Multimodal Imaging Platform for Pre-Transplantation Donor Liver Viability Assessment - Project Summary/Abstract: The global shortage of donor livers poses critical challenges to transplantation, with over 30% marginal donor livers (70% of donation after circulatory death (DCD) livers) discarded due to unreliable viability assessment methods. Current biopsy-based approaches are invasive, limited to single tissue slices, thus fail to provide comprehensive evaluation. This project introduces the Composite Liver Viability Evaluation System (CLVES), a multimodal imaging platform combining polarization-sensitive optical coherence tomography (PS-OCT), optical coherence elastography (OCE), Doppler ultrasound (DUS), and photoacoustic tomography (PAT). By leveraging noninvasive imaging and machine learning, CLVES offers comprehensive structural, biomechanical, and functional assessments, providing accurate predictions of marginal liver viability to expand the donor pool and improve transplantation outcomes. Aim 1 uses PS-OCT and OCE to acquire spatial structural and biomechanical data from marginal donor livers, generating a robust dataset of imaging biomarkers for hepatic steatosis, fibrosis, and stiffness. This dataset will serve as a foundation for subsequent analyses and a benchmark for imaging-guided donor liver evaluation. Aim 2 employs PAT and DUS to collect structural and functional information during normothermic machine perfusion, capturing blood flow, fibrosis, lipid, and oxygenation parameters. Biochemical tests of perfusion agents, whole blood, and bile will generate biochemical index scores as the functional assessment gold standard. These data will be further validated by 3D imaging through three-photon microscopy (3PM) and pathological scoring at labeled sites. Aim 3 establishes correlations between the imaging biomarkers obtained in Aims 1 and 2 and pathological gold standards using machine learning. A multimodal fusion deep learning framework synthesizing multimodal imaging data will be developed to comprehensively predict liver viability. This aim provides critical insights into the relationships between structural, biomechanical, and functional features and liver viability. The proposed research is expected to significantly advance the field of organ transplantation by establishing a novel noninvasive platform to evaluate marginal donor livers. By delivering an integrated imaging system, validated biomarkers, and an AI-driven predictive tool, this project has the potential to expand the donor pool, reduce liver discard rates, and improve post-transplantation outcomes. The multidisciplinary team and robust validation strategy ensure a clear path toward clinical translation, addressing a critical unmet need in transplantation medicine.
