Can 3D Shape and Textural Analysis Differentiate Liposarcomas from Benign Lipomas?

Keywords

abdominal accuracy addition agreement ahead aided although assessed axes basis benign biopsy blindly bone border branch cells class classification classifier classifiers clinical coefficient combination combinations computed computer contingency contours convexity correctly cross define defined described descriptors determine determined diagnosis diagnostic differentiate differentiation eccentricity either electronics ellipse evaluated even exacerbate experienced feature features finding fingerprint fisher frank generated geometrical good grade gray growing heterogeneity hospital identified improved included independently initially inscribed institute inter internal joint kappa learning length lengths linear machine malignant margins mark matrix measuring medial median medical mining misclassified misidentified morphological morphologies occurrence particularly patient pelvic performance post practical preliminary problem processing promises proven quantitative radiologic radiologist radiologists radiology reader regarding remaining resolution retrospectively sampling seeded segmentation sensitivity seventy shape signs simple skeletal slice southern specificity spin spreading studies sufficient surgery table tables technical textural texture tools train tumor tumors turbo unapparent uniformity validation variance vulnerable white widths women years