Histopathological Melanoma Research Articles

Coarse-to-fine feature representation based on deformable partition attention for melanoma identification

In the histopathological melanoma image diagnosis system, manual identification of super-scale slides with dense cells is tedious, time-consuming, and subjective. To deal with this problem, we propose an automatic identification network based on the deformable partition attention to identify lots of dense slides as an assistant. A coarse-to-fine strategy is adopted in feature representation and qualitative identification to improve the identification accuracy of melanomas and nevi. First of all, because it is difficult to extract features in the lesion area with blurred boundaries and uneven distribution, we develop a deformable partition attention module, which integrates the advantage of the attention mechanism and deformable convolution. The module overcomes the limitation of rectangular convolution and gradually refines the channel and spatial features, which enriches feature representation by combining global and local features. Secondly, to address the problem of difficult convergence and poor recognition rate caused by the excessive non-aligned distance between benign-malignant and benign subcategories, we propose a progressive architecture via a coarse sub-network closely followed by a fine sub-network. Moreover, to further increase the inter-class differences and reduce the intra-class disparities, we propose a joint loss function to mine hard samples, which effectively improves the identification performance. Experimental results on the clinical dataset show that the proposed algorithm has higher sensitivity and specificity and outperforms state-of-the-art deep neural networks.

Deep learning outperformed 11 pathologists in the classification of histopathological melanoma images

BackgroundThe diagnosis of most cancers is made by a board-certified pathologist based on a tissue biopsy under the microscope. Recent research reveals a high discordance between individual pathologists. For melanoma, the literature reports on 25–26% of discordance for classifying a benign nevus versus malignant melanoma. A recent study indicated the potential of deep learning to lower these discordances. However, the performance of deep learning in classifying histopathologic melanoma images was never compared directly to human experts. The aim of this study is to perform such a first direct comparison. MethodsA total of 695 lesions were classified by an expert histopathologist in accordance with current guidelines (350 nevi/345 melanoma). Only the haematoxylin & eosin (H&E) slides of these lesions were digitalised via a slide scanner and then randomly cropped. A total of 595 of the resulting images were used to train a convolutional neural network (CNN). The additional 100 H&E image sections were used to test the results of the CNN in comparison to 11 histopathologists. Three combined McNemar tests comparing the results of the CNNs test runs in terms of sensitivity, specificity and accuracy were predefined to test for significance (p < 0.05). FindingsThe CNN achieved a mean sensitivity/specificity/accuracy of 76%/60%/68% over 11 test runs. In comparison, the 11 pathologists achieved a mean sensitivity/specificity/accuracy of 51.8%/66.5%/59.2%. Thus, the CNN was significantly (p = 0.016) superior in classifying the cropped images. InterpretationWith limited image information available, a CNN was able to outperform 11 histopathologists in the classification of histopathological melanoma images and thus shows promise to assist human melanoma diagnoses.

Pathologist-level classification of histopathological melanoma images with deep neural networks

BackgroundThe diagnosis of most cancers is made by a board-certified pathologist based on a tissue biopsy under the microscope. Recent research reveals a high discordance between individual pathologists. For melanoma, the literature reports 25–26% of discordance for classifying a benign nevus versus malignant melanoma. Deep learning was successfully implemented to enhance the precision of lung and breast cancer diagnoses. The aim of this study is to illustrate the potential of deep learning to assist human assessment for a histopathologic melanoma diagnosis. MethodsSix hundred ninety-five lesions were classified by an expert histopathologist in accordance with current guidelines (350 nevi and 345 melanomas). Only the haematoxylin and eosin stained (H&E) slides of these lesions were digitalised using a slide scanner and then randomly cropped. Five hundred ninety-five of the resulting images were used for the training of a convolutional neural network (CNN). The additional 100 H&E image sections were used to test the results of the CNN in comparison with the original class labels. FindingsThe total discordance with the histopathologist was 18% for melanoma (95% confidence interval [CI]: 7.4–28.6%), 20% for nevi (95% CI: 8.9–31.1%) and 19% for the full set of images (95% CI: 11.3–26.7%). InterpretationEven in the worst case, the discordance of the CNN was about the same compared with the discordance between human pathologists as reported in the literature. Despite the vastly reduced amount of data, time necessary for diagnosis and cost compared with the pathologist, our CNN archived on-par performance. Conclusively, CNNs indicate to be a valuable tool to assist human melanoma diagnoses.

Malignant Melanoma: Diagnostic and Management Update.

After studying this article, the participant should be able to: 1. Summarize the changes to the American Joint Committee on Cancer Eighth Edition Melanoma Staging System. 2. List advances in genetic, molecular, and histopathologic melanoma diagnosis and prognostication. 3. Recommend sentinel lymph node biopsy and appropriate surgical margins based on individualized patient needs. 4. Recognize the currently available treatments for in-transit metastasis and advanced melanoma. 5. Describe current and future therapies for melanoma with distant visceral or brain metastases. Strides in melanoma surveillance, detection, and treatment continue to be made. The American Joint Committee on Cancer Eighth Edition Cancer Staging System has improved risk stratification of patients, introduced new staging categories, and resulted in stage migration of patients with improved outcomes. This review summarizes melanoma advances of the recent years with an emphasis on the surgical advances, including techniques and utility of sentinel node biopsy, controversies in melanoma margin selection, and the survival impact of time-to-treatment metrics. Once a disease manageable only with surgery, a therapeutic paradigm shift has given a more promising outlook to melanoma patients at any stage. Indeed, a myriad of novel, survival-improving immunotherapies have been introduced for metastatic melanoma and more recently in the high-risk adjuvant setting.

Nevus count associations with pigmentary phenotype, histopathological melanoma characteristics and survival from melanoma.

Although nevus count is an established risk factor for melanoma, relationships between nevus number and patient and tumor characteristics have not been well studied and the influence of nevus count on melanoma-specific survival is equivocal. Using data from the Genes, Environment and Melanoma (GEM) study, a large population-based study of primary cutaneous melanoma, we evaluated associations between number of nevi and patient features, including sun-sensitivity summarized in a phenotypic index, and tumor characteristics. We also assessed the association of nevus count with melanoma-specific survival. Higher nevus counts were independently and positively associated with male gender and younger age at diagnosis, and they were inversely associated with lentigo maligna histology. We observed a borderline significant trend of poorer melanoma-specific survival with increasing quartile of nevus count, but little or no association between number of nevi and pigmentary phenotypic characteristics or prognostic tumor features.

Nevus count associations with pigmentary phenotype, histopathological melanoma characteristics and survival from melanoma

Nevus count associations with pigmentary phenotype, histopathological melanoma characteristics and survival from melanoma

Inherited variation at MC1R and histological characteristics of primary melanoma.

Variation in the melanocortin-1receptor (MC1R) gene is associated with pigmentary phenotypes and risk of malignant melanoma. Few studies have reported on MC1R variation with respect to tumor characteristics, especially clinically important prognostic features. We examined associations between MC1R variants and histopathological melanoma characteristics. Study participants were enrolled from nine geographic regions in Australia, Canada, Italy and the United States and were genotyped for MC1R variants classified as high-risk [R] (D84E, R142H, R151C, R160W, and D294H, all nonsense and insertion/deletion) or low-risk [r] (all other nonsynonymous) variants. Tissue was available for 2,160 white participants of the Genes, Environment and Melanoma (GEM) Study with a first incident primary melanoma diagnosis, and underwent centralized pathologic review. No statistically significant associations were observed between MC1R variants and AJCC established prognostic tumor characteristics: Breslow thickness, presence of mitoses or presence of ulceration. However, MC1R was significantly associated with anatomic site of melanoma (p = 0.002) and a positive association was observed between carriage of more than one [R] variant and melanomas arising on the arms (OR = 2.39; 95% CI: 1.40, 4.09). We also observed statistically significant differences between sun-sensitive and sun-resistant individuals with respect to associations between MC1R genotype and AJCC prognostic tumor characteristics. Our results suggest inherited variation in MC1R may play an influential role in anatomic site presentation of melanomas and may differ with respect to skin pigmentation phenotype.

Clinical and histologic trends of melanoma

Background: Several changing clinical and histopathologic melanoma trends occurred from the 1950s to the 1980s. Objective: The purpose of this study was to evaluate melanoma trends during the past decade and to compare present trends to those documented during the past four decades. Methods: Sex, age at diagnosis, location, tumor thickness, stage, and histologic subtypes were evaluated from 1984 to 1995 and compared with trends during the past four decades. Results: Most changing trends from the past four decades have slowed or stabilized during the past decade. Conclusion: Complete reporting of all melanomas to central tumor registries is necessary to accurately analyze present and future melanoma trends. Ongoing and new prevention and control strategies beginning at birth may be necessary to continue the positive efforts to curtail the melanoma epidemic.(J Am Acad Dermatol 1998;38:681-6.)