Qualitative CT Criterion for Subsolid Nodule Subclassification: Improving Interobserver Agreement and Pathologic...

Area of Science:

• Pulmonary Medicine
• Radiology
• Oncology

Background:

• Subsolid nodules (SSN) on computed tomography (CT) scans are crucial for diagnosing pulmonary adenocarcinoma.
• Accurate classification of SSNs is essential for predicting invasiveness and guiding treatment.
• Existing classification systems may have limitations in correlating with pathological findings.

Purpose of the Study:

• To evaluate the clinical validity of a novel qualitative CT criterion for classifying SSNs.
• To assess the correlation between the novel SSN classification and the pathologic invasiveness of pulmonary adenocarcinoma.
• To compare the performance of the novel classification with conventional SSN classification methods.

Main Methods:

• CT images of 41 SSNs were independently interpreted by six observers.
• Nodules were classified using both conventional and novel SSN classification systems.
• Interobserver and intraobserver agreement were assessed using the kappa (κ) coefficient; correlations with pathologic invasiveness were analyzed using Spearman correlation coefficients.

Main Results:

• Good interobserver agreement was observed for both classification systems (novel: κ=0.707; conventional: κ=0.702).
• High intraobserver agreement was noted for both systems (novel: κ=0.88; conventional: κ=0.92).
• The novel SSN classification demonstrated a stronger correlation with the pathologic invasiveness of adenocarcinoma spectrum lesions (correlation coefficient range 0.622-0.732) compared to the conventional classification (correlation coefficient range 0.458-0.644).

Conclusions:

• The novel SSN classification system exhibits good interobserver agreement.
• This novel classification shows a superior correlation with pathologic invasiveness compared to the conventional method.
• Further studies are warranted to validate these findings regarding interobserver agreement.