Predicting Invasiveness in Lung Adenocarcinoma with Ground-Glass Nodules Through Machine Learning and Radiomic Analysis of Clinical CT Data

1

Lung adenocarcinoma is the most common subtype of lung cancer, accounting for 55-60% of cases, often detected late due to limited patient awareness.
2

Early detection through screening, particularly low-dose spiral CT, significantly improves prognosis and survival rates in lung cancer patients.
3

Current methods for evaluating ground-glass nodules face challenges, including subjective radiologist assessments and diagnostic errors.
4

Radiomics technology extracts quantitative features from CT images, enhancing the assessment of lung adenocarcinoma and aiding clinical decision-making.
5

This study proposes a framework combining CT-based radiomics and deep learning to predict lung adenocarcinoma invasiveness preoperatively.

Bmc Cancer

by Mingzhi Lin, Longqian Li, Yiming Hui, Bin Li, Yue Li, ChongRui Li, Zhizhong Zheng, Zhuowen Yang
November 3, 2025

1