SurgiCorr: A Physics-Based Computational Framework for Patient-Specific Surgical Corridor Planning in Brain Tumor Surgery.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
BACKGROUND AND OBJECTIVES: Transcortical corridor selection is guided by anatomic intuition and neuronavigation, with no quantitative tool to estimate retraction injury to brain tissue. Such injury depends on corridor-specific tissue properties, opening width, and hold duration, variables existing systems do not model. We present SurgiCorr, a preoperative framework that generates patient-specific corridor candidates, simulates each biomechanically, and ranks them by a composite risk score (CRS) integrating strain, viscoelastic dose, imaging burden, and ventricular proximity. METHODS: MRI volumes were resampled to isotropic resolution, segmented into 5 tissue classes, and meshed with per-element material properties. A surgeon-defined cortical entry patch was sampled to generate candidate corridors, each evaluated by parallel finite element analysis at 4 opening widths (4, 6, 8, and 12 mm). Four risk metrics were computed per corridor: volumetric cumulative strain injury (mm3), strain damage index, damage escalation metric (mm-1), and viscoelastic damage dose. These were combined with fluid-attenuated inversion recovery burden and ventricular proximity into a normalized CRS. RESULTS: In a pediatric supratentorial tumor and an adult with multiple deep-seated metastases, the pipeline generated 176 and 68 candidate corridors, and completed 704 and 272 finite element modeling solves in 68/9 and 28/3 minutes (central processing unit/graphics processing unit), respectively. CRS spanned 0.025 to 0.819 and 0.134 to 0.694 across the 2 cases (case means 0.395 and 0.379), respectively, with the highest-quartile corridors flagged AVOID in each. Viscoelastic damage dose separated the lowest- from the highest-risk corridor by 2.2- and 1.4-fold, respectively. Cortical heatmap produced spatially coherent risk zones. All risk labels are relative to the candidate distribution for the individual patient and do not imply an absolute threshold of mechanical safety. CONCLUSION: SurgiCorr demonstrates the feasibility of physics-based preoperative corridor planning that quantifies retraction injury as a patient-specific, time-integrated biomechanical dose. Prospective clinical validation against surgical outcomes is required before deployment.