Machine Learning in Melanoma Diagnosis. Limitations About to be Overcome

BACKGROUND

Automated image classification is a promising branch of machine learning (ML) useful for skin cancer diagnosis, but little has been determined about its limitations for general usability in current clinical practice.

OBJECTIVE

To determine limitations in the selection of skin cancer images for ML analysis, particularly in melanoma.

METHODS

Retrospective cohort study design, including 2,849 consecutive high-quality dermoscopy images of skin tumors from 2010 to 2014, for evaluation by a ML system. Each dermoscopy image was assorted according to its eligibility for ML analysis.

RESULTS

Of the 2,849 images chosen from our database, 968 (34%) met the inclusion criteria for analysis by the ML system. Only 64.7% of nevi and 36.6% of melanoma met the inclusion criteria. Of the 528 melanomas, 335 (63.4%) were excluded. An absence of normal surrounding skin (40.5% of all melanomas from our database) and absence of pigmentation (14.2%) were the most common reasons for exclusion from ML analysis.

DISCUSSION

Only 36.6% of our melanomas were admissible for analysis by state-of-the-art ML systems. We conclude that future ML systems should be trained on larger datasets which include relevant non-ideal images from lesions evaluated in real clinical practice. Fortunately, many of these limitations are being overcome by the scientific community as recent works show.

Dermoscopy of inflammatory breast cancer

Inflammatory breast cancer is an aggressive and infiltrative malignancy that is often misdiagnosed as an infection because of its symptoms and signs of inflammation, delaying proper diagnosis and treatment. We report a case of inflammatory breast cancer showing correlation between dermoscopic and histopathological diagnoses. We highlight the utility of dermoscopy for skin biopsy site selection.

Melanocortin 1 receptor (MC1R) polymorphisms' influence on size and dermoscopic features of nevi

The melanocortin 1 receptor (MC1R) is a highly polymorphic gene. The loss-of-function MC1R variants ("R") have been strongly associated with red hair color phenotype and an increased melanoma risk. We sequenced the MC1R gene in 175 healthy individuals to assess the influence of MC1R on nevus phenotype. We identified that MC1R variant carriers had larger nevi both on the back [p-value = .016, adjusted for multiple parameters (adj. p-value)] and on the upper limbs (adj. p-value = .007). Specifically, we identified a positive association between the "R" MC1R variants and visible vessels in nevi [p-value = .033, corrected using the FDR method for multiple comparisons (corrected p-value)], dots and globules in nevi (corrected p-value = .033), nevi with eccentric hyperpigmentation (corrected p-value = .033), a high degree of freckling (adj. p-value = .019), and an associative trend with presence of blue nevi (corrected p-value = .120). In conclusion, the MC1R gene appears to influence the nevus phenotype.