An Exploratory Study on Developing an Integrated Approach Combining Multimodal Imaging and Multi-omics Characterization of Tumor Heterogeneity for Precision Diagnosis and Treatment Optimization in Liver Cancer.

Primary liver cancer, mainly including hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC), represents the third leading cause of cancer-related mortality. Enhancing the precision of liver cancer diagnosis and providing early therapeutic efficacy and prognostic evaluation during clinical decision-making hold significant clinical importance. Ultrasound is the preferred imaging modality for liver cancer screening. Contrast-enhanced ultrasound (CEUS) can dynamically visualize the microvascular perfusion of liver cancer lesions. Liver elastography has become a commonly used clinical assessment tool for cirrhosis. Photoacoustic imaging (PAI), an emerging non-invasive functional imaging technique, enables visualization of specific molecules through their spectroscopic characteristics at designated wavelengths. The objectives of this study include: (1) Conducting an observational investigation combining CEUS, elastography, and superb microvascular imaging (SMI) to collect imaging data; (2) Preserving tumor specimens from participants to investigate heterogeneous protein characteristics of primary liver cancer organoids using PAI; (3) Analyzing peripheral venous blood samples to study transcriptomic profiles. Artificial intelligence (AI) technology will be employed to establish models integrating ultrasound radiomics with tumor multi-omics characteristics, aiming to provide novel strategies for precision diagnosis and treatment of liver cancer. Key questions：(1) How to develop a multimodal imaging model combining CEUS, elastography, and SMI for predicting differentiation of liver cancer, microvascular invasion (MVI) and prognosis; (2) Whether PAI can identify heterogeneous proteins in liver cancer organoids through specific spectral recognition; (3) Whether AI can integrate multi-dimensional data to establish models based on ultrasound radiomics and multi-omics features.