Progress in multimodal imaging assessment of the correlation between perivascular adipose tissue and atherosclerosis

Cardiovascular disease remains a leading cause of death and disability worldwide. Adipose tissue dysfunction is a key driver, and perivascular adipose tissue (PVAT) exerts location‑specific effects on the cardiovascular system. This review focuses on PVAT in atherosclerosis, with particular attention to pericoronary PVAT, epicardial adipose tissue (EAT), carotid PVAT, and peri-aortic adipose tissue (PAAT). PVAT homeostasis relies on UCP1‑mediated thermogenesis for metabolic modulation and on paracrine signaling via adiponectin and apelin. Under aging, obesity, or diabetes, PVAT switches from an anti‑inflammatory to a pro‑inflammatory phenotype, accelerating plaque formation through bidirectional inflammatory crosstalk. For clinical assessment, this review highlights progress in multimodal imaging technologies, including CT fat attenuation index, MRI water‑lipid separation, PET/CT, and echocardiography, for noninvasive quantification of PVAT density, composition, and metabolic activity, and compares their advantages and limitations. This review further discusses artificial intelligence (AI)-driven radiomic and radiotranscriptomic approaches that extend PVAT assessment beyond conventional single-parameter metrics. We systematically review the research progress of these techniques in evaluating dynamic PVAT changes and their correlation with atherosclerosis, aiming to clarify the intrinsic correlation between imaging‑derived PVAT functional status and AS development. Graphical abstract PVAT shifts from a protective, thermogenic depot to a pro-inflammatory driver of atherosclerosis with aging and metabolic disease, increasingly characterized by multimodal imaging and AI-driven radiomics.