Agreement of dermatopathologists in the evaluation of clinically difficult melanocytic lesions: how golden is the 'gold standard'?

BACKGROUND: The 'gold standard' for the diagnosis of melanocytic lesions is dermatopathology. Although most of the diagnostic criteria are clearly defined, the interpretation of histopathology slides may be subject to interobserver variability. OBJECTIVES: The aim of this study was to determine the variability among dermatopathologists in the interpretation of clinically difficult melanocytic lesions. METHODS: This study used the database of MelaFind®, a computer-vision system for the diagnosis of melanoma. All lesions were surgically removed and sent for independent evaluation by four dermatopathologists. Agreement was calculated using kappa statistics. RESULTS: A total of 1,249 pigmented melanocytic lesions were included. There was a substantial agreement among expert dermatopathologists: two-category kappa was 0.80 (melanoma vs. non-melanoma) and three-category kappa was 0.62 (malignant vs. borderline vs. benign melanocytic lesions). The agreement was significantly greater for patients ≥40 years (three-category kappa = 0.67) than for younger patients (kappa = 0.49). In addition, the agreement was significantly lower for patients with atypical mole syndrome (AMS) (kappa = 0.31) than for patients without AMS (kappa = 0.76). LIMITATIONS: The data were limited by the inclusion/exclusion criteria of the MelaFind® study. This might represent a selection bias. The agreement was evaluated using kappa statistics. This is a standard method for evaluating agreement among pathologists, but might be considered controversial by some statisticians. CONCLUSIONS: Expert dermatopathologists have a high level of agreement when diagnosing clinically difficult melanocytic lesions. However, even among expert dermatopathologists, the current 'gold standard' is not perfect. Our results indicate that lesions from younger patients and patients with AMS may be more problematic for the dermatopathologists, suggesting that improved diagnostic criteria are needed for such patients.

Automatic differentiation of melanoma from melanocytic nevi with multispectral digital dermoscopy: a feasibility study.

BACKGROUND: Differentiation of melanoma from melanocytic nevi is difficult even for skin cancer specialists. This motivates interest in computer-assisted analysis of lesion images. OBJECTIVE: Our purpose was to offer fully automatic differentiation of melanoma from dysplastic and other melanocytic nevi through multispectral digital dermoscopy. METHOD: At 4 clinical centers, images were taken of pigmented lesions suspected of being melanoma before biopsy. Ten gray-level (MelaFind) images of each lesion were acquired, each in a different portion of the visible and near-infrared spectrum. The images of 63 melanomas (33 invasive, 30 in situ) and 183 melanocytic nevi (of which 111 were dysplastic) were processed automatically through a computer expert system to separate melanomas from nevi. The expert system used either a linear or a nonlinear classifier. The "gold standard" for training and testing these classifiers was concordant diagnosis by two dermatopathologists. RESULTS: On resubstitution, 100% sensitivity was achieved at 85% specificity with a 13-parameter linear classifier and 100%/73% with a 12-parameter nonlinear classifier. Under leave-one-out cross-validation, the linear classifier gave 100%/84% (sensitivity/specificity), whereas the nonlinear classifier gave 95%/68%. Infrared image features were significant, as were features based on wavelet analysis. CONCLUSION: Automatic differentiation of invasive and in situ melanomas from melanocytic nevi is feasible, through multispectral digital dermoscopy.

Precision of automatic measurements of pigmented skin lesion parameters with a MelaFind(TM) multispectral digital dermoscope.

The purpose of this study was to assess the precision of automatic computerized measurement of parameters that may be useful in the differentiation of malignant melanoma from benign pigmented skin lesions, and also to determine the feasibility of quantitative monitoring of skin lesions over time. Ten independent sequences of images were acquired with a MelaFind multispectral digital dermoscope for each of 12 benign or malignant pigmented skin lesions. The sequences of images were processed automatically to provide 10 independent measurements of the various parameters for each lesion. Parameters included lesion area, greatest 'diameter', perimeter, reflectance and asymmetry. The precision of each parameter determination was computed from the mean and standard deviation of the 10 measurements of that parameter. The relative errors in determining the lesion area, 'diameter' and perimeter were found to be 6%, 3% and 4%, respectively. Other lesion parameters that are used in differentiating melanomas from benign skin lesions were also analysed as a function of wavelength. In the blue band (about 430 nm) the relative error was about 7% for the mean lesion reflectance and about 7% for the asymmetry parameter. These results demonstrate the feasibility of using MelaFind for objective quantitative monitoring of changes in pigmented skin lesions over time. As suggested by some studies, such information is useful in the early detection of malignant melanoma. The results show that parameters obtained automatically from MelaFind images are sufficiently precise to allow pertinent parameters to be used to classify pigmented skin lesions.

Can early malignant melanoma be differentiated from atypical melanocytic nevus by in vivo techniques?: Part II. Automatic machine vision classification.

BACKGROUND/AIMS: Differentiation between early (Breslow thickness less than 1 mm) malignant melanoma (MM) and atypical melanocytic nevus (AMN) remains a challenge even to trained clinicians. The purpose of this study is to determine the feasibility of reliable discrimination between early MM and AMN with noninvasive, objective, automatic machine vision techniques. METHODS: A data base of 104 digitized dermoscopic color transparencies of melanocytic lesions was used to develop and test our computer-based algorithms for classification of such lesions as malignant (MM) or benign (AMN). Histopathologic diagnoses (30 MM and 74 AMN) were used as the "gold standard" for training and testing the algorithms. RESULTS: A fully automatic, objective technique for differentiating between early MM and AMN from their dermoscopic digital images was developed. The multiparameter linear classifier was trained to provide 100% sensitivity for MM. In the blind test, this technique did not miss a single MM and its specificity was comparable to that of skilled dermatologists. CONCLUSIONS: Reliable differentiation between early MM and AMN with high sensitivity is possible using machine vision techniques to analyze digitized dermoscopic lesion images.