Low-Dose CT Screening of Persistent Subsolid Lung Nodules: First-Order Features in Radiomics

Area of Science:

• Radiology
• Oncology
• Medical Imaging Analysis

Background:

• Nondisappearing subsolid nodules are common in lung cancer screening.
• Assessing invasiveness changes in slow-growing nodules is challenging.
• Radiomics offers a potential solution for automated detection.

Purpose of the Study:

• Develop an automated radiomics method for invasive tumor identification in subsolid nodules.
• Utilize low-dose computed tomography (LD-CT) data for analysis.
• Evaluate the effectiveness of the developed method.

Main Methods:

• Analysis of LD-CT scans from patients with lung adenocarcinoma.
• Categorization of nodules into invasive and less-invasive groups.
• Assessment of annual changes in nodule histogram features (e.g., skewness, variance) using 3D CT.

Main Results:

• Δskewness emerged as an independent predictor of invasiveness.
• Combining Δskewness and Δvariance achieved 88% true-positive detection rate.
• The radiomics model suggested earlier surgical intervention compared to volume doubling time.

Conclusions:

• LD-CT radiomics shows potential for timely detection of invasive subsolid nodules.
• This technology can improve monitoring of nodules requiring follow-up.
• Automated analysis aids in clinical decision-making for lung cancer screening.