Beyond the Diameter: 3D Surface Area Estimation for More Accurate Polyp Sizing

Accurate and reliable polyp size measurement is integral to effective risk stratification and surveillance in colorectal cancer screening. Traditional diameter-based assessments are prone to variability and fail to capture the complex, three-dimensional nature of these lesions. We present a novel pipeline that integrates state-of-the-art metric depth estimation, robust segmentation, and poisson surface reconstruction to estimate polyp surface area from monocular endoscopic images. By adopting a canonical camera space transformation and focusing on surface area rather than largest diameter, our approach provides a more anatomically faithful representation, and boosts zero-shot generalization in depth estimation. Although the concept of using surface area as a metric for polyp sizing is theoretical at this stage, our experiments on synthetic datasets demonstrate the feasibility of our approach and lay the groundwork for future, more clinically robust applications.