Value of volume-rendering computed tomography for diagnosing different solitary pulmonary nodules and invasion depth in lung adenocarcinoma

We aimed to analyze the value of volume rendering (VR) in diagnosing different solitary pulmonary nodules (SPNs) with diameter less than 1.0 cm and assessing invasion depth in lung adenocarcinoma.

In total, 908 patients with SPN that was confirmed by postoperative pathology were included, followed by an analysis of the imaging characteristics (including microvascular sign, vascular convergence, lobulation, and spiculation) of malignant and benign SPN based on VR. Moreover, the detection rates of imaging signs of three types of malignant SPNs (pure ground grass nodule, pGGN; part-solid nodule; and solid nodule) classified by SPN density and three invasion depths of adenocarcinoma (pre-invasion lesion, PIL; micro invasive adenocarcinoma, MIA; and invasive adenocarcinoma, IAC) were also analyzed.

The microvascular sign detection rate was higher while vascular convergence and spiculation detection rates were lower in malignant SPN than in benign SPN. The microvascular sign possessed high sensitivity (82%) and specificity (72%) in predicting malignant and benign SPNs. The microvascular sign detection rate decreased while vascular convergence, lobulation, and spiculation detection increased with the rising density of malignant SPN. Furthermore, the detection rates of the four imaging signs all increased with the adenocarcinoma invasion depth. Microvascular sign showed good detecting ability in low density SPNs pGGN (81.8%), part-solid nodules (95.8%), and in all three invasion depths of adenocarcinoma (PIL [68.2%], MIS [95.3%], and IAC [87.2%]).

These imaging features distinguished by VR exhibited an excellent differential diagnostic ability of various SPNs as well as invasion depth of lung adenocarcinoma.