ABSTRACT This proposal aims to develop an eXtended Modular ANthropomorphic (XMAN) digital phantom to simulate realistic patient images with both population-based and individualized motions to establish effective motion management in liver cancer radiotherapy (RT). With over 70% of liver cancer patients surgically unresectable, stereotactic body radiation therapy (SBRT) has become an essential non-invasive alternative treatment. However, despite its promise, liver SBRT has a local failure rate of over 20% at 2-3 years, grade 3+ toxicity over 20%, and resulted in deaths for over 10% of cirrhotic patients. Previous studies demonstrated strong associations between the treatment outcome and the RT delivery accuracy. Accuracy of liver SBRT is severely impaired by patient motion, such as breathing and daily motions. Current practice uses a large planning target volume (PTV) margin of 5-10mm to account for patient motion, which leads to increased dose to normal tissues increasing the toxicity and limiting dose escalation to increase tumor control. Lack of effective motion management is a key barrier to improving outcomes in liver SBRT. Several challenges exist for establishing effective motion management: (1) Lack of ground-truth (GT) for evaluation. Liver SBRT is mostly guided by cone-beam CT (CBCT), which has limited tumor visibility. Thus, no GT tumor volume can be defined in CBCT to evaluate the treatment errors. (2) Lack of effective optimization tools: In patient data, various factors affecting motion management compound one another and cannot be separated to be optimized individually. (3) Lack of individualized strategy. The current population-based approach is far from optimal for individual patients. Digital simulation provides GT images and motion patterns, effective tools for optimization with well- controlled parameters, and the possibility for patient-specific simulation. However, current phantoms lack realistic simulation of imaging and motions as well as patient-specific simulation. We will address these limitations by pursuing the following aims: Aim 1: Establish generic XMAN to simulate patient imaging and motions in the entire RT workflow. Aim 2: Establish Patient-specific Adaptive XMAN (PAXMAN) to customize motion management. Aim 3: Translate and apply XMAN and PAXMAN in RT to validate their clinical impact. The hypothesis is that PTV margin can be reduced to 3-5 mm while ensuring tumor coverage and minimizing normal tissue dose using XMAN. Deliverables: This project will deliver a suite of XMAN phantoms, the first of their kind, to simulate realistic patient images and motions in the entire RT workflow, providing essential tools to evaluate and optimize motion management. PAXMAN unlocks the potential to move from population-based heuristic approaches into high-precision patient-specific motion management. Our multi-institution and industry partners have a strong track record of collaboration, ensuring the success of this grant. Overal...