Figure 4.

Correlation between sonometric and histometric measurement of tumour invasion depth in 400 patients with primary malignant melanoma. Simple regression analysis revealed a correlation coefficient R = 0.92 (p < 0.0001), while the slope of the regression line (b = 0.82) was substantially differing from perfect correlation between both measures, in which case b should equal 1.0. As can be seen from the graph, sonometry was systematically overestimating tumour depth. After recalculation of the estimations using a different ultrasound velocity for melanoma tissue of 1564 m/s instead, the value of slope only rose to b = 0.84, which still accounted for a substantial overestimation compared to histological measurement.