Development and Testing of Multivariate Statistical Model To Predict Malignancy of Small (<1.5cm) Pulmonary Nodules.

Improved pulmonary nodule classification utilizing quantitative lung parenchyma features

“…In reviewing the current literature, the size, intensity, shape, texture, and location of the nodule have all been found to be significant features in determining the diagnoses of lung nodules. [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] Several methods have been used to characterize the shape of the nodule including sphericity 8,17 and compactness, 11,17 as well as qualitative descriptors of polygonal shape, 18 edge characteristics, 7,19,20 and the presence of spiculation. 6,9,13 The texture of the nodule has also been quantified through different techniques, such as co-occurrence matrix-based parameters and run-length statistics.…”

Pushing the boundaries

“…In reviewing the current literature, size, intensity, shape, texture, and location of the nodule have all been found to be significant features in determining the diagnoses of lung nodules [19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38]. Several methods have been used to characterize the shape of the nodule including sphericity [22,31] and compactnesss [25,31], as well as qualitative descriptors of polygonal shape [21,32], edge characteristics [20,33,34], and the presence of spiculation [19,21,23,27]. The texture of the nodule has also been quantified through different techniques, such as co-occurrence matrix-based parameters, run-length statistics, and fractional Brownian motion [21,29,30,36,37].…”

“…CT manufacturers (Siemens (162), GE(15), Philips(2), and Toshiba(20)) and18 CT models were used. The most common models were: Siemens SOMATOM Definition(71), Siemens Sensation 16 (46), Sensation Biograph 40 (15), and Toshiba Aquilion (20).…”