101 melanoma in situ and 81 invasive melanomas) were included. The interobserver agreement for melanoma-specific features ranged from slight to substantial. Atypical blue-white structures (κ=0.62, 95% confidence interval 0.59-0.65) and shiny white lines (κ=0.61, 95% confidence interval 0.58-0.64) had a substantial interobserver agreement. These 2 features were also indicative of invasive melanomas >1.0 mm in Breslow thickness. Furthermore, regression/peppering correlated with thin invasive melanomas. The overall agreement for classification of the lesions as invasive or melanoma in situ was moderate (κ=0.52, 95% confidence interval 0.49-0.56).

Discrimination Between Invasive and In Situ Melanomas Using Clinical Close-Up Images and a De Novo Convolutional Neural Network.

Background: Melanomas are often easy to recognize clinically but determining whether a melanoma is in situ (MIS) or invasive is often more challenging even with the aid of dermoscopy. Recently, convolutional neural networks (CNNs) have made significant and rapid advances within dermatology image analysis. The aims of this investigation were to create a de novo CNN for differentiating between MIS and invasive melanomas based on clinical close-up images and to compare its performance on a test set to seven dermatologists. Methods: A retrospective study including clinical images of MIS and invasive melanomas obtained from our department during a five-year time period (2016-2020) was conducted. Overall, 1,551 images [819 MIS (52.8%) and 732 invasive melanomas (47.2%)] were available. The images were randomized into three groups: training set (n = 1,051), validation set (n = 200), and test set (n = 300). A de novo CNN model with seven convolutional layers and a single dense layer was developed. Results: The area under the curve was 0.72 for the CNN (95% CI 0.66-0.78) and 0.81 for dermatologists (95% CI 0.76-0.86) (P < 0.001). The CNN correctly classified 208 out of 300 lesions (69.3%) whereas the corresponding number for dermatologists was 216 (72.0%). When comparing the CNN performance to each individual reader, three dermatologists significantly outperformed the CNN. Conclusions: For this classification problem, the CNN was outperformed by the dermatologist. However, since the algorithm was only trained and validated on 1,251 images, future refinement and development could make it useful for dermatologists in a real-world setting.

Can Dermoscopy Be Used to Predict if a Melanoma Is In Situ or Invasive?

BACKGROUND: The preoperative prediction of whether melanomas are invasive or in situ can influence initial management. OBJECTIVES: This study evaluated the accuracy rate, interobserver concordance, sensitivity and specificity in determining if a melanoma is invasive or in situ, as well as the ability to predict invasive melanoma thickness based on clinical and dermoscopic images. METHODS: In this retrospective, single-center investigation, 7 dermatologists independently reviewed clinical and dermoscopic images of melanomas to predict if they were invasive or in situ and, if invasive, their Breslow thickness. Fleiss' and Cohen's kappa (κ) were used for interobserver concordance and agreement with histopathological diagnosis. RESULTS: We included 184 melanomas (110 invasive and 74 in situ). Diagnostic accuracy ranged from 67.4% to 76.1%. Accuracy rates for in situ and invasive melanomas were 57.5% (95% confidence interval [CI], 53.1%-61.8%) and 81.7% (95% CI, 78.8%-84.4%), respectively. Interobserver concordance was moderate (κ = 0.47; 95% CI, 0.44-0.51). Sensitivity for predicting invasiveness ranged from 63.6% to 91.8% for 7 observers, while specificity was 32.4%-82.4%. For all correctly predicted invasive melanomas, agreement between predictions and correct thickness over or under 1.0 mm was moderate (κ = 0.52; 95% CI, 0.45-0.58). All invasive melanomas incorrectly predicted by any observer as in situ had a thickness <1.0 mm. All 32 melanomas >1.0 mm were correctly predicted to be invasive by all observers. CONCLUSIONS: Accuracy rates for predicting thick melanomas were excellent, melanomas inaccurately predicted as in situ were all thin, and interobserver concordance for predicting in situ or invasive melanomas was moderate. Preoperative dermoscopy of suspected melanomas is recommended for choosing appropriate surgical margins.