Digital twins in radiology: A systematic review of applications, challenges, and future perspectives

Background

Digital twins (DTs) represent a transformative advancement in radiology, integrating multimodal imaging, artificial intelligence (AI), and computational modeling to create dynamic, patient-specific virtual representations.

Methods

This systematic review evaluated DTs applications across different imaging modalities. A total of 24 studies were analyzed, encompassing abdominal, musculoskeletal, interventional, dental, head and neck, cardiothoracic, breast, and general radiology. QUADAS-2 tool was used to assess risk of bias and applicability evaluation of the included studies.

Results

Key findings highlighted the role of DTs in predicting disease risk, optimizing therapies, and improving diagnostic accuracy, with applications including portal hypertension, scoliosis progression, liver ablation, brain tumor characterization, and noninvasive cardiothoracic assessment. Broader uses in general radiology included predictive modeling, automated dosimetry, and radiographer training. DTs are increasingly applied in major clinical domains such as oncology, cardiovascular imaging, and hepatic surgery, underscoring their emerging translational impact in radiology.

Conclusions

While DTs demonstrate significant potential, challenges such as computational demands, data integration, and clinical validation persist. Future research endeavors should focus on standardization, improved computational efficiency, and interdisciplinary collaboration to advance DTs from research to routine clinical practice, heralding a new era of personalized radiology.