2023 World Conference on Lung Cancer (Posters)-P1.13. A Multi-Stage Imaging-Based Deep Learning Classifier for EGFR-TKI Response Prediction in EGFR mutant Non-small Cell Lung Cancer

P1.13. A Multi-Stage Imaging-Based Deep Learning Classifier for EGFR-TKI Response Prediction in EGFR mutant Non-small Cell Lung Cancer - PDF(Abstract)

Back to course

Pdf Summary

This study presented a novel multi-stage imaging-based deep learning classifier for predicting the response to epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI) therapy in non-small cell lung cancer (NSCLC) patients with EGFR mutations. The researchers aimed to accurately assess the efficacy and drug resistance state of EGFR-TKI therapy in order to make personalized treatment decisions and improve patients' prognosis.<br /><br />The researchers proposed a multi-stage CT image-based model that analyzed the dynamic tumor patterns observed in patients' follow-up visits. They conducted a multi-center retrospective study involving 274 patients and 548 CT sections. The results showed that their deep learning-based model accurately predicted drug resistance risk, measured by progression-free survival risk, with 8-month area under the curve (AUC) values of 0.803, 0.910, and 0.858 in three different cohorts. The model successfully stratified patients into high-risk and low-risk groups, showing significant differences in median progression-free survival between the groups.<br /><br />The researchers concluded that their multi-stage imaging-based deep learning classifier outperformed single-stage CT image-based methods for predicting EGFR-TKI response. Additionally, the model provided new visual clues for risk estimation by analyzing dynamic tumor patterns. This suggests the potential for using longitudinal CT image information to accurately monitor EGFR-TKI treatment.<br /><br />In summary, this study presented a novel approach using deep learning and multi-stage imaging to predict the response to EGFR-TKI therapy in NSCLC patients with EGFR mutations. The results demonstrated the effectiveness of the model in predicting drug resistance risk and stratifying patients into risk groups. The findings have implications for personalized treatment decisions and monitoring of EGFR-TKI therapy.

Asset Subtitle

Yang Xia

Meta Tag

Speaker Yang Xia

Topic Screening & Early Detection: Biomarkers/Imaging Technology

Keywords

multi-stage imaging-based deep learning classifier

predicting response

epidermal growth factor receptor-tyrosine kinase inhibitor

EGFR-TKI therapy

non-small cell lung cancer

NSCLC patients

drug resistance

CT image-based model

dynamic tumor patterns

personalized treatment decisions