Accurate assessment of residual liver regeneration following hepatocellular carcinoma resection remains a major challenge due to the dynamic nature of the regenerative process and limitations in conventional image-based evaluation. This study develops a three-dimensional (3D) bioprinted personalized repair scaffold based on multimodal fusion of computed tomography (CT) and magnetic resonance imaging (MRI). Multi-temporal CT/MRI images of patients with hepatocellular carcinoma were acquired, registered, and preprocessed using standardization. A dual-channel deep learning network was employed to achieve high-precision liver segmentation and multimodal fusion 3D reconstruction. Using the deformation field from non-rigid registration, the dynamic changes in residual liver volume after surgery were accurately tracked, and the geometry of the resected cavity was extracted to serve as the basis for scaffold design. Subsequently, the internal pore gradient was optimized by combining regeneration rate data. Using a gelatin methacryloyl (GelMA)/hyaluronic acid (HA) composite bio-ink, a repair scaffold with a personalized shape and functional internal structure was manufactured via digital light processing photopolymerization 3D printing. Experiments showed that, in terms of structural fidelity, the GelMA/HA composite scaffold exhibited excellent performance in overall shape fidelity (94.82%) and key structural wall thickness deviation (12.3 μm). Regarding in vitro cell compatibility, the relative cell proliferation rate achieved 0.85 ± 0.04 under low-serum conditions. In vitro experiments have shown that the scaffold has good cell compatibility and structural stability, providing a theoretical and technical approach for precise repair after liver cancer surgery.