Relationship between histological and computer based assessment of melanoma diameter and thickness in head & neck vs. trunk melanoma.

BackgroundDiameter represents a controversial parameter in the diagnosis of melanoma (MM). The ABCD rule considers it an important diagnostic parameter, while some authors decrease its relevance. Also, its measurement is not always reliable, histological methods being usually used. The current study aims to compare histological and digital measurements and to evaluate the correlation between MM head- and trunk diameter and thickness. Methods The study population was subdivided according to diameter subgroups considering head- and trunk-limbs lesions separately. Digital diameters measured by an automatic software referring to 477 MM images were compared to diameters reported on the pathologist’s records. Clinical and histological information was also considered, and the correlation between diameter and thickness was assessed. Odds ratios (OR) were computed for different diameter subgroups. Results Mean digital diameters of head MMs and trunk limbs MMs exceeded histological measurements by 11% and 20%, respectively. In head MMs, no correlation between diameter and thickness was observable, whereas in the trunk-limbs group a direct relationship between thickness and diameter was noticed. ORs for non in situ vs in situ andfor ≥ 1 mm vs < 1 mm thick were low for small lesions, increasing for larger ones, indicating that the latter are more likely to be thick. ConclusionMM diameter should be assessed digitally to avoid tissue shrinkage after biopsy and imprecise in vivo measurements. Although nearly 10% of MMs might escape an early diagnosis based only on the D of the ABCD rule, MM diameter may be related to its thickness.The evaluation with a computer vision system should be recommended for small pigmented lesions (<6mm) in order to reduce the percentage of misdiagnosed smaller MMs and to better evaluate the parameter E of the ABCDE rule (Evolving lesion).