Sign in Subscribe
Cancer research

Radiomics-Based Machine Learning Model Accurately Classifies Non-Small Cell Lung Cancer Subtypes

Strategic Insights

— 2 min read

Researchers from Sichuan University in Chengdu, People's Republic of China, have developed and evaluated a radiomics-based machine learning model to differentiate between lung squamous cell carcinoma (SCC) and adenocarcinoma (AC) in patients with brain metastases (BMs). The study used T1-enhanced magnetic resonance imaging (MRI) features and identified 833 radiomic features from the segmented lesions using the PyRadiomics package. The model, which performed the best among the ten classifier models built, achieved an accuracy of 0.814, with a sensitivity of 0.726 and specificity of 0.896 in the training dataset.

The researchers obtained promising results, as the model effectively distinguished between SCC and AC based on radiomic features. This research demonstrates the efficacy of a radiomics-based machine learning model in accurately classifying non-small cell lung cancer (NSCLC) subtypes from BMs, providing a valuable noninvasive tool for guiding personalized treatment strategies.

The findings of this study have significant implications for personalized medicine in the treatment of non-small cell lung cancer. The radiomics-based machine learning model can help clinicians identify the subtype of lung cancer earlier and develop targeted treatment plans, potentially improving patient outcomes.

Key Takeaways:

  • Researchers developed a radiomics-based machine learning model to differentiate between lung squamous cell carcinoma (SCC) and adenocarcinoma (AC) in patients with brain metastases (BMs).
  • The model used T1-enhanced magnetic resonance imaging (MRI) features and identified 833 radiomic features from the segmented lesions using the PyRadiomics package.
  • The LightGBM model performed the best among the ten classifier models built, achieving an accuracy of 0.814, with a sensitivity of 0.726 and specificity of 0.896.
  • The model effectively distinguished between SCC and AC based on radiomic features, highlighting its potential for noninvasive non-small cell lung cancer (NSCLC) subtype classification.
  • The study demonstrates the efficacy of a radiomics-based machine learning model in accurately classifying NSCLC subtypes from BMs, providing a valuable noninvasive tool for guiding personalized treatment strategies.
  • Further validation on larger, multi-center datasets is crucial to verify these findings.

Statistics:

  • 173 patients with BMs were included in the study, consisting of 88 with AC and 85 with SCC.
  • 833 radiomic features were identified from the segmented lesions using the PyRadiomics package.
  • The LightGBM model achieved an accuracy of 0.814, with a sensitivity of 0.726 and specificity of 0.896 in the training dataset.
  • The model maintained an accuracy of 0.779, with a sensitivity of 0.725 and specificity of 0.857 in the testing dataset.
  • The AUC was 0.858 for the training dataset and 0.857 for the testing dataset.

Sources:

  • Frontiers in Oncology. Radiomics-based machine learning for differentiating lung squamous cell carcinoma and adenocarcinoma using T1-enhanced MRI of brain metastases. 2025;15:1599853.
  • Qiaoyue Tan, Radiotherapy Physics and Technology Center, Cancer Center, West China Hospital, Sichuan University, Chengdu, People's Republic of China.
  • Xueming Xia, Wei Du, Qiheng Gou. The publisher's contact information for the journal Frontiers in Oncology is: Frontiers Media Sa, Avenue Du Tribunal Federal 34, Lausanne, Ch-1015, Switzerland.

Read more