A mathematical model using computed tomography for the diagnosis of metastatic central compartment lymph nodes in papillary thyroid carcinoma.

OBJECTIVES

The purpose of this study was to establish a potential mathematical model for the diagnosis of the central compartment lymph node (LN) metastases of papillary thyroid carcinoma (PTC) using CT imaging.

METHODS

303 patients with PTC were enrolled. We determined the optimal cut-off points of LN size and nodal grouping by calculating the diagnostic value of each cut-off point. Then, calcification, cystic or necrotic change, abnormal enhancement, size, and nodal grouping were analysed by univariate and multivariate statistical methods. The mathematical model was obtained using binary logistic regression analysis, and a scoring system was developed for convenient use in clinical practice.

RESULTS

30 mm(2) for LNs area (size) and two LNs as the nodal grouping criterion had the best diagnostic value. The mathematical model was: p = e (y) /(1+ e (y) ), y = -0.670-0.087 × size + 1.010 × cystic or necrotic change + 1.371 × abnormal enhancement + 0.828 × nodal grouping + 0.909 × area. We assigned the value for cystic or necrotic change, abnormal enhancement, size and nodal grouping value as 25, 33, 20, and 22, respectively, yielding a scoring system.

CONCLUSIONS

This mathematical model has a high diagnostic value and is a convenient clinical tool.

KEY POINTS

• Papillary thyroid carcinoma has a relatively high rate of metastasis. • CT has unique advantages in evaluating the central compartment. • The mathematical model can improve the diagnosis of CT imaging. • Corresponding scoring system is a convenient clinical tool.