en Radiation Biology Radiation Biology تحقيق بديع Original Research <div style="text-align: justify;"><span style="font-size:10pt"><span style="text-justify:newspaper"><span style="text-kashida-space:50%"><span style="line-height:119%"><span style="font-family:Calibri"><span style="color:black"><span lang="en-US" style="font-size:9.0pt"><span style="font-family:Calibri"><span style="language:en-US">Predictive models have become essential tools in modern oncology, significantly advancing gastric cancer management&mdash;one of the most prevalent and lethal cancers worldwide. This systematic review investigates the application of predictive models derived from public databases, emphasizing their role in improving radiotherapy outcomes. Various modeling techniques are explored, including statistical methods like logistic regression and Cox proportional hazards models, machine learning approaches such as random forests and support vector machines, and deep learning models like convolutional and recurrent neural networks. These models contribute to early detection, prognosis estimation, treatment response prediction, and tumor classification. Notably, in the context of radiotherapy, predictive models enhance tumor delineation, assist in selecting optimal radiation doses, and forecast individual treatment responses, reducing toxicity and improving precision. Other key clinical applications include molecular subtyping, biomarker discovery, and image-based diagnostics, especially through endoscopic and histopathological image analysis. These applications support the development of personalized treatment regimens and improve long-term patient outcomes. Despite their promise, several challenges remain, including inconsistent or imbalanced data, limited interpretability of complex algorithms, and concerns regarding clinical trust and AI transparency. Addressing these issues requires the development of high-quality standardized databases, stronger data-sharing frameworks, the adoption of federated learning methods, and the integration of explainable AI models into clinical workflows. This review concludes that predictive models, when properly validated and implemented, hold substantial potential to transform gastric cancer radiotherapy by enabling more tailored, data-driven treatment strategies, ultimately improving survival rates and quality of life for patients.</span></span></span></span></span></span></span></span></span></div> Stomach neoplasms, radiotherapy, predictive value of tests, machine learning, clinical decision-making. 1101 1114 http://ijrr.com/browse.php?a_code=A-10-1-1442&slc_lang=en&sid=1 Sh. Chen 7900319475328460032829 7900319475328460032829 No Department of Gastric Surgery, Clinical Oncology School of Fujian Medical University, Fujian Cancer Hospital, Fujian Province, China Sh. Wei 7900319475328460032830 7900319475328460032830 No Department of Gastric Surgery, Clinical Oncology School of Fujian Medical University, Fujian Cancer Hospital, Fujian Province, China Y. Wang 7900319475328460032831 7900319475328460032831 No Department of Gastric Surgery, Clinical Oncology School of Fujian Medical University, Fujian Cancer Hospital, Fujian Province, China L. Chen 7900319475328460032832 7900319475328460032832 No Department of Gastric Surgery, Clinical Oncology School of Fujian Medical University, Fujian Cancer Hospital, Fujian Province, China Zh. Lin hxsf803@163.com 7900319475328460032833 7900319475328460032833 Yes Department of Urological Oncology Surgery, Clinical Oncology School of Fujian Medical University, Fujian Cancer Hospital, Fujian Province, China