Using CT texture analysis to differentiate between peripheral lung cancer and pulmonary inflammatory pseudotumor

Area of Science:

• Radiology
• Medical Imaging
• Oncology

Background:

• Distinguishing peripheral lung cancer from pulmonary inflammatory pseudotumor (PIPT) is clinically significant.
• Accurate differentiation is crucial for appropriate patient management and treatment strategies.

Purpose of the Study:

• To investigate the utility of CT-radiomics features in differentiating peripheral lung cancer from PIPT.
• To identify specific radiomics features that can serve as biomarkers for distinguishing these conditions.

Main Methods:

• Retrospective collection of 18F-FDG PET/CT images from 21 PIPT and 21 peripheral lung cancer patients.
• Extraction of CT-radiomics features from regions of interest and screening for robustness using intra-class correlation coefficient (ICC).
• Statistical analysis and receiver operating characteristic (ROC) curve analysis to evaluate the discriminative ability of selected features.

Main Results:

• A total of 435 radiomics features were extracted, with 361 showing good repeatability (ICC ≥ 0.6).
• Twenty features demonstrated the ability to distinguish between peripheral lung cancer and PIPT, primarily from Gray-Level Co-occurrence Matrix, Intensity Histogram, and Shape features.
• The areas under the ROC curves (AUC) for these distinguishing features ranged from 0.717 to 0.748, indicating moderate to good discrimination.

Conclusions:

• CT-radiomics features extracted from non-contrast CT components of PET/CT images can effectively differentiate peripheral lung cancer from PIPT.
• Radiomics analysis holds promise as a non-invasive tool to support the diagnosis of these thoracic conditions.