CT-based peritumoral radiomics nomogram on prediction of response and survival to induction chemotherapy in locoregionally advanced nasopharyngeal carcinoma

• DOI: 10.1007/s00432-023-05590-5
• Publication Date: 2024-01-29T17:03:26Z
• Authors: Fanyuan Zeng; Zhuomiao Ye; Qin Zhou
• Keywords: Nomogram; Concordance; Induction chemotherapy

Abstract Purpose The study aims to harness the value of radiomics models combining intratumoral and peritumoral features obtained from pretreatment CT predict treatment response as well survival LA-NPC(locoregionally advanced nasopharyngeal carcinoma) patients receiving multiple types induction chemotherapies, including immunotherapy targeted therapy. Methods 276 LA-NPC (221 in training 55 testing cohort) were retrospectively enrolled. Various statistical analyses feature selection techniques applied identify most relevant features. Multiple machine learning trained compared build signatures for each region, along with a clinical signature. performance model was evaluated using different metrics. Subsequently, nomogram constructed by best-performing models. Results In cohort, exhibited an AUC 0.816, outperforming other model’s calibration curve showed good agreement between predicted observed outcomes both sets. When predicting survival, concordance index (C-index) 0.888 cohort 0.899 indicating its robust predictive ability. Conclusion conclusion, combined model, incorporating features, outperformed chemotherapies. These findings highlight potential utility suggesting individualized planning decision-making.