Search papers, labs, and topics across Lattice.
This study explores the use of PET/CT imaging combined with deep learning to predict key mutations (EGFR, TP53, KRAS) in non-small cell lung cancer (NSCLC) through a multi-label learning approach. The research highlights the limitations of traditional single-gene classification by demonstrating that joint prediction can enhance accuracy for certain mutation pairs, achieving AUC improvements for KRAS and TP53. These findings indicate that the effectiveness of multi-label learning is mutation-specific, suggesting tailored modeling strategies for better radiogenomic predictions in NSCLC.
Jointly predicting mutations in lung cancer can significantly boost diagnostic accuracy, but only for the right combinations of genes.
Lung cancer remains one of the leading causes of cancer- related mortality worldwide. Although targeted therapies have improved outcomes for patients with non-small cell lung cancer (NSCLC), they rely on mutation profiling through tissue biopsy, an invasive procedure with several limitations. This study investigates PET/CT-based radio- genomic prediction of epidermal growth factor receptor (EGFR), tumour protein 53 (TP53), and Kirsten rat sarcoma viral oncogene (KRAS) mutations using deep learning. We further evaluate whether pairwise multi-label learning improves mutation prediction compared with conventional single-gene classification. To the best of our knowledge, this is among the first studies to systematically investigate multi-label learning for PET/CT radiogenomic mutation prediction in NSCLC. Experiments were conducted on a novel UK-based radiogenomics cohort. Joint pre- diction of KRAS and TP53 improved AUC from 0.58 to 0.64 for KRAS and from 0.69 to 0.71 for TP53. For the EGFR/KRAS pair, only EGFR benefited from joint learning, while no improvement was observed for the EGFR/TP53 pair. These findings demonstrate that the effectiveness of multi-label learning depends on the specific combination of gene mutations being modelled, suggesting that mutation-specific modelling strategies may be preferable for PET/CT radiogenomic prediction.
