Related Experiment Video
Updated: Jan 8, 2026
Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions
Published on: April 5, 2024
Preoperative selection of best-fit imaging targets and image-guided surgery improves tumor delineation and surgical
Nayeon Choi1, Dayong Shin, Han-Sin Jeong
1Department of Otorhinolaryngology-Head and Neck Surgery, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, South Korea.
Background:
Image-guided surgery (IGS) using biological markers enhances tumor eradication while preserving normal tissues. However, defining optimal imaging targets is challenging due to phenotypic changes that occur during tumor evolution or treatment. This study proposes a new surgical approach involving preoperative tumor analysis followed by individualized IGS and assesses its significance in surgical oncology.
Methods:
Using in-vivo tumor models (FADU and A253 tumors), preoperative tumor analysis was performed to assess the relative expression of various markers using immunohistochemistry. The markers with the highest signal-to-background ratios (SBRs) were selected for molecular imaging. The results of individual tumor biology-based imaging were compared to tumor pathology, and the accuracy of imaging was evaluated using a lattice grid segmentation method. In addition to treatment-naïve tumors, we simulated two in-vivo tumor models in the post-radiation and post-chemotherapy settings and investigated the diagnostic ability of individual tumor biology-based molecular imaging in these settings. Finally, the surgical outcomes of IGS were assessed.
Results:
In treatment-naïve tumors, integrin α5β3 and CA19-9 for FADU tumors and CEA and GLUT1 for A253 tumors had the best SBRs. Tumor imaging using these markers predicted tumor boundaries with a high concordance with those on tumor pathology. In the post-radiation and post-chemotherapy settings, different markers exhibited high SBRs in tumor analysis, indicating that tumor phenotypes changed after treatment. Similarly, tumor imaging using these selected markers improved the anatomical delineation of tumors. Surgical completeness and outcomes (recurrence) were significantly better in the individual tumor biology-based IGS group compared to the conventional (unaided) surgery group.
Conclusions:
This proof-of-concept study demonstrates that tumor biology-based molecular imaging is feasible and can accurately delineate tumor boundaries. Preoperative analysis allows for the selection of optimal imaging targets with high SBRs, potentially enhancing surgical precision. However, its generalizability and clinical relevance require validation across multiple tumor models with sufficient sample size, extended follow-up, and clinical studies.
