JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Dual biomarkers CT-based deep learning model incorporating intrathoracic fat for discriminating benign and malignant pulmonary nodules in multi-center cohorts

  • 0School of Computer Science and Technology, Harbin University of Science and Technology, Harbin, China.
Physica Medica : Pm : an International Journal Devoted to the Applications of Physics to Medicine and Biology : Official Journal of the Italian Association of Biomedical Physics (aifb) +

|

December 17, 2024

|

December 17, 2024

View abstract on PubMed

Summary

This summary is machine-generated.

This study shows that incorporating intrathoracic fat (ITF) imaging into deep learning models significantly improves the discrimination of benign and malignant pulmonary nodules. Higher intrathoracic fat index (ITFI) is linked to a reduced risk of malignancy.

Area Of Science

  • Pulmonology
  • Radiology
  • Artificial Intelligence

Background

  • Body composition, especially fatty tissue, is a recognized prognostic factor in lung cancer.
  • Current clinical practice lacks methods to combine fatty tissue analysis with pulmonary nodule assessment for malignancy discrimination.

Purpose Of The Study

  • To develop a deep learning (DL) model utilizing dual imaging markers, including intrathoracic fat (ITF), for predicting malignancy in pulmonary nodules.
  • To explore the predictive value of ITF in differentiating benign from malignant pulmonary nodules.

Main Methods

  • A cohort of 1321 patients with pulmonary nodules from three centers was analyzed.
  • Deep learning was employed for feature extraction from computed tomography (CT) images of pulmonary nodules and ITF.
  • Multimodal information, including nodule and ITF data, was used for classification.

Main Results

  • The DL model incorporating ITF and pulmonary nodule data achieved superior performance (AUCs ranging from 0.899 to 0.922) compared to models using only nodule data.
  • Intrathoracic fat index (ITFI) was identified as an independent factor, with each unit increase correlating to a 9.4% decrease in malignancy risk.

Conclusions

  • Intrathoracic fat (ITF) shows potential as a noninvasive imaging biomarker for auxiliary prediction in pulmonary nodule assessment.
  • The integration of ITF into DL models enhances the accuracy of differentiating benign and malignant pulmonary nodules.

Keywords: