Recent advances in large models demonstrate significant prospects for transforming the field of medical imaging. These models, including large language models, large visual models, and multimodal large models, offer unprecedented capabilities in processing and interpreting complex medical data across various imaging modalities. By leveraging self-supervised pretraining on vast unlabeled datasets, cross-modal representation learning, and domain-specific medical knowledge adaptation through fine-tuning, large models can achieve higher diagnostic accuracy and more efficient workflows for key clinical tasks. This review summarizes the concepts, methods, and progress of large models in medical imaging, highlighting their potential in precision medicine. The article first outlines the integration of multimodal data under large model technologies, approaches for training large models with medical datasets, and the need for robust evaluation metrics. It then explores how large models can revolutionize applications in critical tasks such as image segmentation, disease diagnosis, personalized treatment strategies, and real-time interactive systems, thus pushing the boundaries of traditional imaging analysis. Despite their potential, the practical implementation of large models in medical imaging faces notable challenges, including the scarcity of high-quality medical data, the need for optimized perception of imaging phenotypes, safety considerations, and seamless integration with existing clinical workflows and equipment. As research progresses, the development of more efficient, interpretable, and generalizable models will be critical to ensuring their reliable deployment across diverse clinical environments. This review aims to provide insights into the current state of the field and provide directions for future research to facilitate the broader adoption of large models in clinical practice.
Humans ; Diagnostic Imaging/methods* ; Precision Medicine/methods* ; Image Processing, Computer-Assisted/methods*

Objective:To evaluate the accuracy of variation of carotid artery hemodynamic parameters combined with passive leg raising (PLR) test in predicting supine hypotension syndrome (SHS) after spinal anesthesia in the patients undergoing cesarean section.Methods:Sixty-four parturients who were at full term with a singleton fetus, at 37-42 weeks of gestation, aged 18-40 yr, with body mass index of 18-30 kg/m 2, of American Society of Anesthesiologists physical status Ⅰ or Ⅱ, undergoing elective cesarean section, were enrolled in this study.The variation of carotid artery diameter (ΔD), variation of velocity time integral (ΔVTI), and variation of carotid blood flow (ΔCBF) before and after PLR were measured using ultrasound.Patients were divided into SHS group and non-SHS group (NSHS group) according to whether SHS after spinal anesthesia occurred.Pearson correlation was used to analyze the correlation between ΔD, ΔVTI, ΔCBF and systolic blood pressure (SBP) after spinal anesthesia.The receiver operating characteristic curve was used to assess the accuracy of ΔD, ΔVTI and ΔCBF in predicting SHS. Results:ΔVTI was negatively correlated with SBP after spinal anesthesia ( r=-0.539, P<0.01), ΔCBF was negatively correlated with SBP after spinal anesthesia ( r=-0.475, P<0.05), and ΔD had no correlation with SBP after spinal anesthesia in group SHS ( P>0.05). The critical values of ΔCBF, ΔVTI, and ΔD combined with PLR in predicting SHS after spinal anesthesia were 15.5%, 10.1%, and 6.0%, respectively, the sensitivity was 92.9%, 57.1%, and 96.4%, respectively, and the specificity was 53.1%, 81.2%, and 75.0%, respectively, and the areas under the curve were 0.873, 0.681 and 0.846, respectively. Conclusion:The ultrasound-measured ΔCBF and ΔD of carotid artery combined with PLR can be used as a reliable method to predict SHS after spinal anesthesia in the patients undergoing cesarean section, and the ΔCBF combined with PLR has a higher accuracy.