Artificial intelligence in personalized oncology: current approaches and application prospects
Year: 2026
Type: article
Abstract: Background . Personalized oncology aims to select optimal treatment strategies by accounting for the molecular and clinical characteristics of the tumor, whereas traditional approaches based on histological type and stage do not always achieve sufficient efficacy. Objective . To systematize contemporary approaches to the application of artificial intelligence (AI) in personalized oncology, with emphasis on predicting tumor chemosensitivity and radiosensitivity and on the role of transcriptomic data in the development of predictive models. Methods . A narrative literature review was conducted in PubMed/MEDLINE and peer-reviewed scientific journals, analyzing publications on the application of machine learning and deep learning methods in diagnosis, prediction of disease course, selection of chemotherapy and radiotherapy, and integration of omics data in oncology. Results . The principal areas of AI application in oncology were reviewed, along with the types of data employed (medical images, clinical records, and genomic, transcriptomic, and proteomic profiles) and models for predicting sensitivity to chemotherapy and ionizing radiation; the role of transcriptomics was considered separately. Key limitations were identified: a shortage of high-quality data, heterogeneity of data formats, model overfitting, the difficulty of clinical validation, and ethical and regulatory issues. Conclusion . AI has become firmly established in the armamentarium of personalized oncology; however, realizing its potential requires systematic clinical validation of models, the development of decision support systems, and digital twins of tumors.
Authors: I. A. Solovev, D. A. Golubev
Institution: Syktyvkar State University
Language: English
Subfield: Health Informatics
Field: Medicine
Domain: Health Sciences
Open Access status: diamond
OpenAlex ID: W7225463926